IX. Chapter Nine: OIG Analysis of Broader Problems with the INSs Implementation and Use of IDENT
A. The Design of IDENT
When IDENT was conceived, Congress and the INS recognized its importance in identifying and tracking recidivists who repeatedly crossed the border. However, they also noted its importance in identifying criminal aliens. In 1989, when Congress first appropriated money to the INS for a biometric database system, it stated that the funding was designed to address the problem of the revolving door of aliens crossing the border as well as to immediately identify those criminal aliens who should remain in the custody of INS. The INS noted that IDENT could provide the capability for identifying aliens who had been apprehended previously, aliens who were suspected of criminal activity, and aliens with outstanding warrants. In 1996, when Congress provided more money to expand the use of IDENT, it directed the INS to
operate a criminal alien identification system . . . to assist Federal, State, and local law enforcement agencies in identifying and locating aliens who may be subject to removal by reason of their conviction of aggravated felonies, subject to prosecution . . . [because they are] not lawfully present in the United States, or otherwise removable. Such system shall include providing for recording of fingerprint records of aliens who have been previously arrested and removed into appropriate automated fingerprint identification systems.
The INS identified IDENT as the system serving this purpose.
Yet, several critical decisions in the design of IDENT demonstrate that the INS viewed IDENT primarily as a tool for the Border Patrol to identify and track recidivist border crossers, not as a tool for identifying or locating criminal or wanted aliens. First, the INS decided to take two fingerprints of persons enrolled in IDENT, rather than ten fingerprints. The INS concluded that fingerprint checks of the thousands of aliens who were apprehended each day needed to be done quickly, in two minutes or less, to avoid the problems and expense of detaining aliens for longer periods of time. As a result, a two-print rather than a ten-print system was adopted.
Second, as noted earlier, when the IDENT lookout database was initially deployed, the INS decided to populate it with records on a day- forward basis. The INS records in WIN and some INS records of bad boys informally maintained in Border Patrol sectors were placed into the lookout database when IDENT was first deployed, but the INS made no systematic effort to include lookout records for those aliens who had been previously deported or who had extensive criminal records. We had difficulty finding anyone in INS who could tell us the exact basis for this decision or what analysis was conducted regarding the costs and benefits of including previous records into the lookout database. According to Tim Biggs, the Chief of the INS Biometric Information Systems Section, the INS made the day-forward decision because it concluded that the cost of finding and placing into the lookout database the fingerprints of previously deported aliens or criminal aliens would be too much. He estimated this would have cost $12 to $14 million dollars, although he and others in the INS were not certain on what that estimate was based or what it would have bought. Others suggested that the earlier fingerprint records were of poor quality. But it is not clear that the INS ever carefully examined this issue in terms of the costs and benefits of including past records in the lookout database.
Third, the INS did not integrate IDENT with other INS or criminal databases that were text-based as opposed to based on biometrics, and the INS did not reach any agreement to integrate IDENT with criminal biometric databases once they became operational. The INS determined that it was too difficult to link text-based databases, where there is uncertainty about the identity of persons who gave varying names and birth dates, with biometric databases, where an aliens identity is established through fingerprints. And as we describe in Chapter X, when the FBIs automated fingerprint identification system, IAFIS, became operational in July 1999, there was no agreement on integrating IDENT and IAFIS. The current integration plans are only in the preliminary stages and will take several years to implement.
The INSs design of IDENT reflected the predominant sentiment within the INS that IDENT was primarily a system for identifying and counting recidivists. Certainly, this is a critical function. Before IDENT was implemented, the INS had virtually no way to identify recidivist illegal border crossers. The INS used paper forms that relied on the names aliens provided and that were scattered in various Border Patrol stations. That information was virtually useless in determining how often the alien had been apprehended previously. Unless the aliens admitted to previous apprehensions, or the same Border Patrol agent remembered previously apprehending the aliens, it was unlikely that the Border Patrol could identify how often the aliens had been apprehended. With IDENT, the number of times aliens have been apprehended can be identified and recidivist aliens can be prosecuted. According to INS officials, IDENT has reduced recidivism because of this deterrent effect.
In addition, IDENT provides a way to analyze information and develop intelligence on patterns of alien border crossings, allowing the INS to allocate its resources more efficiently. IDENT also has demonstrated its usefulness as a tool to help identify alien smugglers amongst groups of apprehended aliens by linking the same individual to separate groups of aliens.
We agree that IDENT provides an important and useful tool for the INS, particularly for the Border Patrol. But we believe that the INS devoted insufficient effort to ensuring that it was also being used throughout the INS to identify criminal or wanted aliens, which is another important objective of IDENT. Outside the Border Patrol such as in the Investigations and Intelligence Divisions INS employees did not view IDENT as a tool that could assist their work. We found that many supervisors in program offices outside the Border Patrol never required IDENT to be used or monitored its usage. They viewed IDENT solely as a tool for the Border Patrol, not for their offices.
B. The INSs Implementation of IDENT
We interviewed various INS managers and employees to understand how the lack of interest in IDENT outside the Border Patrol came about. In 1994, when Congress provided substantial funding for IDENT, the INSs Office of Information and Resource Management (OIRM) was given major responsibility for the development and deployment of IDENT. In 1995, OIRM invited representatives from all INS enforcement programs including representatives from Investigations, Intelligence, Detention and Deportation, and the Border Patrol to participate in a review of the IDENT project plan.
In approximately August 1995, the INS formed the Biometric Users Group (BUG), composed of representatives from all INS components. The main purpose of the group, which met quarterly until January 1998 when it was disbanded, was to develop requirements for IDENT based on INS users input. According to Tim Biggs, BUG was not successful because the INS components rarely sent the same representatives to the meetings, and much time was spent explaining past decisions to new representatives.
Biggs also stated that program offices, such as the Investigations Division, were not greatly involved in the development of IDENT. Biggs initially told us that OIRM was open to Investigations suggestions about IDENT. Yet, as we discussed the matter in more detail, Biggs acknowledged that the Investigations Division may have formed the impression that OIRM was not receptive to Investigations suggestions because OIRM frequently did not adopt them. Biggs stated that IDENT was not designed for Investigations or Intelligence functions.
Ronald Collison, the Associate Commissioner for OIRM, told the OIG that all INS programs were involved with the development of IDENT, but he admitted that OIRM viewed the Border Patrol as the primary user of IDENT. For example, when the program offices were discussing whether to use a two-print or ten-print system, Collison stated that all the program offices were in favor of a two-print system, except for Investigations. Thus, Collison said that when the ten-print concept was rejected, Investigations might not have perceived the advantages to them of using IDENT. He stated that any requirements articulated by investigators that were related to the ten-prints would have been excluded from IDENT. Similarly, he stated that IDENT was not marketed to Investigations and that OIRM only talked about the lookout database when discussing response times to fingerprint queries.
When we interviewed Investigations personnel, many told us that they could not identify anyone from Investigations who had been involved in the development of IDENT. During our review, however, we did find several Investigations employees who were involved in a limited way. For example, one Investigations employee was assigned to participate in the IDENT user group, but he stopped attending after going to three or four meetings. He told us that once the decision was made that IDENT would only capture two-prints rather than ten-prints of apprehended aliens, IDENT was no longer of interest to Investigations. Several other Investigations employees confirmed his assessment. During our review, we heard repeatedly from Investigations personnel that the standard for law enforcement organizations is a ten-print system and that comparing ten fingerprints is the only accepted way to make a positive identification for prosecutions.
One Investigations employee said that whenever Investigations personnel attended IDENT user-group meetings, they were told that IDENT was for recidivists and bad boys, and not something that Investigations would be interested in using. Another Investigations employee who was knowledgeable about the development of IDENT said that Investigations was not involved in designing it that the Border Patrol was the user of IDENT and the only INS component involved in its development.
As a result, the Investigations Division regarded IDENT as an OIRM and Border Patrol initiative. According to the head of the Investigations Division, it largely opted out of using it. He reported that the Investigations Divisions association with IDENT can probably be characterized more as observer than participant.
The lack of involvement or commitment to IDENT by program offices such as the Investigations Division was evidenced in various ways. When IDENT was deployed in INS District Offices, the Investigations Division did not carefully consider how INS investigators could use it. Even when INS investigators received training on IDENT, it was not geared towards their duties. And unlike in the Border Patrol, no one in Investigations Headquarters (or in the District Offices) required investigators to use IDENT or monitored their usage. Often, IDENT machines were installed in Investigations offices, used briefly or sporadically, and then not used at all.
Several Houston investigators told the OIG that the IDENT machines in the INSs Houston Investigations office did not work. A supervisory investigator in the Houston office said that there were many problems with the IDENT equipment, and it appeared to be out of service more than it was in service. Another investigator said he tried to get IDENT to work but could not. He said that when the contractor came to train the agents, the trainer got the system to work in training mode but stated that the system was overloaded and operating beyond capacity. The agent said he thought IDENT was defective, so he did not rely on it or use it. Other Houston investigators echoed this sentiment. According to a Houston Investigations supervisor, by September 1998 the Houston IDENT machines were no longer operational and no one in Investigations used them. But instead of attempting to get any defective machines repaired, Houston investigators simply ignored IDENT.
We also found that INSs Intelligence offices were not involved in the development or deployment of IDENT. The INS Acting Assistant Commissioner for Intelligence told us that the information collected by IDENT is not useful to the type of intelligence work done by his office. He said that it was his understanding that the Border Patrol or OIRM was responsible for IDENT. We also found a lack of knowledge about IDENT among the INS Intelligence officers to whom we spoke in the field. In the Houston District Office, the Intelligence officer who compiled the Weekly Intelligence Reports, including the report containing information about Resendez, said he had no working knowledge of IDENT.
After IDENT was deployed, some INS officials raised concerns that OIRMs control over the project had created problems for program offices. Notes from a meeting conducted in May 1997 about the INSs quarterly Review of Priorities stated that the [l]ack of communication and policy coordination between [O]IRM, ENFORCE, the field, and Programs, continues to plague the IDENT project. These issues have caused IDENT deployment to be behind schedule, and adversely affected usage rates. Because of this and other difficulties identified with IDENT, the notes contained a recommendation that a cross-programmatic IDENT Coordinator be empowered to finalize policies and coordinate implementation.
In response to this recommendation, the INS Commissioner directed the INS Office of Policy and Planning (OPP) to recommend a solution. OPP convened several meetings with representatives from various INS offices, including Field Operations, Investigations, Inspections, Border Patrol, and OIRM. OPP then drafted a seven-page document labeled DRAFT, dated August 11, 1997, called IDENT Coordination Proposal. In this document, OPP identified concerns over the coordination and decision-making issues and problems surrounding IDENT implementation. It identified additional concerns about IDENT, such as the lack of communication between parties, poor data integrity, and inconsistent data and technical policies. In addition, many INS offices have expressed concern over the need for an individual empowered to make timely decisions on policy-oriented issues affecting more than one office [emphasis in original]. The proposal recommended with regard to IDENT coordination, There needs to be an immediate, temporary, structure to address the urgent IDENT problems and concerns. It also recommended that a more institutionalized, long-term, structure [be] developed within the INS to ensure future biometrics coordination. The proposal suggested that the IDENT coordinator should coordinate and monitor the many arms of INS working on the IDENT project. Moreover, due to the urgent, sensitive, and cross-programmatic nature of the current IDENT issues, it is recommended that the IDENT coordinator report directly to the Deputy Commissioner, or at a minimum, to the newly created EAC for Programs and Policy [emphasis in original]. However, we were told by an INS official that these recommendations were never implemented.
C. IDENT Training
1. Training Recommendations in the OIGs 1998 Inspection Report and the INSs Response
In its 1998 Inspection report on IDENT, the OIG identified various problems with its implementation and use. Among other issues, the OIG report suggested that the problems in IDENT usage by INS personnel were compounded by a lack of IDENT training provided to INS field personnel. The report noted that few INS agents and even fewer supervisors were trained to use IDENT. The report stated that many employees were not familiar with the fingerprint match confirmation process or with other features of IDENT. The report recommended that the INS needed to develop and implement a strategy for sufficiently training all INS personnel using IDENT.
In a memorandum responding to our training recommendation, the INS Office of Management only partially addressed the need for such training. This memorandum stated:
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.
The INSs formal response to the OIG claimed that Border Patrol agents are becoming more comfortable using IDENT in day-to-day operations. The response said that the INS planned to incorporate IDENT training into the course work at the INS academies, and that the INS planned to continue using the train the trainer concept upon initial deployment of IDENT and thereafter with each system upgrade. The response also noted that the INS had developed a Web Site on the INS Intranet that provided training information on IDENT.
On March 24, 1998, the OIG replied that the INSs approach to training was reasonable but incomplete. We noted that the INS response did not address how the INS would evaluate the effectiveness of training. We pointed out that the train the trainer concept had had limited success because not all agents were trained and those that had received IDENT training did not have to demonstrate that they had mastered the basic skills. We recommended that the INS develop a method to ensure that its employees successfully learned how to properly use IDENT.
On December 21, 1998, the INS responded to open OIG recommendations, including the training recommendation. The INS memorandum again stated that training on IDENT was provided with all installs and system upgrades. The memorandum added that OIRM had arranged individual training sessions for as many participants as site management needed. The memorandum noted that the training was evaluated through course evaluation forms completed by training participants and that OIRM surveyed personnel in charge of sites regarding the training. The memorandum acknowledged that while these survey methods were anecdotal, the INS believed that the surveys could pinpoint issues that the users of IDENT believed required attention.
On March 9, 1999, the OIG replied that the INSs response still did not address how it would evaluate the effectiveness of IDENT training and that the INS proposals did not provide a direct measure to ensure that INS officers successfully learn how to properly use IDENT.
The next response the OIG received from the INS about the OIG Inspection reports training recommendations was dated July 27, 1999, after the controversy over the Resendez case arose. A memorandum written by the INS Executive Associate Commissioner for Management stated that IDENT terminals were scheduled to be installed in the INS training academies in September 1999 and would be used to provide IDENT training as part of the academy experience. The response said that OIRM would monitor IDENT usage in the field, including periodic assessments at field locations. The response further claimed:
The OIG report findings are based upon a review by their Inspections Division conducted in 1997 at a time when IDENT was not being used consistently. Today IDENT is being used where deployed as an integral part of INS enforcement operations. Agents and their supervisors have a much greater awareness of the IDENT processes for capturing biometric information for aliens apprehended. We believe that IDENT is being used much more effectively. While training does not represent the problem it did 2 years ago, as outlined in the OIG report, it is still an area in which we will strive to make improvements.
Finally, in response to our questions during the Resendez review about the training provided to INS employees, the INS responded in writing that INS employees had received on-the-job training in the field or more formal training when IDENT was first installed and during upgrades. The INS response stated that self-paced training was available to INS employees through the IDENT site on the INSs World Wide Web site. The INS noted that the Border Patrol Academy has not yet established a course of training for students on IDENT. The INS did not include any information about new officer training on IDENT for offices outside the Border Patrol.
2. The INS Failed to Train INS Employees Adequately on IDENT
As we learned through our review of the Resendez case, the INSs claims about the effectiveness of its training were wrong. Certainly, the INS had allocated significant resources for IDENT training to its employees, particularly in the Border Patrol. But the training was ineffective. When we interviewed INS employees in various offices connected to the Resendez case, their knowledge about IDENT was still severely lacking. Many INS employees with whom we spoke, particularly those outside the Border Patrol, had little knowledge about IDENT. Most INS employees we interviewed were unaware of the policies or procedures related to IDENT lookouts. Even those in the Border Patrol who regularly processed apprehended aliens had significant gaps in their knowledge about IDENT and the procedures that should be used.
We found that the INS placed the initial training emphasis on the Border Patrol. Less emphasis was placed on IDENT training for the Investigations and Detention and Deportation (D&D) offices. The training that was provided to Investigations and D&D employees did not explain the benefits of IDENT to their work or how they could effectively use IDENT to advance their missions. And few INS Intelligence employees at Headquarters or in the field, who could potentially play a role in suggesting ways to locate aliens such as Resendez, ever received IDENT training.
We also found that when training was provided, neither OIRM nor INS managers ensured that the employees who received the training actually understood how to use IDENT. This was particularly important given that the INS adopted a train the trainer approach, in which some agents were trained in IDENT and were expected to inform other agents how to use it.
We attempted to review in detail the training that was given in the Border Patrol sectors where Resendez was apprehended in 1998 and 1999 (El Paso and Del Rio), in the Houston Investigations office (the office that was contacted after the murder of Dr. Benton by law enforcement authorities who were searching for Resendez), and to INS Central Region intelligence officers. The following sections describe what we found in those offices.
a) Training in the El Paso and Del Rio Border Patrol Sectors
Training records provided to us by the INS indicate that it made a significant effort to train Border Patrol agents in the El Paso and Del Rio Border Patrol Sectors. In the Del Rio Sector, INS contractors provided training on IDENT to 193 out of approximately 550 Border Patrol agents in 1996, shortly after IDENT was deployed there. In addition, a one-day training course on ENFORCE (which included basic training on IDENT) was provided to 171 Del Rio Border Patrol agents in 1998, and a two-day ENFORCE upgrade training course was provided to 31 Del Rio Border Patrol agents in 1999. In the El Paso Sector, INS contractors provided training on IDENT to 127 out of approximately 700 Border Patrol agents in 1995 and 1996; one-day ENFORCE training was given to 278 agents in 1997; and two-day ENFORCE training to another 215 agents in 1997.54
In total, approximately 80 classes were given to El Paso Border Patrol agents and 55 classes to Del Rio Border Patrol agents. The classes were small and usually included 6 to 10 trainees. According to the IDENT training instructors and their training materials, the IDENT training consisted of a 90-minute to 2-hour course covering how to enroll persons into the IDENT recidivist database; how to search for a prior IDENT record by FINS or fingerprints; how to access, read, and work with IDENT transactions and enter new ones; how to use IDENT menu listings; and how to verify that the alien being processed was in fact the same person as the person in an IDENT record. The instructors guide contained a section defining a lookout and explaining how an agent should react to a lookout. There was no training on how to have a lookout entered in IDENT.
However, this training was frequently affected by problems because of a lack of planning and coordination among the various contractors. A Border Patrol official who was involved in a working group that planned IDENT training stated that OIRM did a poor job of coordinating the set-up activity among the installers, local personnel, and trainers. She said, for example, there was confusion about who would ensure that IDENT had been installed and was working properly when the contractors were sent to train Border Patrol agents. As a result, the training equipment was often not set up on time or was not working properly when training began. This official stated that she had raised this issue on many occasions over the past three years to two OIRM officials who were part of the working group, but that neither of them addressed the problem. She said that the remainder of the attendees of the working group were the contractors who were the source of the problem.
As a result, the training was less effective than it could have been. The trainers notes we received from the INS contain many examples of technical problems that precluded training altogether. For example, the notes reflect sessions when the class was unable to log onto ENFORCE for the entire training day; system breakdowns that ended classes; ENFORCE and IDENT lock-ups; and occasions when the entire ENFORCE system was inaccessible.
Border Patrol agents confirmed to us that these training problems existed, describing occasions when due to technical problems only one IDENT terminal was working and all of the trainees were required to gather around the trainer and observe as she worked on a single terminal. The agents said that it was much more difficult to learn under these circumstances because they did not have the opportunity to practice.
Most significantly, once the training was provided, neither OIRM nor the Border Patrol offices took any action to confirm that the Border Patrol agents actually understood the most important concepts of the training. This was particularly important because the majority of Border Patrol agents were not trained directly and had to learn about IDENT from their colleagues. Less than half of the employees in the Del Rio and the El Paso Sectors received the initial training directly. Due to the high turnover among Border Patrol agents, the vast majority of current agents in Del Rio and El Paso have not received any direct training. This resulted in widespread confusion among the agents with whom we talked about what a lookout is, who enters them in the lookout database, how they are entered, and what action an agent should take in response to a lookout.
b) Training in the Houston Investigations Office
In October 1997, an INS contractor provided 4 classes that included IDENT training in the Houston District Office for 10 agents in Investigations, 3 Investigative Assistants, and 12 D&D employees. In April 1998, INS contractors trained another 11 agents in Investigations, 6 Investigative Assistants, and 9 D&D employees. In June 1998, INS contractors provided a four-day training course on ENFORCE to 7 agents in Investigations, 6 Investigative Assistants, and 9 D&D employees.55 Among the Houston special agents who received IDENT training were Kelly Dozier, David Jennings, and Susan Friesenhahn.
According to the evaluations they completed, most of the IDENT trainees in the Houston Investigations office rated the training very highly in terms of the information that was conveyed. The trainers told us that the training included basic information about lookouts, such as how to recognize a lookout, what a lookout indicates, and how to react to lookouts. While the trainers did not have the appropriate equipment to demonstrate an actual lookout, they described and provided illustrations of lookouts. However, because the IDENT portion of the training in Houston was developed primarily for the Border Patrol, it did not provide any in-depth discussion of how IDENT could specifically benefit Investigations employees in their work.
At least one trainer recognized the problem of universal training for different groups of employees. She wrote in her training notes that the different groups process aliens in different ways and the training should be geared towards their particular work. For example, the training did not convey how investigators could use IDENT to support additional charges against aliens, such as reentry after deportation. The training did not specifically discuss how to enter an alien in the lookout database. Border Patrol agents, to whom the training was geared, usually had custody of the individual being processed. In contrast, investigators might want to enter a lookout for an alien who was not in custody but who the investigators were seeking. This can be done by sending a ten-print card to the BSC. But the Houston investigators to whom we spoke about IDENT did not understand that they could initiate a lookout without having an alien present.
The trainers who conducted the Houston training told us they were unaware themselves of the precise procedures for enrolling lookouts in IDENT. One said that she always informed the students that they had the ability to enroll aliens in the lookout database, but they needed to determine the precise procedures from their supervisors. Yet, Investigations supervisors in the Houston District Office acknowledged to us that they did not understand the IDENT system and, consequently, could not have provided any further guidance.
We do not credit the claims made to us by Houston investigators who attended IDENT training that they never received any training about lookouts. While it is possible that some investigators were not told how to enroll aliens into the IDENT lookout database, we concluded that IDENT training informed them in general about the existence and use of the lookout database, since the emphasis in IDENT training was on its ability to identify previously apprehended aliens (through the recidivist database) or certain criminal aliens (through the lookout database).
c) Training for Intelligence Employees
Intelligence officers were not trained in IDENT and did not use it in their work. There is no standard training program for the various individuals working in Intelligence positions within the INS and Border Patrol, but the limited training provided to Intelligence officers did not include any discussion of IDENT.56 None of the field intelligence personnel to whom we spoke had been apprised of the potential of IDENT to Intelligence work. The outgoing Chief of the Operations Branch, Headquarters Intelligence, who was responsible for developing Intelligence training for INS Investigations, Inspections, and Border Patrol personnel, confirmed that training for these field offices does not include anything about IDENT. The INS Acting Assistant Commissioner for Intelligence told the OIG that he was only peripherally aware of IDENT.
Intelligence personnel in the field told us that since they were not required to learn about the IDENT lookout database, they had not done so. For example, the intelligence officer for the Houston District Office said that he missed the IDENT training when he was a special agent and had never been trained on IDENT since that time. Intelligence officers in the Del Rio and El Paso Border Patrol Sectors were not familiar with IDENT and its potential benefits to the Intelligence function. Central Region Intelligence officers said they had never been trained in the capabilities of IDENT. As a result, the only method intelligence officers used to warn Border Patrol agents or investigators of dangerous aliens was to send bulletins by electronic mail or to create posters with the aliens photograph that could be posted at a central location within the Border Patrol station or checkpoint. This was the suggestion offered but never followed through with by the Central Region Intelligence officers in the Resendez case. Not surprisingly in view of their lack of training, they did not think to use IDENT to notify INS employees about Resendez.
54 Some employees participated in both IDENT and ENFORCE training. The principal focus of ENFORCE training was on entering and retrieving biographical information about aliens. ENFORCE training did not discuss entering lookouts. The training was provided before ENFORCE machines were available in the Houston office, and even the trainers recognized that training the agents who did not yet have ENFORCE would be problematic, since they would likely forget what they learned before they had the opportunity to use it.
55 Some of the same employees attended both the IDENT and ENFORCE training.
56 The INS Acting Assistant Commissioner for Intelligence told us that there is no method for identifying which Intelligence personnel in the field need training on intelligence duties. He also said that because some of these employees perform their Intelligence duties as a collateral function, they may not receive any intelligence training at all.