3. Allegations that agents were not permitted to respond to sensor hits

During the investigation, the OIG also received allegations from witnesses that agents were not permitted to respond to sensor hits. The allegations took three forms: 1) Some said supervisors instructed agents to ignore hits and tell Sector Communications that there were no agents available to respond when there had in fact been available agents; 2) some said agents were given sensor maps indicating that agents should never respond to certain sensors; and 3) some witnesses said agents were instructed to respond "10-4" to Dispatch as if they were responding to the sensor, but not actually go to the area of the sensor to search for alien traffic. These actions allegedly were designed to artificially limit apprehensions or lower gotaway estimates.

a. Alleged orders to ignore sensor hits

The OIG received several broad allegations that supervisors had precluded agents from responding to sensors because they did not want apprehensions. Bonner alluded to this allegation during the San Diego radio program when he said agents are "being forced to sit in stationary positions on the line and not allowed to go respond to the sensors." Out of 118 responses on this subject from people at Imperial Beach, Chula Vista, and Brown Field, the OIG received eight claims that agents were not allowed to respond to sensors even if available. The OIG received no such complaints from the other stations.

i. Imperial Beach

The OIG received three complaints at Imperial Beach that agents were being told to ignore sensors even when adequate manpower to respond was available. One agent said he had gone to investigate a sensor "200 yards from his `X,'" but a supervisor arrived, asked him where his `X' was, and told him to get back to his area. When he asked the supervisor, "Are you telling me you don't want me to respond to sensor hits and work traffic?" the supervisor replied, "Yes, that's exactly what I am telling you." The agent returned to his position.

Another agent generally alleged that agents would try to respond when Dispatch called out sensor hits, but supervisors would tell them not to. This witness cited no specific instances of this problem and testified that he frequently left his X to apprehend aliens even though he was not supposed to. In addition, a Union official reported hearing that an agent assigned to a particular X had not been allowed to respond to a sensor in an adjacent area to the north even though he "had time." He testified further that agents were upset because they had to stay in high-visibility positions while there was traffic activating sensors to the north.

In contrast to these three witnesses, 43 witnesses at Imperial Beach, including 27 line agents, testified that agents generally were allowed to respond to all sensors. Indeed, 35 said they had never been ordered to ignore a sensor or had never ordered someone to ignore a sensor. The exceptions noted were when there was a shortage of manpower or the sensor was out of the agent's area. While most of the supervisors testified that they had never told an agent he could not respond to a sensor hit, several said they did not allow agents assigned to fixed positions to leave their area to respond to sensors. One noted that he had told agents they could not respond to a hit because there was no one to stay at the X while the agent responded. Several line agents recalled having been told that if they left their assigned areas, they would be leaving a hole in the first line of defense. Significantly, no witness claimed that he could not respond to sensor activations within his assigned area.

In essence, the claims of the complaining agents were factually consistent with the accounts of those who did not complain. The only real difference was in the refusal or inability of the complainants to accept or understand the Gatekeeper strategy, which called for agents to maintain high-visibility positions at designated posts. Because we have found that the high-visibility strategy was an appropriate one and implemented in good faith, we have no basis to second-guess the judgment of the supervisor mentioned by the first complaining agent. Although there may have been instances where the temporary absence of an agent from his assigned position would not have harmed the effectiveness of the operation, that judgment must be left to the supervisors on the scene. Individual agents who lacked the perspective necessary to make the judgment do not and should not have the discretion to abandon their assigned positions whenever they wish to do so. We believe that there is no evidence that the decision to keep agents in their assigned places was made in bad faith or with the intent to prevent apprehensions. Indeed, the evidence clearly demonstrates that the goal was to maintain a deterrence posture sufficient to keep further incursions by alien traffic to a minimum.

ii. Chula Vista

The three complaints at Chula Vista were essentially the same as those at Imperial Beach. One witness testified that he was regularly told by "all" supervisors to ignore sensor hits. He said supervisors would frequently respond to Dispatch that there were no units available when he believed there were units available. A second witness testified that he had been told to ignore sensors outside his assigned area. One time, while assigned to a position along the border, he said he had responded to a sensor outside his area but had been rebuked by a supervisor. Finally, a Union official testified that he had been told by agents that supervisors had instructed them not to worry about aliens setting off sensors. He said he and the agents interpreted this to mean that supervisors did not want apprehensions. He had never personally been told to ignore a sensor, and he declined to identify who had complained to him. Forty-three other witnesses, however, testified that they were not improperly told to ignore sensors. Although half acknowledged being told or telling others that agents assigned to Xs should not respond to sensors outside of their assigned area, these witnesses did not view this as improper. Many line agents noted that if they left their assigned areas to respond to sensors, more aliens would enter the country in the hole left behind.

Again, the chief difference between the two agents who complained about being told not to leave their assigned positions and the other agents appears to have been their attitude toward the Gatekeeper strategy. We cannot fully assess the Union official's claim without specifics that he was unable to provide, but there is nothing to support his interpretation of the reasons behind supervisors' orders. To the contrary, there is considerable evidence that supervisors explained the reasons for the restrictions on out-of-area responses, that agents understood those reasons, and that the restrictions were entirely appropriate.

iii. Brown Field

The two complaints at Brown Field mirrored the others. One agent claimed that, during 1995 and 1996, supervisors had not allowed available agents to respond to sensors. He limited this somewhat by testifying that agents generally could respond to sensors south of the Otay Mountain truck trail, but not those north of the trail. A second agent said he was aware of one incident where another agent who left his position to respond to a sensor was ordered back by a supervisor.

As at the other stations, these complaints hinged on the issue of whether agents could choose whether to abandon their assigned positions. The other witnesses all appeared to understand that the strategy required that agents generally remain in position rather than seek out traffic elsewhere. Other agents also noted that the decision whether to permit agents to respond to sensors north of the trail were made on a case-by-case basis because of the distance involved in getting to these northern locations. We found no evidence that the line agents who wished to chase after traffic were in a better position to judge whether such action was appropriate than their supervisors were.

The evidence was consistent across the three stations. A mere handful of agents appeared to resent the fact that they did not have the authority to decide where and how they would work. The overwhelming majority of agents, however, seemed to understand how the individual agent positions fit into the overall Gatekeeper plan. Moreover, their testimony reflected an appreciation that the decisions had to be left to the supervisors as opposed to the individual, competing desires of a few agents. The evidence was consistent that agents were free to respond to sensors within their areas of responsibility; however, they were not allowed to respond to sensors outside their assigned areas without supervisory approval. There is no evidence that supervisors made such decisions for improper reasons. As a result, the claims that supervisors were precluding agents from responding to sensors as a means of keeping apprehensions low are unfounded.

b. Sensor list with orders not to work particular sensors

The OIG received a one-page typed list of sensors at Imperial Beach that included a description of their locations. Next to the description for the 7-1 sensor was typed "**never work this sensor." Next to the 9-1 description was typed "**never work**." The person who gave us this document said he had found it in July 1996 in a Border Patrol vehicle and thought it might have been left by the previous driver, a trainee on the previous shift. The witness said he spoke with the trainee, who said he had obtained what he called "the map" in a training class. He declined to identify the trainee but promised to see if he would be willing to speak to the OIG. When asked why he thought the instructions not to work these sensors had been issued, the witness said agents may have been told to ignore these sensors, which are in the northernmost tier, "to maintain what management considered to be vital positions along the border." He said he had given a copy of the list to Union officials prior to speaking to the OIG and they had told him to report the matter to the OIG. When the OIG sought to speak with this witness further,201 in an effort to identify the trainee, he refused to be interviewed. Thereafter, when compelled to speak by INS, he said he had never spoken to the trainee about the list. However, after being shown a list of newer agents at Imperial Beach, the witness did identify the trainee.

When interviewed, the former trainee stated he had never been instructed to ignore any sensors. When shown the list produced by the earlier witness, the former trainee testified that he had never seen it before. After looking at it further, he conceded that he might have seen it previously in the control room at the Imperial Beach Station. He said he had never asked anyone about the list or the instructions not to work the two sensors. He further said he had not kept or made a copy of the list or put it in a vehicle and that the "instructions" not to work the two particular sensors had not come from management, but had been typed on the form as "a joke." He noted that he and the other agents did not ignore any particular sensors.

The only other people who claimed to have seen this list prior to their interview were Union officials, none of whom had seen the list in the ordinary course of their Border Patrol duties. One official testified he was shown the list by a Border Patrol Agent in October or November 1996. He did not know where the agent had obtained the document and did not keep a copy for himself.202  The official believed the document may have been given to trainees but had no first-hand knowledge of its origins or purpose. A second Union official indicated that he had been shown the list by an agent who had given it to the OIG. A third Union official believed he had been previously shown the document but was not certain; he had no information as to who created the document, when it was created or allegedly distributed, or why those two sensors had been singled out. A fourth Union official testified that he had received a copy anonymously in his desk drawer at the Station. He believed it was a page from the Trainee's Manual and had looked at it but thrown it away.

Two other agents said they had heard that copies of a map indicating that the 7-1 and 9-1 sensors should not be worked had been distributed to trainees. Neither witness knew who had made the map or how it was distributed and had never actually seen the document before.

Forty-nine other witnesses from Imperial Beach - at every level of experience and from every shift - testified that they had never seen the list before and did not know anything about it. The PAIC expressed surprise when shown the list and declared he knew nothing about it. He noted that he had seen the packet the trainees are given and the list had not been in it. None of the sensor coordinators we interviewed professed to have seen or heard of the list. Witnesses said they were puzzled by the document because, unless no units were available, sensors were not ignored at Imperial Beach. Each witness who tried to offer a reason why these two sensors might not be worked noted that their northern location made them difficult to respond to in time to apprehend any alien traffic setting them off. Because of the sensors' proximity to residential neighborhoods, many agents said that unless agents were nearby when the sensors were activated, the alien traffic could disappear into the neighborhoods before agents arrived. Other agents expressed surprise that the document said not to work these two sensors because the bike unit had worked these sensors "all the time."

After interviewing 57 witnesses regarding the list, we found no one who knew who had created it, why it was created, when it was created, or how it was distributed, if at all. No one, including the former trainee, could testify that the list was an official document, let alone that anyone had been given it and actually instructed to ignore these two sensors. Furthermore, even if agents had been told to ignore these sensors, there was no evidence that it had been done for inappropriate reasons. As noted earlier in this report, these two sensors were the two northernmost sensors at Imperial Beach and could not be worked effectively when agents were concentrated along the border. When units were assigned to the third tier, these sensors were worked. In the absence of any evidence supporting the claim that large numbers of agents were given this document and improperly instructed to ignore sensors, we found this allegation to be unsubstantiated.203

c. Alleged automatic sensor response

Two agents, from El Cajon and Brown Field respectively, complained that agents had been told to respond "10-4" whenever Dispatch called out sensors hits, regardless of whether they intended to actually investigate. They complained that this was done to keep the reported gotaways down.

i. El Cajon

According to one former supervisor at El Cajon, there had been a disagreement between Dispatch and the station regarding the calling out of sensors. Dispatch had claimed that no one was responding to many of its calls regarding sensor activations. The station had contended that Dispatch was not calling out all of the sensor activations. To resolve the problem, this supervisor designated a person each night to respond to every Dispatch broadcast of sensor activations in El Cajon. This person - often called the "OK man" - was always someone assigned to the fixed position near the Tecate lights. His job was to wait 30 seconds after hearing Dispatch call out a sensor hit. If no one else responded, he was to respond "10-4." The supervisor explained that the point was not to suggest that this person was investigating the sensor hit but merely to acknowledge that the Dispatch broadcast had been heard. If Dispatch later claimed that there was no response to a hit, the supervisor would thus be able to respond that Dispatch had failed to call the hit out.

The supervisor's account was confirmed by the PAIC, who noted that Sector Headquarters and Dispatch had expressed concern about the lack of response to broadcasts of sensor activations. Agents had therefore been told that, as a "courtesy," they needed to at least acknowledge every sensor broadcast and then work the sensor traffic if they could. Other witnesses also mentioned a directive from Sector that every sensor needed to be acknowledged. Two electronic messages provided additional corroboration. The first, dated September 16, 1996, went from the APAIC to the supervisors at El Cajon and reported that, during a visit to El Cajon, the CPA had "voiced a concern about the response to sensors." The message asked the supervisors to ensure that someone always transmitted a response to Dispatch. A month later a second electronic message from the APAIC to El Cajon supervisors reminded them to ensure that all sensor calls by Dispatch were acknowledged. It noted that the station had failed to do so lately and it had been "noticed by Sector."204

None of the agents regularly assigned to El Cajon whom we interviewed complained that this practice was intended to be misleading or was in any way inappropriate. Several noted that having someone acknowledge the transmission prevented Dispatch from using valuable radio time to repeat the sensor hit several times until it obtained some response that could be entered into the ICAD system. The one complaint we received was from an agent who had been detailed to El Cajon for a short period. When initially interviewed by OIA, before the OIG became involved in the investigation, he said he had been assigned to be the OK agent while on detail to El Cajon, and that this required him to call in 10-4 even though he was not working in the vicinity of the alarm or able to investigate the sensor traffic. He believed this was done so the station would not have to report gotaways.

This agent did not repeat his complaint in his subsequent OIG interview. Perhaps he did not do so because his claim made no sense. As has already been noted, raw sensor data was not used to calculate gotaways because it provided no basis for knowing whether the sensor was activated by illegal alien traffic and, if so, how much. Moreover, a "10-4" response did not indicate that any aliens had been apprehended. Once an agent had investigated a sensor alert he was then supposed to radio to Dispatch the results of the investigation, indicating, for example, whether the sensor had been activated by alien traffic and whether any of the traffic had been apprehended or had escaped. It is only this later information that could have affected gotaway estimates, and no one ever suggested that it was falsified.

Perhaps this lone witness, detailed to El Cajon only temporarily, simply was not told the reason for the practice and assumed that there was a sinister explanation for it. The line agents permanently assigned to El Cajon testified that they understood that the purpose of requiring a response was merely to acknowledge having heard the transmission and not to indicate any particular investigatory response.205   The dispatchers had a similar understanding. Although they acknowledged that, on receiving such a response, they entered "en route" into the ticket, they entered "unidentified traffic" and closed out the ticket if they received no further information. In sum, we believe that the practice was not part of any effort to mislead.

ii. Brown Field

One agent at Brown Field testified that on several occasions in 1995 and one occasion in 1996, two particular supervisors had told agents during muster that the sensors were recording too many gotaways and that, whenever a Sector dispatcher called out a sensor hit, the agents should radio "10-4" even if the agent would be unable to investigate the sensor activation.

Although there was no evidence that Brown Field had a designated "OK" man, we found some evidence of a policy that all broadcast sensor hits should receive some response even if merely to acknowledge the transmission. Unlike El Cajon, however, there was no clear rule, and even many of the supervisors did not agree on what was required. The PAIC said that, while he had never instructed agents to just "10-4" a sensor hit with no intent of investigating the call, he had advised agents to respond even if they cannot work the sensor immediately, then catch up with the traffic as it moves north. The APAIC testified that there was no standard operating procedure for responding to sensors, and he was not aware of any procedure allowing agents to just respond "10-4" with no intent of ever responding to the hit. Some supervisors asserted that agents were supposed to respond "10-4" to every sensor dispatch to prevent Dispatch from calling out a sensor repeatedly. They noted, however, that the agent was also to inform Dispatch if he could not investigate the hit. One of the two supervisors mentioned by the complaining agent testified that he had never instructed agents to merely "10-4" a sensor dispatch with no intent of responding. He noted, however, that because Dispatch called out a sensor until an agent responded, he had advised agents to inform Dispatch that they could not investigate the hit at that time but would try to respond as soon as possible. The second supervisor acknowledged a policy to ensure every sensor dispatch was responded to; he explained that this prevented Dispatch from calling a sensor hit out repeatedly.

None of the witnesses supported the complainant's belief that this procedure was used to hide evidence of gotaways. Line agents and supervisors alike testified that the purpose of responding "10-4" was to merely acknowledge the transmission and preclude the need for Dispatch to repeat a sensor call. Furthermore, as noted in the El Cajon discussion, such a procedure would have had no impact on gotaway estimates. Although the supervisors may have complained about too many sensor gotaways in the same discussion as announcing a need to acknowledge every sensor call, there is no evidence that these two issues were linked and that merely acknowledging a sensor call would have affected the gotaways count.

C. Summary of the OIG's findings regarding sensors

None of the complaints the OIG received regarding sensors was substantiated. Although we found that particular sensors had been inhibited at Imperial Beach and Chula Vista, we found no credible evidence that this was done to limit apprehensions or reduce gotaway estimates. The sensors were inhibited when there was insufficient manpower to respond to these sensors and to maintain the strategic positions along the border. When additional manpower became available, these sensors were fully activated. The tactic of inhibiting sensors was consistent with the chosen strategy and that strategy was not inappropriate or executed in bad faith. Indeed, as time has passed, the strategy has proven its capacity to accomplish its intended goal of deterring traffic in the areas where it has been implemented and moving the traffic eastward.

The OIG also found that numerous sensors originally planted in positions distant from the border had been pulled and replanted closer to the border. There was no malfeasance in this, however. Such movements were simply a result of a new strategy being implemented with limited resources. Forced to choose between waiting for new sensors to become available or moving sensors forward to assist the new strategy, management chose the later option, and refilled the northern sensor locations as additional sensors became available. Their choice to enhance the critical border positions first was neither illogical nor ill-conceived, and, over time, has proved to be correct.

None of the claims that agents were inappropriately precluded from responding to sensors was substantiated. The complaint that supervisors were radioing that no units were available when units were purportedly available rested on a belief that agents assigned to Xs along the border were available to respond to sensor hits outside their area of responsibility. This belief disregarded the fact that a basic premise of the strategy was to maintain a constant presence at the border to deter entry. While optimally there always would have always been sufficient manpower to maintain border positions and respond to sensor hits, the fact remains that supervisors frequently had to make difficult choices with the limited resources available. We found no evidence that their choice to maintain border integrity was inappropriate.

Although we obtained a copy of a list purportedly instructing trainees not to work two particular sensors at Imperial Beach, we found no one who was given this list as part of their duties or who was instructed not to work the sensors. The trainee who allegedly left the list in a Border Patrol vehicle denied having done so, denied ever possessing the list, denied being given the list in his training course, and denied ever having been told to ignore any sensors. There was no evidence that this list was an official document distributed to any agent.

Finally, although we found that several stations had a policy that every sensor hit called out by Sector Dispatch had to be acknowledged, we found no evidence that this was intended to hide any information. We did find some confusion as to whether a "10-4" response was intended to merely acknowledge the dispatch or to indicate that an agent was investigating the sensor hit. This confusion needs to be resolved. Regardless, this response had no impact on estimated gotaways because only subsequent apprehension or gotaway information was used to measure effectiveness.

As with many of the allegations against Gatekeeper, those here primarily stemmed from a lack of adequate communication as to why particular actions were taken. Although supervisors may have assumed that their reasoning was obvious, clearly not all agents understood the policy and how particular actions supported that policy. All of the blame cannot be placed on the supervisors, however. If agents do not understand a policy, they should ask to have it explained. Furthermore, just because an agent does not agree with a particular policy does not mean that it demonstrates bad faith on the part of supervisors. Opening the lines of communication, however, may address some of the suspicions and lack of trust that has fueled such allegations.

201 The first interview, conducted in the field while the witness was on duty, was not a formal interview conducted under oath, but primarily a meeting arranged for the purpose of giving the OIG a copy of the list. During it, the witness was asked several questions about the list.

202 The official did not identify the agent because he had failed to ask the individual for permission to disclose his identity pursuant to the Union/OIG agreement.

203 We also located several diagrams of sensor locations at Brown Field with the notations not to call out or not to work particular sensors. No one brought these to our attention or complained about them. When questioned, the sensor coordinator explained that certain of the sensors were National Guard sensors and the notations were from a time when the sensors had been planted pending official approval. During that relatively short period, these sensors were not called out. The sensor coordinator explained the reference in a second diagram to particular sensors not to be worked by noting that agents needed to be told not to work these sensors too closely because they had a tendency to tip off the alien traffic to the sensors' location. The PAIC offered the same explanation for this document. As no witnesses complained about these directives nor contradicted the explanations offered, we have no basis to find there was anything inappropriate suggested by these documents.

204 An August 1996 memorandum by the APAIC at Campo to all agents and supervisors at Campo and Boulevard alluded to the same concern about the lack of responses when Dispatch called out sensor hits. It instructed agents to give "some sort of response on each and every report of sensor traffic," and noted that if an agent was unable to investigate the report he was to respond "NUA" (no unit available).

205 Many agents at other stations where there was a similar requirement that all sensor hits called out by Dispatch be responded to in some way agreed that responding "10-4" did not indicate the agent was actually investigating the sensor activation. This view was not universal, however, as some agents believed a "10-4" meant the agent was responding to the scene. This inconsistency needs to be resolved by clarifying what agents should respond when they are acknowledging the transmission but not actively pursuing the sensor traffic. Although "NUA" (no units available) is appropriate when no one will respond, it may not be known immediately whether a unit is available. Thus, "10-4" should merely acknowledge the transmission to avoid repetitive transmissions. If an agent is actually responding, he could indicate he is "en route" by giving that code instead. Whatever resolution is arrived at, it should be clearly communicated to all agents so a consistent practice can be established.