2. Allegations that sensors were inappropriately moved closer to the border
Bonner first claimed that sensors were inappropriately moved closer to the border in his July 15, 1996, testimony before the State Assembly subcommittee. He gave no details, merely stating that "[m]any electronic sensors, which provide an indication of movement of illegal entrants north of the border, have been deactivated or removed in the western portion of the San Diego Sector." In his August 1996 OIG testimony, he said that the only "northern" sensors he knew that were moved were those at Imperial Beach. He claimed that agents had told him which sensors had been moved, but he had not recorded this information and had since forgotten it. He had no first-hand knowledge regarding this allegation.
There was universal agreement among witnesses that sensors had been moved at the various stations. It also was agreed that sensors may be moved legitimately to enhance detection of alien traffic as it shifts to different trails to avoid detection by the Border Patrol. Agents observing new alien traffic routes frequently suggest sensor repositionings. The issue thus is whether productive sensors were removed and placed in less productive locations to hide the existence of illegal alien traffic and prevent agents from apprehending these aliens. Although we questioned agents at every station on this point, we received complaints about sensor movement only at Imperial Beach, Chula Vista, and Brown Field.
a. Imperial Beach
The OIG interviewed 93 witnesses regarding the movement of sensors at the Imperial Beach Station. Over 40 testified that "northern" sensors (sensors farthest from the border) had been moved closer to the border. Although most said the movement was for legitimate reasons, seven line agents at Imperial Beach alleged impropriety. Witnesses varied on how many sensors were moved and when. Some said "all" of the northern sensors; some specified the 60 series, the 40 series, or the 70 series, while others did not believe that northern sensors were moved. Witnesses also varied on timing, with some saying the sensors were moved pre-Gatekeeper and others saying they were moved when Gatekeeper began or later.
From Sector Electronics, which is responsible for planting and maintaining sensors, we obtained records charting when and where each Imperial Beach sensor had been planted or moved from the late 1980s forward. We also obtained copies of sensor log entries - which chronologically report every action taken regarding any sensor, including planting, moving, pulling, or checking its condition - for December 29, 1994, through October 1996.191 Although the individual sensor records were not perfect, they were substantially complete, generally missing just the pull or replant dates of a few sensors. The log, however, contained all of this data from December 29, 1994, to the time the records were produced. In most cases, the missing data from before December 1994 could be estimated from other information. From these records, we could determine which sensors had been planted prior to Gatekeeper and which of these had been moved from their original locations.
As shown on the chart below, 34 sensors were planted at Imperial Beach prior to January 1994: 16 were in what was to become the third tier under Gatekeeper, the furthest tier from the border; 14 were in the area that would become the first tier under Gatekeeper, and 4 were in second tier. In the first eight months of 1994 at least 22 of these 34 sensors - including 15 from the third tier - were pulled. Most were replanted elsewhere, although some were replaced with new ones in the same position.192 Nineteen of the sensors planted in that same time period were planted in the first tier, and three were planted in the second tier. None was planted in the third tier. In the last four months of 1994, six sensors were pulled: four from the first tier; two from the second tier, and none from the third tier. Eight sensors were replanted in that same time period, seven in the third tier within the first three weeks of Gatekeeper. Four more third-tier sensors pulled in early 1994 were replanted in the first two months of 1995, and one was replanted in August 1995. Thus, 12 of the 16 third-tier sensors pulled prior to Gatekeeper were replanted in the third tier within the first year of the Operation.
The PAIC at Imperial Beach at the time the third-tier sensors were pulled explained that sensors could be used both to report the presence of alien traffic and to indicate the direction the traffic was taking. Lacking enough sensors to fill both functions prior to Gatekeeper, he filled the reporting function positions first. Reasoning that it makes more sense to have the doorbell at your front door to tell you people have arrived rather than merely a later signal to indicate that someone is already in your house, he moved the sensors closer to the border. This movement was in conjunction with his testing of new strategies of deploying agents closer to the border - strategies that presaged the Gatekeeper plan. Although none of these experiments lasted for long due to manpower constraints, the sensors were kept closer to the border until Gatekeeper began.
The PAIC testified that when he ordered the sensors closer to the border in early 1994, he did not specify precisely where they should be located but merely directed the sensor coordinator not to put any north of where agents were deployed. The sensor coordinator - a line agent - was responsible for selecting the exact sites. He testified that the PAIC had told him that he wanted to know as soon as the aliens had entered the country and had asked him to put sensors at every major crossing point a certain distance from the border. The PAIC selected which northern sensors were to be moved south based on sensor reports, and the coordinator determined where the sensors were to be placed based on agent input. Sector Electronics personnel then planted them in the selected sites.193 Both the PAIC and the sensor coordinator testified that, as traffic patterns changed, particularly in response to changes in agent deployment, sensors were moved to deal effectively with the new patterns.
The sensor coordinator noted that, because northern sensors had to be pulled to fill the additional southern locations, he wrote a memorandum requesting more sensors for Imperial Beach. We obtained a copy of this memorandum, written on August 15, 1994. It requested 60 additional sensors to permit coverage of the third tier and to replace old sensors. It included a map that proposed sites for 92 sensors, including approximately 25 in the third tier, based on observation of alien traffic patterns during all three shifts. The sensor coordinator noted that not all agents agreed on every location and that it was impossible to make every agent happy.
A memorandum from a subsequent PAIC to his FOSs, written in January 1996, indicates that the PAIC had requested 40 additional sensors for the Station, to be planted as close to the border as possible "to signal immediate entries." A volunteer line agent was to be selected to solicit input from all of the agents at the Station to pick the "ideal" locations for these sensors. The sensors already in place would remain as "directional" sensors, with the goal being to work the traffic closer to the border. This memorandum was written shortly after a meeting at which the CPA asked all PAICs to audit all sensors for effectiveness and to move ineffective ones to better locations.
From this evidence, we conclude that the witnesses who claimed most of the northern sensors were pulled were correct. But the claim that the movement occurred during Gatekeeper was incorrect. Indeed, over half of the sensors pulled from the third tier were replaced shortly after Gatekeeper began. In addition, the sensors most identified as critical third-tier sensors, the 60 series, were among the ones replanted. Thus, the witness who claimed that the 64-1 sensor was pulled during Gatekeeper and never replaced was wrong. The 7-1 and the 9-1, two other sensors frequently mentioned by witnesses, also were replanted in the third tier. The only third-tier sensors pulled in early 1994 and never replaced were the 63-1, the 45-1, the 14-1, and the 8-1; none of these was identified by witnesses as being significant. The other third-tier sensors were replanted at some point, although some were later re-pulled in 1995 and 1996. Around that same time two new third-tier sensors also were planted, including an additional 60 series sensor.
The timing of this movement is significant. The witnesses who complained about it generally suggested that it was done to prevent agents from catching aliens or to decrease the number of gotaways reported - all to make Gatekeeper look better. The sensors, however, were moved at least six months prior to Gatekeeper's implementation. And a mere three weeks after Gatekeeper began, seven of the third-tier sensors, including the key 60 series sensors, were back in place. Moreover, the sensors were not being used for data on gotaways.
Two witnesses suggested that the sensors were moved because "in the supervisors' minds" if there are no sensors in the area, there are no aliens there. These witnesses did not identify the supervisors who purportedly believed this and offered nothing in support of their claim. We do not find it credible. The testimony of the PAIC and the sensor coordinator explaining the reason for the shift is reasonable. It is supported both by the documents and the fact that the third-tier sensors were replanted once Gatekeeper began. As discussed earlier in this report, we believe that agent claims of this sort are simply symptomatic of certain agents' inability either to grasp or to accept the essence of the Gatekeeper strategy, which sought to concentrate scarce resources on the twin goals of deterrence and apprehension closer to the border.
Movement of sensors continued during the period in late 1996 for which we obtained the sensor records. Although the number of sensors in the third tier today is not as high as prior to Gatekeeper, the levels of traffic penetrating Imperial Beach at all, let alone the third tier, has dropped so dramatically that the need for sensors in the third tier is significantly lower than prior to Gatekeeper. Thus, we found no evidence that the movement of sensors at Imperial Beach was for improper reasons.
b. Chula Vista
The OIG interviewed 55 witnesses concerning the movement of sensors at the Chula Vista Station. Forty-five witnesses testified that "northern" sensors were moved to locations closer to the border. Only four expressed concern that this movement was not for legitimate reasons.
Witnesses tended to generalize and merely indicate that the "northern" sensors had been moved. Ten who identified specific sensors mentioned that the "valley" sensors, located several miles from the border north of Highway 905, had been pulled. One witness claimed the "docks" sensors had been moved closer to the border. Another said sensors in two particular canyons had been moved south.
We had a harder time reconstructing sensor movements for Chula Vista from Sensor Electronics records than we had for Imperial Beach. At Imperial Beach, when a sensor moved to a substantially different location it was given a new number to avoid confusion. Chula Vista used the same sensor number, forcing us to decipher the map for each sensor to determine whether it had moved to a location in the same area or to a completely new area. Chula Vista also renumbered a group of sensors, even though they had not been moved. And Chula Vista had not created the detailed sensor analyses like those at Imperial Beach, which identified the tier for each sensor and its relative productivity. Finally, Chula Vista's boundaries changed over time, making it difficult to determine which station was responsible for the movement of a particular sensor.
Because of these difficulties, we focused only on sensors said to have been moved south. The first such group was the five valley sensors, the 402-1, 403-1, 404-1, 405-1, and the 406-1. Two of these were pulled on July 7, 1995. One was moved on October 13, 1995, and then pulled on December 28, 1995, the day the other two sensors were pulled as well.194 Four new sensors were placed in the valley (though in different locations) beginning in September 1996.
The second group was the docks sensors. Because these were not identified to us by number, we selected a group based on descriptions in the sensor maps and in ICAD. We tracked five sensors in this area, the 203-1/209-1, 208-1/270-1, 213-1, 215-1/265-1, and 251-1/211-1.195 The 203/209 sensor, initially planted and pulled before Gatekeeper, was planted in a new location in October 1995. The 251/211, the 213, and the 215 were planted prior to Gatekeeper and had not been moved by October 1996. The 208\270 was planted in November 1995, and there is no evidence that it had been moved by October 1996.
The third group was the canyon sensors. Here, too, we selected a representative group that included every sensor whose description referred to those canyons. These sensors were the 305-1, 307-1, 315-1, 316-1, 322-1, 325-1, 326-1, 329-1, 338-1, 340-1, 343-1, and the 350-1. Four of these were pulled from their canyon locations on May 4, 1995, although one was replanted on March 28, 1996. Two were pulled on May 12, 1995; another was pulled on July 11, 1995; and the remaining five sensors were not pulled.196
These records indicate that in fact some "northern" sensors were moved during Gatekeeper. The canyon sensors were the first significant post-Gatekeeper moves of sensors at the Chula Vista Station. The repositioning appears to have been made at the suggestion, a few weeks before, of a line agent who wrote a memorandum in April 1995 suggesting that the 300 series sensors be moved closer to the border because no one was available to respond to them. This agent testified that he had reviewed the ICAD printouts for these sensors and, upon noting that no one was responding to them, decided to write the memorandum.
The PAIC at the time confirmed he had ordered the northern sensors moved because the deployment of agents closer to the border and the expansion of Chula Vista's area of responsibility to include the Otay Port of Entry left insufficient manpower to work those sensors. The sensor coordinator who assumed responsibility for the Chula Vista sensors in December 1995 gave essentially the same explanation. Indeed, he wrote a memorandum in January 1996 suggesting that the location of secondary sensors - like those in the canyons and the valley - be adjusted to improve their efficiency. Additional corroboration comes from the fact that, prior to being moved the valley sensors were regularly inhibited because of the lack of manpower. Rather than being continually inhibited, it made more sense to relocate them to where agents could respond to them.
Only one agent mentioned the canyon sensors when complaining about sensors being moved, and said supervisors "implied" these sensors were moved because they did not want to know that traffic was going north of the front lines. This agent stated his belief that a lot of alien traffic made it north without being counted as gotaways because of these moves. Another agent complained that the northern sensors were moved to lower gotaway figures. We did not find any factual basis for these suspicions. Not only does the evidence show an appropriate motivation, but, as has been noted, since sensors were not used to measure gotaways the movement did not affect these estimates.197
One agent complained about the valley sensors being moved because he said there was a unit who worked this area "exclusively" and apprehended a lot of aliens. No one else spoke of such a unit, and numerous witnesses, including line agents, said these sensors were moved because there was no one to work them. Although there possibly may have been a unit assigned in 1994, there is no evidence that in mid-to late-1995 these sensors were being worked and were producing large numbers of apprehensions or that the subsequent movement was intended to hide the presence of alien traffic in the area.198
A Union official complained that all of the sensors north of a few hundred yards of the border had been moved toward the border and that some were placed in unproductive areas. When pressed to identify which sensors were in unproductive areas, he could not specify a single one. He also said that, at the time of his interview, there were no sensors more than a few hundred yards from the border. The records indicate, however, there were at that time three sensors in the valley area, at least four sensors in the docks area, and at least five in the canyons.
Based on the testimony of the witnesses responsible for the movement of the sensors, the testimony of the overwhelming majority of the Chula Vista witnesses (who found no impropriety in the movement of sensors) and the supporting documents, we find there is no credible evidence that the Chula Vista sensors were moved for inappropriate reasons.
c. Brown Field
The OIG interviewed 30 witnesses at the Brown Field Station regarding the movement of sensors. As at Imperial Beach and Chula Vista, there was no question that many sensors had been moved from northern locations to places closer to the border. Most of the witnesses testified that this was in conjunction with the redeployment of agents to positions closer to the border. Only two complained that the movement was inappropriate; both singled out the "lakes" sensors in this regard.
The systematic movement of northern sensors at Brown Field did not take place until February 1996 when sensors north of the Otay Mountain Truck Trail (the "lakes" sensors) were moved closer to the border. The sensors also were renumbered to reflect the patterns of alien traffic so agents could easily determine which direction the traffic was moving. According to the PAIC, this sensor movement paralleled a manpower shift to the south as part of Gatekeeper. When additional sensors became available, they were replanted north of the mountain. He said the sensors were never moved to unproductive areas and noted that had they been the supervisors would have "bitched and moaned" about good sensors being taken away. He also noted that the decisions regarding sensor placement were dictated by the station's needs, not Sector Headquarters.
The PAIC's account was corroborated by two sensor coordinators at Brown Field, both line agents.199 They testified that sensors were moved south of the Otay Mountain Truck Trail (which is approximately one mile from the border) at the direction of the PAIC to assist agents in determining where alien traffic was entering the country and what direction it was taking. The two picked locations based on other agents' suggestions and on analysis of sensor and apprehension records. Then they informed the PAIC of their decisions. As the station obtained more sensors, a number of the northern sensors were replaced. Neither agent had any knowledge of sensors being placed in unproductive areas.
The accounts of the PAIC and the sensor coordinators are fully supported by a series of memoranda. In February 1996, the sensor team wrote that, having conducted an analysis, they had identified a need for 140 additional sensors to cover the Station's area of responsibility, and they proposed specific locations. In March 1996, the PAIC forwarded a version of this memorandum to the CPA and requested the additional sensors. His cover memorandum claimed that the number of aliens crossing Otay Mountain had increased dramatically since Gatekeeper and that this traffic was taking new routes - routes that agents could not easily cover because of the plant growth and the condition of the truck trail. Only with the addition of new sensors could the mountain traffic be effectively worked.
In May 1996 another memorandum from a Brown Field supervisor explained that, in the wake of a change in agent deployment in April 1996 designed to push traffic east, the Brown Field sensors also had been reorganized. On the basis of a sensor effectiveness analysis, more sensors had been placed near the border to detect crossings, additional sensors had been put to the north to monitor the traffic, and all sensors had been renumbered to assist in tracking groups. The new deployment had led to significantly increased sensor activity on the eastern portion of the station, and the memorandum indicated additional locations where sensors were still needed. This memorandum supports the claims that sensors were needed closer to the border and highlights efforts to monitor sensor activity and ensure that sensors were placed in productive areas.
One agent alleged that line agents were never told why the lakes sensors were removed, and he related that, as a result, "we began making our own rumors" as to why supervisors were moving "good" sensors. He then claimed that supervisors had the lakes sensors removed because they did not want high numbers of aliens being reported. He noted that when he asked a member of the sensor team why the sensors were being moved, the reply was "I don't have to tell you why. You should know why." This agent interpreted this response to mean sensors were moved for inappropriate reasons, although he had no evidence to support this interpretation other than his suspicions. Indeed, when first interviewed, he had testified that he believed the sensors were moved as "part of the Gatekeeper strategy" and that that he was not aware of any sensors being moved from high traffic areas to low traffic areas.200 Under these circumstances, and given the overwhelming evidence that the sensor moves were appropriately motivated, we find this agent's claims of impropriety not credible.
Another agent claimed that he had observed sensors moved from areas of high traffic in the lakes region to areas of low traffic near the border. He did not specify which sensors were in low traffic areas, and his basis for believing that particular areas were low traffic areas is unclear as well. In considering his claim, we note that he underestimated the number of sensors at Brown Field at the time of his interview by over 60 percent - believing there were approximately 60 in place when there were approximately 160. This suggests that he was not well-informed regarding the sensors. Although he conceded that agents were free to make suggestions about sensor locations, he made no claim to have reported his belief that sensors were not well-placed or to have sought the movement of these particular sensors to better locations. To the extent that he may have been suggesting that any locations near the border were less productive than the lakes area, he simply did not understand the new strategy.
This same agent noted that a scope operator was sent up to the lakes area in May or June 1996 for intelligence purposes. Thereafter, sensors were replaced there. This incident suggests that the supervisors were indeed interested in determining the best locations to place the new sensors that were being received and wanted to learn if traffic was still traversing that area. The subsequent placement of sensors there demonstrates the intention to respond to the traffic when resources became available. In sum, there is overwhelming evidence that the lakes sensors were moved in support of the new strategy and were placed in appropriate locations for that strategy, and that the PAIC and the sensor coordinators intended to maximize the use of available sensors rather than waste them in unproductive locations.
d. Summary of the OIG's findings regarding sensor movement
Although many sensors in locations distant from the border were moved closer to the border, we found no credible evidence that this movement was intended to hide the presence of alien traffic or to lower the effectiveness of sensor use. The movement forward was concurrent with and consistent with the new strategy of working alien traffic closer to the border. Credible testimonial and documentary evidence demonstrates that sensor location decisions were made at each station and driven by the desire to work alien traffic more effectively. The only evidence of Sector involvement was a directive to the stations to measure their sensor effectiveness and to locate sensors to maximize their effectiveness. Moreover, the decisions as to specific sensor sites were made by line agents serving as sensor coordinators, often based on the input from their fellow agents. They uniformly testified that they had the freedom to use their best judgment when locating sensors, as long as they were consistent with the station's strategy.
There was ample evidence that decisions to relocate sensors from northern locations arose out of the change in strategy and the lack of sufficient sensors to cover all locations adequately. Priority was given to areas closest to the border work; when additional sensors and agents became available, sensors were replaced in northern locations as appropriate. Thus, we find the allegations that sensors were moved from northern locations to hide alien traffic or to underestimate alien traffic are not substantiated.
191 All sensor log records prior to December 29, 1994, had been discarded prior to our request. The maps, which are considered the primary records, are maintained indefinitely.
192 If a sensor was replanted in its identical location, it was usually pulled originally because it was malfunctioning or being replaced with a different unit.
193 One Union official claimed that the sensor coordinator had told him that the PAIC would not let him place the sensors where he wanted and had forced him to put them in unproductive locations. The sensor coordinator denied this and said he was always allowed to put the sensors in the most effective areas that he selected. Numerous other witnesses said they had no reason to believe that sensors at Imperial Beach had been intentionally located in unproductive areas.
194 These sensors were then planted closer to the border. At some point, sensors near the border were given the numbers previously used to identify the valley sensors.
195 Because four of these sensors were renumbered during Gatekeeper, we use both numbers.
196 Of all the 300 series sensors, regardless of location, 11 sensors planted prior to Gatekeeper were never moved; 10 were pulled prior to Gatekeeper, and 27 were pulled or moved after Gatekeeper began.
197 The PAIC at the time testified that he was aware that alien traffic was getting away but he was not concerned about it because he was used to thousands of gotaways prior to Gatekeeper.
198 It should be noted that these valley sensors were not planted until July 1994. Thus, unlike the 60 series sensors at Imperial Beach, they did not have a lengthy record of productivity prior to Gatekeeper that might suggest that their movement post-Gatekeeper was an effort to adversely affect a well-established working area.
199 A third line agent on the Brown Field sensor team refused to be interviewed by us. He was interviewed by OIA prior to the OIG's involvement in the investigation, and we obtained a copy of the memorandum recording the substance of this interview. Because we had no basis for believing that his testimony would support any claims of wrongdoing and we had reason to believe his refusal was not related to this investigation, we did not seek to compel him to testify.
200 This agent was interviewed under oath twice. During the first interview, he claimed that virtually every type of wrongdoing alleged by the Union was occurring at Brown Field. He claimed to have first-hand knowledge of the wrongdoing but refused to identify the alleged perpetrators or provide much detail of the incidents, citing a fear of retaliation. When re-interviewed only two months later as a confidential informant pursuant to the Union/OIG agreement, he claimed to have "forgotten" the identity of the persons who had allegedly engaged in wrongdoing and retreated from a number of his initial claims. He did identify, however, the individual who allegedly had made the comment about the sensor movement. That person was interviewed by the OIG and was adamant that the sensors were moved only for appropriate reasons.