B. Allegations that INS's public claims of success overstated Gatekeeper's actual results
In the initial allegation received by the OIG in February 1996, the complainant alleged that the San Diego Sector had overstated the extent of Gatekeeper's success, giving the impression that the results achieved at Imperial Beach had been duplicated across the Sector. Bonner similarly claimed that the INS and the Border Patrol were misleading the public about Gatekeeper's success. Because no complaint focused on any particular public statement, the OIG reviewed several hundred newspaper articles and television news segments about Gatekeeper. We interviewed ten people who have served in DOJ Public Affairs, INS Public Affairs, or the Sector Public Information Office (PIO) and also questioned DOJ and INS management regarding the public dissemination of Gatekeeper information.
1. Overview of the effort to publicize Gatekeeper
Media coverage was an integral part of the launch of Gatekeeper and was intended to serve two purposes. First, it was intended to tell Americans generally and Californians specifically that the Clinton Administration was taking action to address the problem of illegal immigration along the San Diego border. Second, it was to convince potential illegal immigrants that they could no longer expect unimpeded travel across the border. Information about the size and seriousness of the program was relayed in Spanish to the Mexican media in an effort to deter individuals considering whether to leave home for the northward trek.
To handle the numerous requests for information and tours, a media command center was temporarily erected at the Imperial Beach Station. This was a proactive effort, with daily updates of apprehension statistics and commentary on how the program was faring. Although the quantity of coverage declined after the first few weeks of operation, there were additional flurries of activity surrounding the announcement of additional resources or major new initiatives, or assessing progress at the six-month and yearly anniversaries of the operation.
Testimony from a broad range of witnesses indicated that this process was chaotic. Neither the Border Patrol nor INS was experienced with or well-suited to managing the public relations aspects of such a significant operation. Personnel from DOJ Public Affairs were therefore brought in to direct the media portion of the operation. One particular OPA staffer became the primary contact person for media inquiries. Although she was not experienced in evaluating Border Patrol operations, she essentially took it upon herself to become the primary public advocate for Gatekeeper. As a result, she became known within the Border Patrol and INS for her interpretation of statistics and pronouncements on Gatekeeper's progress.
Within DOJ and INS management, however, there were concerns about the efforts to publicize Gatekeeper. The Attorney General recalled having cautioned Commissioner Meissner and other INS officials about claiming success and making promises they could not keep. Her recollection in this regard is corroborated by that of former Deputy Attorney General (DAG) Gorelick and notes of meetings the Attorney General had with INS management. Commissioner Meissner also said she recalled fearing that announcements about Gatekeeper had "too much enthusiasm." She was concerned not about the accuracy of data that had been released to the media but about the conclusions drawn from it.
Others at INS headquarters were apparently also concerned about the potential for overstating the significance of early data. Reporting on his canvassing of personnel in the Western Region, the San Diego Sector, and INS headquarters, an Assistant Chief Patrol Agent in Border Patrol Headquarters noted, in a November 4, 1994, memorandum to William Slattery, the Executive Associate Commissioner for Field Operations, that "all parties contacted" had agreed that it was best not to release any preliminary statements regarding the success of Gatekeeper. The memorandum noted that the statistical data available at that stage of the operation were "ambiguous and should not be utilized to justify a preliminary finding of success."
Despite these concerns, there was little oversight of the public relations operation. Although some official media releases were reviewed by the DAG's or Commissioner Meissner's office, much of the information, particularly early in Gatekeeper, apparently was not reviewed by anyone outside the Sector. Moreover, WRD Gus de la Viņa testified that, when he was CPA, he never reviewed the releases, leaving that task to agents in Sector PIO. He said the PIO agents would ensure that there were no numbers in the press releases "because numbers were open to interpretation." The press releases issued at the beginning of Gatekeeper, however, generally were laden with statistics because that was what the media was requesting and what the DOJ Public Affairs specialist was offering.
The resulting public relations campaign was a poorly coordinated, insufficiently supervised, deluge of information from all levels - DOJ headquarters, INS headquarters, and the Sector. Some of the information publicly reported by these various sources did not agree. Some of the public claims were more wishful thinking than hard fact. Most significantly, there was apparently little accountability for the flow of information.
2. Review of Gatekeeper publicity
The OIG's review of the press releases and news reports from Gatekeeper did not find examples of inaccurate apprehension statistics. The underlying facts of Gatekeeper's performance were consistent and supportable. This is probably because this type of data had been gathered for years and was purely objective. The OIG's review did uncover, however, several areas where information was overstated or, as Commissioner Meissner described it, "ahead of the operations on the ground."
The first such example was the immediate claim that Gatekeeper was working. The first day's press release claimed that apprehensions at Imperial Beach were three times what they had been on the same date the previous year. Although this information was accurate, the release contended that this fact was an early indicator that the strategy of increasing apprehensions at Imperial Beach to push traffic east "is beginning to work" and that Gatekeeper was responsible for "this significant improvement in controlling the border." This, the meager data did not support. Nor did it support similar claims of success in subsequent updates. These updates did not indicate who prepared them but the evidence suggests it was the DOJ Public Affairs personnel in San Diego.
Such unsupported claims were picked up by the media. A November 1, 1994, article in the Los Angeles Times quoted a Border Patrol release suggesting that the operation was "deterring border crossers and increasing the agents' effectiveness." A November 5, 1994, article in The Press-Enterprise quoted a Sector PIO agent as claiming that the Border Patrol believed that agents "in the Gatekeeper area" were apprehending "nearly every person" who tried to cross illegally.
These are just a few samples of the early claims of success for Gatekeeper. Although none contained demonstrably false information, each drew conclusions that could not yet be substantiated. Such claims, however, should be attributed to the bad judgment of certain individuals, not a Departmental effort to mislead. Indeed, in their public pronouncements, officials from Washington were appropriately cautious. In a January 1995 press conference the Attorney General would say only that the Border Patrol had made "great strides" with Gatekeeper and that the program was "having an impact." The official DOJ release concerning the Attorney General's announcement of new resources for San Diego quoted her as saying that Gatekeeper "had begun to make a difference." Later that same month, in a televised meeting with editors of the Los Angeles Times, Commissioner Meissner said it was too early to judge the impact of Gatekeeper in California because spring is the time of the heaviest alien traffic. She declared the program was a multi-year effort that was not going to "razzle-dazzle overnight." Unlike several spokespeople in the San Diego Sector, who were sometimes quite vague as to the scope of the operation, the Attorney General and the Commissioner also made clear in their statements that only traffic in Imperial Beach was being targeted at first. Official progress reports and briefing materials also routinely defined the relevant area.220
Our review of media materials also found cases in which accurate data was reported, but without necessary qualifications. When data from two different years was compared to show, for example, that a station's apprehensions had increased or decreased, the reader was not told that station boundaries had shifted during that time and that the later figures might therefore overstate or understate the actual fluctuation.221 Similarly, statistics comparing apprehensions at the checkpoint stations over time failed to note whether the checkpoints had been closed for any of the reporting periods involved. In addition, reports of additional manpower assigned to the Sector failed to indicate how many of these agents were filling positions created by attrition versus adding additional agent positions.222
These problems were caused by several factors. First, the people providing the information were not told how it was going to be used; they therefore were not in a position to warn the user, for example, that one data set could not properly be compared to another without qualification. Second, those providing the data were not always aware themselves of the need to qualify. For example, data concerning the number of newly hired agents could come from someone lacking information as to the number of agents who had left the San Diego Sector. Third, those analyzing the various data were not sufficiently aware of the nuances of operations or statistical methods to ask the right questions.
It must be noted, however, that several factors mitigated the damage caused by these problems in this case. Although factors such as the shifting station boundaries certainly affected the data comparisons at the margin, the movement of traffic eastward over time was of such a magnitude that reports of such movement were essentially correct. Similarly, the changes in station boundaries were generally not of such a magnitude that they had a measurable impact on any given comparison. Moreover, inferences that caused overestimates in manpower levels did not affect the primary issue of whether Gatekeeper was succeeding at apprehending, deterring, or shifting alien traffic. While available manpower may have an impact on the success of Gatekeeper, it was not in itself a measure of success. We also found no evidence that these problems were the result of intentional efforts to deceive the public.
There was, however, one factual claim repeatedly made that was demonstrably false. On numerous occasions the Border Patrol contended that it was fingerprinting every person apprehended in San Diego. This statement was inaccurate at the beginning of Gatekeeper and remains untrue today. For the first 11 months of Gatekeeper, IDENT machines were installed only in the three western stations, San Clemente, and the San Ysidro POE; no one apprehended in the eastern portion of the Sector was being entered into IDENT.223 Even at the stations where IDENT was installed, significant numbers of apprehended persons were not being fingerprinted. Although there was a clear directive that stations with IDENT machines were to enter everyone who was apprehended, that directive was not followed.
It is unclear whether the sources of these claims were aware that they were false. At a minimum, they should have recognized that because the machines were only in a few selected locations, aliens apprehended at the other stations were not being fingerprinted. Whether the sources had any knowledge that not all persons apprehended at the three most western stations were being fingerprinted is uncertain. Because the public statements were not attributed to anyone in particular, we could not determine what their sources knew. As late as September 1995, Robert Bach, INS Executive Associate Commissioner for Policy and Planning, indicated in a memorandum to Commissioner Meissner that all adults at the three westernmost stations were being fingerprinted. Thus, even internal documents contained this error, even after some within the organization were aware of its inaccuracy.
The erroneous claim that all persons were being fingerprinted highlights the difference between what the stations were directed to do and what they actually did. This was not the only example we found where the Commissioner incorrectly assumed that one of her directives was being carried out.
3. Conclusions regarding Gatekeeper's public claims of success
Despite the voluminous amount of information released regarding Gatekeeper, there was no evidence of any concerted effort to mislead the public regarding the operation's success. The public reports of apprehension data were consistent with the internal data being reported by the stations and Sector Intel. Although some press releases were far too quick to cite this data as proof that Gatekeeper was succeeding, such claims, made after a few days of operation, were so patently premature that the public was unlikely to have been swayed. Moreover, the majority of the claims, particularly from higher level officials, tended to exhibit more restraint.
The impact of the efforts to sell the program to the public was also tempered by the presence of a critical and well-informed media that did not rely solely on the information contained in the releases. Thus, there was an article in February 1995 in the San Diego Union-Tribune arguing that Gatekeeper was "bordering on failure." Members of the media frequently went to the border to evaluate claims of progress first-hand. Indeed, one reporter apparently tested the operation by sneaking across the border from Mexico. Additional articles, two months later, pointed to rising apprehension figures and questioned whether Gatekeeper was making any difference. As alien traffic levels exploded in East County, the media wrote articles critical of Gatekeeper's progress. At the same time, critics in Congress issued their own releases decrying Gatekeeper's failure to meet the standards set by the INS's Hold the Line operation in El Paso. This environment, characterized by substantial skepticism about Gatekeeper, was not conducive to the establishment of a program to deceive.
While isolated pieces of information were not placed in sufficient context or, in the case of the IDENT claims, were demonstrably false, this data was generally peripheral to the central issue of whether Gatekeeper was having a material impact on the flow of illegal alien traffic. We find that these lapses, although problems that must be addressed, were not part of any organized plan to deceive.
These problems must be addressed in the future. Conclusions regarding the import of statistical information should be made by persons with sufficient training and knowledge to make them. Better coordination between those who generate statistical information and those who analyze it is needed to ensure that issues affecting the significance of the data are adequately communicated along with the data. And sufficient oversight and accountability need to be incorporated into the dissemination process.
Before making sweeping statements of compliance with any directive, INS's leadership should demand the data necessary to determine the extent to which a program is in place. INS and the Border Patrol should also recognize that unequivocal, sweeping claims that something was or is done on every occasion will almost certainly be inaccurate and thus should be avoided, absent clear supporting data.
220 As noted, infra, at 313, in relation to the Congressional delegation's visit to the Sector, although one witness recalled Commissioner Meissner's claiming progress across the first 14 miles of the border, the written handouts explicitly distinguished between the progress made at Imperial Beach versus the next two stations.
221 This problem was not limited to public reporting of information. Internal documents also repeatedly failed to note that factors such as changes in station boundaries, agent deployment, or operating methods could affect the reliability of the statistical comparisons. Indeed, it was not until after we raised this issue that the SW Border Weekly began placing an asterisk next to figures affected by shifting station boundaries, alerting readers that that comparisons could be distorted.
222 The Attorney General testified that she found this particular problem extremely vexing. She testified that because of this recurring issue she did not want manpower information publicly reported.
223 Interestingly, at the same time the Border Patrol was claiming it was fingerprinting everyone, newspapers correctly reported that the machines were not yet installed across the Sector. These two facts were never linked together to alert the reader that the claim at best applied to only a few stations.