This investigation by the Department of Justice Office of the Inspector General (OIG) examined the Immigration and Naturalization Service's (INS) "Citizenship USA" initiative of fiscal year 1996.
Beginning in 1993, the demand for naturalization began to increase at a staggering rate and application backlogs developed at INS offices throughout the country. By June 1995, INS was receiving applications for naturalization at a rate twice as high as it had the previous year. INS projected that without a serious effort to reduce this application backlog, by the summer of 1996 an eligible applicant would have to wait three years from the date of application to be naturalized as a U.S. citizen.
On August 31, 1995, INS Commissioner Doris M. Meissner announced "Citizenship USA" (CUSA), an initiative to reduce the backlog of pending naturalization applications to the point where an eligible applicant would be naturalized within six months of application. The goal of the initiative was to reach this level of processing "currency" within one year. The effort focused on the workload in the five districts in the country-dubbed "Key Cities" for CUSA-which then had the largest application backlogs: Los Angeles, New York, San Francisco, Miami, and Chicago. To reach the CUSA goal, INS dramatically increased its naturalization workforce in the Key Cities, opened new offices dedicated to naturalization adjudication, and engaged new processing strategies in an effort to "streamline" the naturalization process.
In the spring of 1996, however, just as CUSA moved into its most aggressive phase, media reports began to question the integrity of INS naturalization processing. Initially, INS' off-site testing program came under fire.
Beginning in 1991, a naturalization applicant was able to satisfy the English proficiency and knowledge of U.S. history and government requirements of naturalization by passing a test administered by certain private entities. This off-site testing program was expanded during CUSA. In April 1996, INS learned that ABC Television, following up on local media reports, was researching fraudulent testing practices by one of INS' largest off-site testing contractors. In June 1996, ABC aired a program that exposed many of the vulnerabilities of INS' outside testing program.
In May 1996, The Washington Times published an article about INS employees who criticized the acceleration of naturalization processing. The report quoted INS employees who questioned the motives of CUSA. The National Performance Review (NPR) of the Office of the Vice President had targeted CUSA for "reinvention" and NPR officials had traveled to the Key Cities to facilitate INS efforts at hiring new staff. This link between the naturalization initiative and the Vice President's Office during an election year fueled speculation and media stories that the rush to naturalize approximately one million applicants during fiscal year 1996 was an attempt to swell voting rolls with new citizens who were anticipated to vote for Democratic candidates, including President Clinton and Vice President Gore. In the wake of such reports, Members of Congress sought information from INS concerning various aspects of CUSA.
As Congress began to inquire, the media reports continued. Of particular concern were reports that some INS offices were naturalizing applicants so quickly that applicant criminal history reportsgenerated by the Federal Bureau of Investigation (FBI) after INS submitted applicant fingerprint cards for analysiswere arriving in INS offices only after the applicant had been sworn in as a United States citizen. These and other allegations of flaws in naturalization processing suggested that INS had sacrificed naturalization processing integrity in the name of processing applicants more quickly. In September 1996, the Subcommittee on National Security, International Affairs, and Criminal Justice of the House Committee on Government Reform and Oversight (the Subcommittee) held its first hearings concerning CUSA to explore the nature and extent of these processing flaws and the motives behind INS' accelerated naturalization initiative.
The portrait of naturalization processing that emerged from the September hearings did not allay congressional concerns. Although Executive Associate Commissioner T. Alexander Aleinikoff described the efforts INS had made to improve the off-site testing program, his testimony confirmed that off-site testing had lacked standards and not been monitored by INS. In addition, INS employees from Los Angeles, Chicago, and Dallas testified to the extraordinary rush imposed on naturalization adjudications during CUSA. That rush, according to these witnesses, meant that INS had naturalized people without ensuring that they were eligible. Some of the shortcuts to which these witnesses testified were that INS had not properly trained the new adjudicators hired for CUSA, had not conducted thorough applicant criminal history checks, had not provided applicants' permanent files to adjudicators for review before making decisions on naturalization applications, or, when those files had been available, had discouraged the thorough review of the file to determine whether the applicant had lawfully obtained the prerequisite permanent residency status.
The Subcommittee and other congressional subcommittees continued to investigate these allegations about CUSA and to seek information from INS, from the White House, and from the Office of the Vice President. Although Commissioner Meissner asserted that political motives had not influenced CUSA, documents provided to Congress suggested otherwise. In particular, e-mail messages to the Vice President and others from Douglas Farbrother, an NPR employee who was assigned to work on naturalization "reinvention" efforts in March 1996, connected CUSA's goals with the goal of naturalizing one million new citizens in time for the November 1996 election.
In the meantime, in response to congressional requests the Justice Management Division (JMD) of the Department of Justice engaged an outside accounting firm, KPMG Peat Marwick, to oversee a systematic review of CUSA naturalizations that INS would conduct using INS employees. The KPMG-supervised review first concentrated on determining whether each person naturalized during CUSA had a fingerprint check conducted by the FBI. Subsequent reviews would determine how many persons had naturalized during CUSA despite a disqualifying criminal history. The KPMG-supervised review continued over the course of the next two years, but even its preliminary results were troubling. In March 1997, JMD reported to Congress that of the 1,049,867 persons INS had then identified as having naturalized between August 31, 1995, and September 30, 1996,1 the fingerprint cards of 124,111 had been returned by the FBI as "unclassifiable," meaning that the fingerprints submitted had not been suitable for comparison. For an additional 61,366 persons, the FBI had no record of having conducted any fingerprint check. This data, therefore, indicated that for 18 percent of those persons naturalized during CUSA, INS had not conducted a complete criminal history background check.
This information was troubling to Congress not only because of what it reflected about criminal history checking procedures and thus the integrity of CUSA adjudications, but also because it suggested that INS had done little to improve its fingerprint processing procedures since 1994, when both the OIG and the General Accounting Office (GAO) had issued reports critical of those procedures and had recommended specific improvements that INS had agreed to undertake.
In late April 1997, then-Inspector General Michael Bromwich announced the investigation that is the subject of this report. It was to become the largest investigation of its kind ever undertaken by the OIG.
The OIG assembled the CUSA investigative team in July 1997. It included 19 agents, 3 analysts, 4 attorneys, and clerical staff working in each of the Key Cities and in Washington, D.C. The team analyzed documents and data from various sources including INS, congressional subcommittees, and the KPMG-supervised review. Those materials included policy memoranda, tapes of INS e-mail messages, applicant files, and the database created for the KPMG-supervised review team that identified the more than one million citizens naturalized during CUSA. Among the applicant files the team reviewed were those that the KPMG-supervised review team had identified as files of applicants who were naturalized despite a disqualifying criminal history. Ultimately, the team analyzed over 80,000 pages of documents.
In addition to this collection and analysis of documents, we conducted 1,829 interviews. We interviewed INS clerical employees, adjudicators, supervisors, training staff, computer analysts, and contractor employees. We interviewed the District Directors and other managers in the Key Cities. We interviewed personnel at the Vermont, Nebraska, Texas, and California Service Centers. We interviewed INS Headquarters staff in the Offices of Programs, Field Operations, Management, General Counsel, and Congressional Relations. We interviewed the managers at INS Headquarters who had responsibility for CUSA, including project manager David Rosenberg, Associate Commissioner Louis (Don) Crocetti, Executive Associate Commissioner William Slattery, Executive Associate Commissioner T. Alexander Aleinikoff, Deputy Commissioner Chris Sale, and Commissioner Doris Meissner.
Our interviews were not limited to INS employees.2 We interviewed managers and staff at the Federal Bureau of Investigation concerning criminal history checking procedures. We interviewed Department of Justice managers, including Assistant Attorney General for Administration Stephen Colgate, former Associate Attorney General (now Solicitor General) Seth Waxman, former Deputy Attorney General Jamie Gorelick, and Attorney General Janet Reno.
To examine any collateral influences on CUSA, we interviewed NPR staff members Douglas Farbrother, Robert Stone, Laurie Lyons, and the NPR Chief and Senior Advisor to the Vice President, Elaine Kamarck.
We sought interviews with other officials at the White House and Office of the Vice President. Among those who agreed to be interviewed were former Chief of Staff Leon Panetta and former Deputy Assistant to the President Kevin O'Keefe. Presidential advisors Harold Ickes and Rahm Emanuel, among others, declined our requests for interviews. Vice President Gore declined our request for an interview, but submitted written answers to our questions.
In the end, the OIG found that the CUSA initiative was developed as Commissioner Meissner and others asserted to Congress, that is, as a backlog reduction initiative designed to decrease naturalization processing times. We did not find that CUSA was developed, implemented, or otherwise directed to further inappropriate political ends. We did find, however, that the integrity of naturalization adjudications, already vulnerable before CUSA, suffered badly as a result of INS' efforts to process naturalization applications more quickly. We summarize our findings below.
Our investigation was extremely broad and addressed the many defects in CUSA that had been identified in the congressional hearings. We examined how CUSA was designed and implemented, and to what extent collateral influences played a role in the formulation and execution of the program. To isolate the impact of CUSA on naturalization processing integrity, we also considered the state of naturalization processing in the years preceding the program.
Naturalization processing before CUSA already displayed significant weaknesses that compromised the quality of adjudications. The central standard governing the naturalization adjudication inquirywhether the applicant was a person of "good moral character"was subject to varying interpretations in the Field. Administrative files or "A-files," the mechanism by which INS maintained an applicant's immigrant history, were often unavailable to adjudicators and thus not subject to review. As had been documented in both the OIG and GAO reports of 1994, INS' criminal history checking procedures were poorly administered. All of these factors were known to Commissioner Meissner and her staff, and yet INS decided to launch CUSA and thus accelerate a processing system that already was in need of considerable repair.
When Commissioner Meissner and her staff decided in the summer of 1995 to launch CUSA, they clearly sought to reduce the massive naturalization backlogs and reach "currency" in processing.3 They were not acting out of partisan political motives. They represented to the public and to Congress that INS would reduce the naturalization backlog within one year, an ambitious goal under the best of circumstances. However, they failed to address known system weaknesses before implementing a program that they knew would tax that system as it never had been taxed before. Given the known weaknesses in the system and the lack of commitment to repair the deficiencies, the promise of backlog reduction within one year also meant a certain recklessness about the quality of the resulting adjudications.
When it became clear, by early 1996, that CUSA was behind scheduleadjudicators had not been timely hired and offices were not yet open and productiveinstead of adjusting the goal of the program INS adjusted its approach to getting the work done: additional pressure was added to swiftly increase production. On March 1, 1996, Headquarters issued a memorandum to the Field that specifically instructed that all applications for naturalization received by the end of that month had to be completedeither denied or approved and the applicant sworn in-by September 30, 1996.
The delays in the implementation of CUSA caused concern outside INS. Representatives of CBOs were concerned that naturalization applicants would not receive the service they had been promised, and some complained to INS and to political officials.
Commissioner Meissner attended a meeting at the White House on February 9, 1996, with Harold Ickes, Elaine Kamarck, and others on the subject of the progress of the CUSA program. At that meeting, Kamarck offered the assistance of the NPR to INS in reaching its backlog reduction objective. Later that month and in early March 1996, NPR staff members Robert Stone, Douglas Farbrother, and Laurie Lyons met with INS officials and traveled to the Key Cities. NPR offered its assistance in streamlining hiring procedures, finding other federal government sources for available personnel to work on the project, and securing suitable office space. NPR's work did not have a discernible effect on those procedures at that time, however, because most of the facilities procurement and hiring efforts were already underway.
The evidence shows that the greatest influence NPR had on CUSA was an indirect one. NPR staff member Douglas Farbrother, concerned after his Key City tours that CUSA was not moving quickly enough, met with Deputy Attorney General Jamie Gorelick and senior INS officials on March 24, 1996, to discuss CUSA. Farbrother's participation at the meeting was cut short by the Deputy Attorney General, who reacted with anger at Farbrother's approach to Department of Justice affairs and asked Farbrother to leave the meeting. However, Farbrother's concerns about CUSA's progress did cause the Deputy Attorney General to inquire of INS officials about their efforts to meet their CUSA objectives.
As a result of the meeting with the Deputy Attorney General (DAG), INS redoubled its efforts and forcefully reiterated to the Field its commitment to achieve naturalization "currency" by the end of the fiscal year, then only six months away. INS Headquarters announced an "Expanded Naturalization Initiative" that authorized increased hiring and other resource expenditures provided that the District Directors of the Key Cities committed themselves to meeting the September 30 goal. INS Headquarters encouraged efforts to "streamline" the naturalization process, including methods of reducing the length of interviews. The length of time during which offices had been, at least by policy, expected to wait before adjudicating an application without the applicant's permanent file was reduced from six months to 30 days.
Although we found no explicit instruction from INS Headquarters to ignore standards that had previously existed, neither did we find any instructions concerning how to ensure quality during this period of heightened production. Once INS had rededicated itself to the goal of CUSA, the principle of increased production was pursued at the expense of accuracy in the determination of applicant eligibility, and a process previously regarded as lacking safeguards became even more vulnerable.
The evidence shows no other direct or indirect impact of the involvement of NPR on the CUSA initiative. By May 1996, NPR officials were no longer involved in the program. That said, the evidence does suggest that the motives for the involvement of NPR were mixed. Elaine Kamarck, Robert Stone, Laurie Lyons, and the Vice President asserted that good government-in this case naturalization backlog reductionwas the reason for NPR's involvement in CUSA. Farbrother, as was illustrated by his e-mail messages provided to Congress, believed that the CUSA program had a deadline that was directly connected to the upcoming election. He told the OIG that eligible applicants had to be naturalized in time to register to vote in November 1996, when he presumed that such voters would support the Clinton/Gore ticket. He also told the OIG that he had learned from Elaine Kamarck that this was one motive for NPR's involvement. Kamarck, Stone, and Lyons do not corroborate Farbrother in this regard, but it is clear that the end result of CUSA's successthe naturalization of one million people by September 30, 1996was appreciated by community-based organization (CBO) representatives and by some administration officials who agreed to be interviewed as a goal that could have electoral benefits.
Thus, there were some who hoped for a political benefit from INS' naturalization effort. We did not find evidence that officials of INS or the Department adopted this as a goal of CUSA or as a reason to press the goals of the initiative, and did find instances in which such a characterization by CBO representatives was expressly rejected. Nonetheless, many INS employees questioned the legitimacy of CUSA because they suspected it grew out of partisan purposes.
In sum, NPR's influence was one stimulus among several that recommitted INS to its production deadlines, and efforts to meet those deadlines undermined the quality of naturalization adjudications during fiscal year 1996. As to the many other allegations concerning INS' failure to safeguard naturalization integrity, our investigation confirmed the most serious: INS advanced its production goal despite the known risks that accelerated production would pose to its proper evaluation of applications for citizenship. Our specific findings regarding these processing errors, and our recommendations to improve the naturalization process, are described below.
Our report first provides an overview of the CUSA program and addresses the specific allegations concerning its design and implementation. We then address the naturalization processing weaknesses according to three specific subject matter areas: interviews and adjudications (including issues regarding adjudicator training), file practice and policies, and criminal history checking procedures.
Soon after she became Commissioner of INS in October 1993, Commissioner Meissner began to plan a major naturalization initiative for fiscal year 1995, a primary component of which would be to encourage eligible permanent residents to apply for citizenship. The promotion of naturalization for eligible applicants was something the Commissioner had often publicly supported.
During 1993 and 1994, however, INS had not kept pace with the increasing demand for naturalization. The pending caseload had quadrupled in three years, from 135,652 in 1992 to 481,580 in mid-1995. By March 1995, having learned from District Directors that they were overwhelmed by naturalization backlogs, Commissioner Meissner understood that encouraging applicants to apply for naturalization was incompatible with the increasing workload in the Field. Accordingly, her promotion of naturalization was transformed into a concentration on backlog reduction. That same month, Commissioner Meissner declared that of INS' work on applications for various benefits, addressing the naturalization backlog would be of the highest priority.
On August 31, 1995, Commissioner Meissner announced the CUSA initiative and its goal of reducing naturalization application processing time to six months by the following summer. Reaching that goal would mean that INS would have to process more than one million cases in fiscal year 1996. She also announced that the initiative would focus on five large districts that were then facing the largest backlogsLos Angeles, New York, San Francisco, Miami, and Chicago.
While it has been alleged that INS chose the five Key City Districts because of the anticipated voting behavior of the potential citizens in these geographical areas, we found no evidence to support this claim. INS officials asserted that it was the size of the backlog in the largest Key City Districts, and not any characteristics of the applicants, that influenced their choices. Witness interviews and the statistical evidence support their assertions.
Members of Congress also alleged that INS had actively solicited more than half of the total applicants naturalized during CUSA as a means to swell the voting rolls in anticipation of the 1996 election. The evidence shows that Headquarters in fact scaled back its emphasis on naturalization promotion during CUSA in comparison to what it had encouraged during the previous year, although this change in priorities away from naturalization promotion had not been clearly communicated to the Field. Accordingly, we found evidence that some promotional efforts lingered and those efforts could be characterized as "soliciting" applicants for naturalization during CUSA. However, we found no evidence to support the allegation that INS continued to solicit naturalization applications because of a desire to increase the number of potential voters in the 1996 election.
In publicizing CUSA, INS described the naturalization initiative as being about "the three P's: people, process, and partnerships. The "people" referred to the number of new temporary employees INS intended to hire; the "process" was included in the CUSA slogan as a reminder that INS intended to make the process more efficient; and "partnerships" referred to the new ways in which INS would involve CBOs in the process. This three-part strategy was what INS used to reach the CUSA goal.
In the implementation of each component of the CUSA strategy, however, INS single-mindedly concentrated on how each could serve to increase the rate at which INS processed naturalization applications. We found that INS paid little attention to the impact this increased rate of processing would have on the quality of adjudications. One example from each category will help illustrate this point.
First, the "people" of the CUSA planthe hundreds of new adjudicators with no previous INS experience hired to interview applicants and determine their eligibility for citizenshipwere superficially trained and unprepared for anything but the most routine adjudication of an "approvable" application, yet their assignments were not limited to such cases. In many instances, INS assigned the new adjudicators to work in environments without adequate supervision by experienced personnel. We found the most egregious instance of this inappropriate deployment of new hires in the Garden City, New York, CUSA office, where approximately 100 new, temporary adjudicators were assigned to work under the guidance of only six experienced employees.
As to "process" improvements implemented during CUSA, INS officials similarly focused on the importance of swiftly "moving" cases and not on the consequences of such acceleration. INS employed new methods before they had been validated and before involved personnel were appropriately trained. INS' implementation in February 1996 of a "Direct Mail" system for naturalization applications was one such example.
Direct Mail was a method by which applicants sent applications to the service centers rather than the district offices for initial processing. It had already been used for other INS benefit applications, and had been under consideration for naturalization applications for several years. The theory of Direct Mail was that it made processing more efficient by assigning the initial tasks to regional service centers that were accustomed to handling a large volume of paperwork, thus relieving district offices of clerical burdens.
INS failed to adequately train its service center and contractor staff for the transition to Direct Mail. One consequence of this lack of training was illustrated by the Vermont Service Center's failure to properly process applicant fingerprint cards during the summer of 1996. This shortcoming further compromised New York District's ability to have a completed criminal history check for applicants who applied through Direct Mail, as described in our chapter on criminal history checking procedures.
The transition to Direct Mail during CUSA was also flawed because INS changed the fundamental design of Direct Mail at the eleventh hour. The historical concept of Direct Mail gave the service centers responsibility only for the initial receipt and processing of applications and did not include responsibility for ordering and maintaining A-files. At the urging of the site coordinator for the Los Angeles District, two months before Direct Mail began the design was changed to shift responsibility from the districts to the service centers for the finding, storing, and maintaining applicant files until the time that the naturalization interview was scheduled at district offices. The change was theoretically beneficial for Los Angelesbecause the California Service Center was close to the District Office and already housed many of the necessary filesbut not for other Key Districts. The design change meant a dramatic movement of files from districts to the service centers and back again, although the consequences of this file movement were not considered by those who advocated the transition to Direct Mail. As a result, files were often unavailable when adjudicators needed them for the naturalization interview.
Direct Mail's success also depended to a large extent on the reliability of INS' computer systemssystems on which the service centers and the districts had to rely, for example, for accurate data about naturalization cases, for file transfers, and for the scheduling of interviews and ceremonies. INS had recognized before CUSA that NACS, the automated system used for naturalization processing, was overloaded with data and was unreliable, and yet CUSA was not delayed until NACS could be overhauled. During Direct Mail, INS made repeated adjustments to NACS' faulty scheduling system because without the scheduling of interviews CUSA would grind to a halt. However, INS did not improve NACS' file-ordering functions or other failures that were less related to production (without the permanent file, adjudicators could proceed with a temporary file, as described below). NACS was further impaired by the huge demands CUSA processing placed on it, and essentially collapsed under the strain.
Finally, concerning the "partnerships" prong of the CUSA strategy, INS Headquarters encouraged the Field to work closely with CBOs by conducting interviews at sites in the community where CBO representatives would pre-screen applications and make applicants feel more at ease with the process. Although INS Headquarters recognized the need to establish criteria for CBOs working with INS and the need to issue guidance to the Field in managing its relationship with the CBOs, INS Headquarters failed to do so. As a result, districts developed their own community outreach procedures on an ad hoc basis, and some were more vigilant than others about processing integrity concerns.
We did not find, however, as was alleged during the summer of 1996, that INS formed partnerships with particular CBOs to increase the number of persons who applied for citizenship or to integrate partisan voter registration efforts. To the contrary, we found INS' approach to partnerships was disorganized and unplanned, thus permitting occasional abuses of the system, particularly in the Chicago and Los Angeles Districts.
By the time of the March 1996 "Expanded Naturalization Initiative," INS was behind schedule. Headquarters stepped up its emphasis on the importance of meeting production goals. The goal of backlog reduction was translated into numerical goals, goals often perceived as quotas in the Field. To achieve the numerical goals, the Field had to increase and maintain a high rate of interviewing in the summer of 1996 and had to increase either the size or the frequency of naturalization ceremonies. Steps taken to enhance production included extending the number of work hours in a day and the number of workdays in the week. Adjudicators throughout the Key Cities perceived that they were working under considerable pressure to quickly complete, and not necessarily thoroughly review, applications for naturalization.
We did not find that the pressure to work harder, to work longer hours, or to work more days, was by itself the cause of employee complaints about the production pace or its deleterious effects. We found that what was debilitating to adjudicators-and thus to the quality of their work-was this emphasis on working long hours combined with having to work without adequate adjudicative standards and guidance, without the requisite tools necessary to perform a thorough adjudication, and with supervisors who often emphasized the importance of increasing the number of cases completed. In such an environment, many adjudicators were led to believe that it was quantity, not quality, that mattered.
Many of the allegations made by Members of Congress directly implicated the integrity of the naturalization interview, including concerns that adjudicators were not properly trained to perform their duties, that they lacked sufficient time to conduct the interview properly, that they were pressured or directed to limit their inquiries, and that the standards by which applicants were judged became more lenient during CUSA. We found that these allegations were supported by the evidence. We found that many of these problems pre-existed CUSA, and although there was no evidence that INS deliberately repealed pre-existing safeguards or purposefully sought to lower the bar for citizenship, INS' poor planning of CUSA strategies and the rush to produce further diluted the quality of naturalization administration, and did so in predictable ways.
We note at the outset that it is impossible to precisely quantify the number of CUSA adjudications that were affected by these or other problems detailed in our report. We cannot establish how many applicants who were approved for naturalization would have been found ineligible had appropriate procedures been followed. The KPMG-supervised review identified some of these cases, but even in that review the only cases that could be considered as having been wrongly adjudicated were those for which there was evidence available after the fact that the adjudicator had made the wrong decision (for example, the existence of a disqualifying criminal history or the absence of the requisite number of years as a permanent resident). No amount of after-the-fact study can determine what might have been revealed to a better-trained adjudicator with adequate supervision and the appropriate tools to conduct the review and determination. If such an interview is never conducted, the potentially disqualifying information is never revealed.
We further note that the number of applicants naturalized during CUSA who were in fact ineligible for citizenship was not the most important question informing our inquiry. Indeed, it may have been that the huge applicant pool was full of persons who were qualified for citizenship and did not pose an actual threat despite INS' vulnerable procedures. The more important inquiry was what safeguards existed so that the benefit of U.S. citizenship is conferred equitably, consistently, according to a properly administered set of laws, and with enough integrity to minimize the risk of conferring it on ineligible applicants. As a result of that inquiry we found that necessary protective measures that did exist were improperly administered, while others were never established.
Before we detail our findings concerning naturalization adjudications during CUSA, we provide a brief explanation of the naturalization interview process.
All applicants for naturalization must be interviewed under oath before they are approved for citizenship. The interviewing officer, a District Adjudications Officer (DAO), makes the determination of eligibility.
The interview typically consisted of two prongs, a testing prong and an evaluation of the applicant's other qualifications for citizenship. Before and during CUSA,4 the applicant could satisfy the reading and writing portion of the English proficiency and Civics knowledge requirements either by passing a test administered at the interview or by presenting a certificate from an authorized outside organization attesting to his or her proficiency. Otherwise, DAOs tested the applicant's ability to speak, read, and write English, and tested his or her knowledge of the history and government of the United States. If the applicant presented a valid certificate from an outside testing entity, then the DAO at interview only verified the applicant's oral language skills. If the applicant was unable to satisfy the English or Civics requirements, the case would be continued and the applicant would be rescheduled for one additional opportunity to demonstrate his or her proficiency or knowledge.
DAOs reviewed the applicant's written responses to questions on the naturalization application (the Form N-400) in order to determine whether he or she met the other eligibility criteria for citizenship. The N-400 included questions designed to elicit information that would inform the DAO's determination of whether the applicant was a person of "good moral character" (such as whether the applicant had ever been arrested for a criminal offense), a central requirement of naturalization eligibility, and whether he or she was "attached to the principles of the Constitution." If the applicant's responses triggered concern that he or she did not satisfy the prerequisites for citizenship, the adjudicator could either deny the application or continue the case for additional review or for the presentation of additional documentation.
If the application was approved, the applicant was told (or later notified by mail) of the naturalization ceremony date. If the applicant applied in a jurisdiction where the federal court maintained the responsibility for administering the naturalization oath, the ceremony would usually be scheduled approximately two months later. Otherwise, applicants could be sworn in by INS officials. Between the time of the approval and the ceremony, the applicant had to remain eligible for naturalization. Information bearing on eligibility received by INS between the date of the interview and the date of the ceremony could lead to a reconsideration of the case.
Because of the short-term goal for CUSA and the brief period of employment anticipated for temporary adjudicators, INS devised a special training that was much shorter than the customary training for adjudicators that lasted 12 to 16 weeks. Its basic premise was that the training could be reduced to approximately one week of classroom instruction and one week of on-the-job training by extracting from the traditional officer training curriculum those portions relevant to adjudication of the naturalization application.
This abbreviated training was not intended to prepare new adjudicators to assume a full-fledged adjudicative role. Instead, the training was based on two assumptions: (1) that cases would be screened in advance so that the new adjudicators would only be presented with simple, straightforward cases and (2) that adjudicators would function in a "primary-secondary" environment that would allow them to immediately pass on to experienced DAOs cases that had not been recognized as complex during the screening.
Headquarters, however, did not direct the Field to implement a pre-screening or a primary/secondary environment or even advise the Field that the training had been premised on these assumptions. As a result, none of the Key City Districts screened the cases before they were received by the new adjudicators and only one office, the El Monte office in the Los Angeles District, implemented a primary/secondary environment.
In every district, many of the CUSA adjudicators and supervisors interviewed by the OIG criticized the training for temporary adjudicators. They told the OIG that the training did not provide adjudicators with enough information about the law or the overall immigration process to evaluate applicants properly, and did not train them to understand the contents of an A-file or to understand a rap sheet.
The weaknesses of the training were exacerbated in some districts by the failure to provide adequate supervision in the Field. Headquarters left the decision concerning how many supervisors to assign to any given location to the Field, and some districts, notably the San Francisco and the Miami Districts, adopted local procedures in an attempt to support the new hires with access to experienced personnel. In others, like in the New York District as noted above, local officials staffed the largest naturalization site in the country-the Garden City office-with more than 100 temporary officers and only 6 permanent employees, 5 of whom worked as supervisors.
The Immigration and Nationality Act requires that an applicant be of "good moral character" in order to be approved for citizenship. Federal regulations provide a list of elements that preclude a finding of "good moral character," many of which pertain to an applicant's criminal record. The questions on the N-400 focus on eliciting information about these precluding characteristics. However, the regulations also specifically state that the absence of a precluding factor does not necessarily mean that the applicant is, in fact, of "good moral character." Although the law provided DAOs with guidance concerning the boundaries beyond which an officer could not find that an applicant was of "good moral character," within those boundaries the evaluation of the applicant's character was left to the discretion of the officer.
INS recognized before CUSA that the "good moral character" standard was applied inconsistently in the Field and yet did not clarify it before implementing CUSA. In the absence of standards, INS had left the determination of whether, when, and to what extent to explore issues not completely fleshed out in the answers to the N-400 to the subjective judgments of individual adjudicators. The broader the exploration, the more likely the adjudicator would discover potentially disqualifying issues if they existed. This determination had always represented a kind of balancing test between the perceived likelihood of obtaining relevant information against the investment of time involved.
During CUSA, time was at a premium. Many supervisors discouraged adjudicators from making any inquiries beyond the questions specified in the N-400 when determining whether the applicant was a person of "good moral character." Such inquiries were neither specifically sanctioned nor specifically prohibited in the laws governing naturalization. Moreover, they took time. The problem with narrowing the inquiry to the boundaries of the N-400, however, was that it made the interview less likely to uncover potentially disqualifying information if there was any to be found. Thus, this discouragement contributed to the adjudicators' sense that approvals were favored over the thorough exploration of an applicant's eligibility.
Well before CUSA began, INS was aware that its language and government testing components of the naturalization interview were poorly and disparately administered. INS had not developed an objective standard or effective test for determining whether an applicant spoke English at the requisite minimum level. INS recognized the need to improve in these areas and had planned to develop both a standard and a test for spoken English, and a revised regulation to improve the administration of outside testing.
However, these new standards and improvements were not targeted for implementation in time to affect CUSA cases. In the meantime, DAOs continued to apply the same ambiguous standards to the increasing number of cases. By the end of CUSA, on September 10, 1996, EAC Aleinikoff testified before the Subcommittee that "lack of standardization among INS offices [had], for some time, led to inconsistent standards" in the testing of English and Civics. INS had taken no meaningful steps to insulate CUSA adjudications from the adverse affects of the weaknesses it had identified before the program began.
Absent an objective standard for determining the minimum level of spoken English required of an applicant, DAOs relied on the general guidance that governed their determination before CUSA-whether the applicant spoke ordinary English, or English at an "elementary" or third-grade level. This guideline was not further defined. New adjudicators during CUSA attempted to formulate some understanding of minimum proficiency at a time when they were under pressure to avoid unnecessary continuances and to complete cases. When no real standard existed, however, management pressure to reduce the number of continuances reinforced interpretations in favor of approval. We found that there was a widespread agreement among the adjudicators we interviewed that INS did not properly enforce the English-language requirement of naturalization.
We also found that the manner in which the Civics test was administered at the naturalization interview during CUSA varied from district to district and even among adjudicators in the same district.
Finally in regard to English and Civics testing, we found that fraud by outside testing entities had not been adequately addressed during CUSA. This failure represented a continuation of a problem that began with the inception of the outside testing program in 1991. Despite periodic reports from the Field that it was encountering evidence of fraud by testing organizations, INS had not instituted a national monitoring system. Additionally, although the Field had advised Headquarters that it needed stronger guidelines and standards in order to conduct local oversight, Headquarters had not responded. Over time, the Field developed the view that Headquarters was not interested in oversight of the testing entities and therefore the Field did not devote its resources to that purpose. Despite the perception of fraud in outside testing, INS plans for 1996 called for the expansion of the program.
INS officials undertook preliminary steps in fiscal year 1996 to address fraud, but their timing was more closely linked to increasing pressures from the media and Congress rather than an urgency to ensure the integrity of the program before its planned expansion. In any event, these efforts were too little and too late to have any beneficial impact on the integrity of the testing program during CUSA. Many adjudicators reported to the OIG that they regularly encountered applicants who presented passing certificates but who could not speak even simple English, from which they inferred fraud in the outside testing program.
INS officials we interviewed often characterized their approach to CUSA as "business as usual" only on a much larger scale. Thus, any failure to provide its adjudicative corps guidance in the determination of naturalization eligibility did not pose a "new" risk, just a more prevalent one because so many more applications would be adjudicated without such guidance. We identified one way, however, in which INS officials knew that naturalization applications received during CUSA might pose a higher risk of ineligibility than those received during previous years: more than one million persons became eligible for naturalization on December 1, 1995, who had obtained their permanent residency through an INS program widely regarded as having been rife with fraud.
The Legalization program of the late 1980s had included provisions under which "Special Agricultural Workers" (SAWs) could obtain residency (first temporary, then permanent) in the United States by demonstrating that they had worked in certain agricultural jobs during specified time periods. The applicants proved their qualifications by providing documentation to INS attesting to their work history. There was wide consensus within INS that the use of fraudulent documents had been prevalent in the SAW program. Although estimates varied (at one Headquarters meeting, there was an estimate that 70 percent of all SAW cases had involved fraud), no one denied its existence. For that reason, before CUSA began the Chair of the U.S. Commission on Immigration Reform, Barbara Jordan, had recommended that INS pay special attention to applications for citizenship by SAW applicants. In addition, the Chairman of the Senate Judiciary Committee's Subcommittee on Immigration, Alan Simpson, had received assurance from INS that it would respond appropriately to evidence of previous immigration fraud by all applicants for naturalization.
Despite these warnings and promises, however, evidence of fraud was not adequately explored and, in some instances, was completely disregarded. New adjudicators were not trained to recognize or explore potential issues of fraud by SAW applicants and no guidance was sent to the Field to ensure that experienced examiners were alert to the importance of this issue. Even worse, many adjudicators believed that they were prohibited from reviewing documents in an applicant's file that pertained to the SAW application or believed that if evidence of SAW fraud somehow emerged they could not consider it in the determination of citizenship.5
Throughout CUSA, INS Headquarters encouraged the Field to "streamline" the naturalization interview. As a result of the "Expanded Naturalization Initiative" of March 1996, the Office of Programs (Examinations) issued its "Naturalization Process Changes" memorandum on May 1, 1996, that further encouraged such streamlining. The Naturalization Process Changes memorandum recommended that the Field consider innovative examination strategies, noting that "interviewing pace" needed to be "recalibrated to take advantage of the time-saving strategies." The El Monte interview, anticipated to be pared down to five minutes, was lauded as a CUSA model, even though at the time the memorandum was distributed the site had only been open for business for less than one month and the impact of its approach on the quality of naturalization adjudications had not been reviewed. Even without validation, although the testing of Civics and written English proficiency had been eliminated from the El Monte interview,6 the notion that a thorough review of the applicant's immigration history, evaluation of his or her spoken English, consideration of all of the answers provided in the N-400, and determination of "good moral character" could be reduced to a 5-minute exchange between the applicant and a superficially trained, inexperienced DAO showed INS Headquarters' disregard for the importance of the naturalization interview.
"Off-site" processing, where DAOs would go to a location in the community to interview applicants whose applications had been "pre-screened" by a CBO, was also encouraged as a time-saving strategy. The Naturalization Process Changes memorandum held out the Chicago District's off-site processing as the model to emulate, and yet Central Regional officials had reviewed Chicago's program the previous year and sharply criticized it as reducing the naturalization interview to the point of meaninglessness, and as a program without safeguards to ensure that representatives of the CBOs did not place inappropriate pressure on INS staff. Nonetheless, the Chicago off-site processing program was successful at completing many thousands of cases and at pleasing the participating CBOs, and thus it earned high marks at INS Headquarters as an effective CUSA strategy with none of its impact on the quality of those adjudications.
We found that INS did not properly administer its file practices during CUSA.
The A-file is the repository of all immigration-related documentation regarding the applicant. The permanent file contains the applicant's immigration history, and only by reviewing it could an adjudicator ensure that the applicant had obtained his or her residency status lawfully and as represented on the N-400. Although INS' automated systems reflected information such as when, where, and how the applicant had obtained residency status, only the applicant's file could reveal the actual information submitted in order to obtain the previous benefit. Review of such information was crucial if INS was to be vigilant for fraud in obtaining any previous immigration benefit. Nevertheless, even before CUSA began, INS was relying to a significant degree on temporary files, which typically contained only the naturalization application and a printout from the INS Central Index System showing the applicant's immigration status.
Pursuant to a policy established in 1980, adjudicators could use a temporary file if there had been diligent efforts for six months to obtain the permanent file. Unfortunately, no standards defining the meaning of diligence were promulgated and the seriousness of the effort to locate the file varied from district to district. In Los Angeles, the nation's largest district, for example, there was essentially no requirement to search for the A-file before using a temporary file. Moreover, none of the Key Districts before CUSA required that the efforts to locate the file be made before the interview, the time when the file was most relevant.7
During CUSA it became even less likely than before that adjudicators would have the permanent file in time for the interview. First, INS failed to adequately and timely shore up its records staff, despite the obvious burdens that would be placed on them by massive data-entry projects and new interview sites. We offer one illustration concerning the work of the Naturalization Data Entry Center (NDEC) as an example of the consequences of inadequate records-related planning during CUSA.
At the very beginning of CUSA, the Los Angeles District mounted a major effort at the building housing the California Service Center to data-enter naturalization applications that had been submitted but never entered into the system. Under INS' automated procedures, immediately after data-entry, the computer requests the applicant's permanent file from the district office records room that has the file. The NDEC project, completed during several weeks in August and September 1995, resulted in a sudden request to the Los Angeles District records room for 170,000 A-files, a number that equaled more than one normal year's worth of file requests. This request was made to a records staff that had not increased in size and that continued to have responsibilities for other INS programs. As a predictable result, the records staff was immediately backlogged and was even less able than it had previously been to produce permanent files for applicant interviews.
The NDEC project also data-entered approximately 50,000 applications from the Miami District that had not been entered into INS' system. Project planners, however, had failed to consider that INS' computer system automatically requests that files be sent to the location of the "requestor," or the person entering the application data. This meant that A-files relating to Miami applications that would ultimately be adjudicated in the Miami District were "ordered" by the automated system to be sent to the Los Angeles District. As a consequence of insufficient coordination between the NDEC staff, CUSA planners, and district records rooms, the applications were being shipped back to Miami, while the corresponding permanent files for these cases were moving in the opposite direction to Los Angeles.
In addition to the problems engendered by this huge, one-time data-entry effort, INS' file practices exhibited broader systemic problems. New sites were located in different buildings separate from each district's main records room and therefore a file had to travel farther, and through more intermediaries, to arrive at the interview site. This contributed to the decreased likelihood of A-file usage during CUSA. For example, an A-file requested from outside the district by the CUSA satellite office in San Jose would first arrive at the San Francisco District Office, then be sent to the records room of the San Jose Sub-office, and finally to the CUSA site itself. The result was predictablea greater dependency on temporary files in San Jose.
The implementation of Direct Mail for the naturalization application also increased reliance on temporary files, as we discussed above. The increased file migration required by new CUSA sites and the transition to Direct Mail had a predictably adverse impact on A-file availability, particularly during a time of dramatically heightened production.
In addition to the adverse but inadvertent impact that resulted from planning weaknesses, we found processing changes that reduced the emphasis on the importance of using the A-file. First, the site that was regarded as the premier example of "naturalization streamlining," the El Monte office in the Los Angeles District, had a design that permitted only a cursory review of whatever file was available as the interview was taking place. Second, on May 1, 1996, the Benefits Division (Office of Examinations) modified the longstanding requirement that a 6-month diligent search for the A-file was a prerequisite for reliance on a temporary file by reducing search time to just 30 days. Ironically, the potentially negative impact of the policy change was mitigated by the fact that in most of the Key City Districts the instruction was either ignored, unknown, or deemed irrelevant because temporary file use was already so common. In the one district, Miami, that did change its practices in response to the new policy despite managers' disagreement with it, the result was that more cases were adjudicated with temporary files.
The vulnerability of relying on temporary files is often not readily apparent. The information in the file may not be facially disqualifying but may raise issues that lead to evidence that the applicant is not eligible. In some instances, however, the information in the file may be quite explicit. INS' response to the possibility of fraud by SAW applicants illustrates both aspects. Although INS assured Congress that it would investigate whenever evidence of SAW (or other) fraud was discovered during interview and deny citizenship whenever that evidence was substantiated, such investigation was generally dependent upon the presence of an A-file. Without it, suspicions of fraud were unlikely to arise and even less likely to be fully explored.
When the initial statistics from the KPMG-supervised review of CUSA cases became available in early 1997, Commissioner Meissner acknowledged that INS' primary policy regarding fingerprint checks, the "presumptive policy," had been flawed. The policy, established in 1982, permitted INS offices to presume that an applicant had no criminal record if no rap sheet was returned to INS after a designated amount of time had passed (the "presumptive period") since the fingerprint card had been sent to the FBI for analysis. At the same time that INS officials conceded the flaws in the presumptive policy, they continued to maintain that they had made every effort to safeguard the systemindeed, to improve it.
We examined the INS fingerprint checking process in detail. We found that although the presumptive policy was inherently flawed, the myriad problems with INS' criminal history checking procedures could not be attributed only to having adopted an ill-conceived policy many years before CUSA. INS had been repeatedly warned about systemic weaknesses, including but not limited to the presumptive policy, and failed to respond to them before launching CUSA. The volume of cases, the new processing strategies like Direct Mail, and the accelerated rate of production during CUSA all exacerbated these problems. The one innovation INS implemented in the name of improving fingerprint processing-the opening of a centralized Fingerprint Clearance Coordination Center in June 1996-was poorly planned, poorly timed, insufficiently staffed, and inadequately explained to the Field. Instead of improving INS' fingerprint processing during CUSA, it served to weaken procedures that were already deficient.
We also reviewed INS' "bio-check" procedures. The bio-check is a comparison of applicant biographic data to data available at the FBI and CIA to determine whether the applicant is the subject of any intelligence, counter-intelligence, organized crime, or terrorism investigations. In addition to investigative data, this check could reveal an applicant's criminal history from a foreign country or outstanding domestic or international arrest warrants. We found that INS bio-check procedures suffered from the same kinds of deficiencies that marred INS' processing of naturalization applicants' fingerprint cards and resulting criminal histories. As sometimes happened in INS' processing of fingerprint cards, staff ignorance of appropriate bio-check procedures also resulted in the destruction of relevant material returned from the FBI.
In searching for explanations for how INS could so poorly administer, and for so long, the few methods in place for checking naturalization applicants' backgrounds, we identified several factors. First, INS regarded the pool of naturalization applicants as a group of persons at very low risk for a disqualifying criminal history. Second, INS approached criminal history checks with an attitude that the cost of obtaining a definitive criminal history check for every applicant outweighed their value, given the presumed low risk of a criminal history. Finally, thorough criminal history checking procedures, like other procedural safeguards such as applicant file review, were overshadowed by the priority of completing more than a million cases in one year, and thus these checks suffered from the mismanagement that affected other areas of naturalization processing.
We also examined INS' response to congressional inquiries about criminal history checking procedures. We found that INS answers to congressional inquiries concerning criminal history checking procedures during CUSA were replete with mistakes that could have been avoided had INS officials paid sufficient attention to the information then available from the Field.
Our report addresses INS criminal history checking procedures in considerable detail. For a full understanding of the risks taken by INS in regard to applicant criminal histories, we refer the reader to those sections of the report. What we offer here is only a synopsis of the broader findings we made as a result of our investigation.
INS had no definitive policy statement or articulation of the rules regarding fingerprint card processing before or during CUSA. Accordingly, there was no definitive source to turn to for a description of the presumptive period or the proper procedures to be followed if the FBI rejected a fingerprint card.
According to Headquarters officials' testimony before Congress, the presumptive period was 60 days. Although all districts understood that they could adjudicate cases in the absence of a response from the FBI, they did not uniformly know that the waiting period was this long. The policy in Miami, for example, was to wait 45 days. The understanding of the length of the presumptive period varied within New York, Chicago, and Los Angeles Districts, and some supervisors in each district believed that the presumptive period was 45 days. This difference was not insignificant. In 1996, approximately 77,000 cards took less than 60 days but more than 45 days for the FBI to process (excluding backlogs and delays in submission).
In the years before CUSA, INS Headquarters did not recognize or emphasize the importance of resubmitting rejected fingerprint cards. Although INS was aware that many thousands of fingerprint cards were returned to submitting offices because they were not suitable for analysis, neither INS Operations Instructions (OIs) nor other official sources described what the Field should do upon receiving a fingerprint card rejected by the FBI because of "masthead errors," or errors in filling out the required biographical data at the top of the fingerprint card. If such cards were not resubmitted, the applicant's criminal history was never checked by the FBI. In regard to fingerprint cards rejected because the fingerprints themselves were not suitable for comparison-deemed "unclassifiable" by the FBI-since 1994 INS Headquarters had instructed the Field that the decision to obtain new fingerprints from the applicant was discretionary.
Given the lack of guidance, different districts adopted different approaches. We found that two Key City Districts, New York and San Francisco, had made efforts before CUSA to resubmit cards rejected from the FBI, while Los Angeles, Miami, and Chicago had not. Even in those districts that did attempt to resubmit rejected cards, however, we found that INS had no systematic way of delaying the naturalization interview until the second card had been analyzed. These districts did not recalculate the presumptive period to begin from the date of the second submission of fingerprints, and thus applicants could be naturalized before a check had been completed by the FBI.
INS officials asserted to Congress that reliance on the presumptive period before CUSA had not been problematic because the large backlogs expanded the time between the receipt of the application and the date of interview to many months, and, in some places, more than one year. They implied that the presumptive period, whether understood as a period of 45 or 60 days, therefore would be respected in every case. The fallacy in this reasoning, however, was that it incorrectly assumed that the Field sent fingerprint cards for analysis immediately upon receipt of the application. In the Los Angeles and the Chicago Districts, for example, fingerprint cards were not detached from the application and sent to the FBI until these districts took steps to schedule the applicant's interview-usually just two months before the interview date. It also assumed that when rap sheets or rejected cards were returned by the FBI, they were promptly processed, another assumption that was not borne out by practices in the Field.
INS had also failed to emphasize, before CUSA, the importance of having any available criminal history report available for the adjudicator at the naturalization interview, when it could be used as a tool for questioning the applicant. Incoming criminal history reports had to be timely interfiled in applicant files, and those files used at interview, if the reports were to influence the adjudication of the application. As early as 1994, we found that in the Chicago District rap sheets were stored in a box and not interfiled, and if an applicant had a relevant criminal record the adjudicator had to search the box to find it. Also, in the Los Angeles District it was customary to review incoming rap sheets after approval of the application but before ceremony, relying on a last-minute "pull-off" system if the record revealed a criminal history that should have disqualified the applicant. As the volume of applications increased and the processing time decreased, this pull-off system gave way under the strain.
As described in a February 1994 inspection report by the OIG, we found that INS was not resubmitting rejected fingerprint cards and not timely interfiling rap sheets. We reported that FBI checks were a necessary procedural safeguard in the naturalization process and recommended that INS resubmit rejects and ensure that criminal history reports returned by the FBI be available to adjudicators at interview.
Although in 1994 INS concurred with the OIG only in part that fingerprint checks were necessary for naturalization applicants, it agreed that its procedures were in need of repair. Accordingly, INS promised to take a series of ameliorative steps. We found as a result of this investigation that INS took no substantive steps to repair the deficiencies we previously highlighted other than to form a working groupthe Fingerprint Enhancement Working Groupwhose recommendations were not implemented before CUSA. As Associate Commissioner Crocetti told the OIG during this investigation, INS' response to the OIG's 1994 inspection report was "lip service."
In March 1994, after the OIG issued its report, for a brief period INS suspended all fingerprint checks for naturalization applicants as a cost-cutting measure. The suspension was quickly criticized by Congress and reversed. Congressional concern, however, led to a follow-up review by the GAO of INS' fingerprint checking procedures in relation to naturalization applications. GAO issued its report in December 1994.
The GAO report reiterated the importance of proper fingerprint processing procedures, but it went further than the OIG report by calling for the elimination of the presumptive policy.
INS agreed with the GAO's recommendation to clarify its fingerprint processing procedures. However, INS resisted as too costly the elimination of the presumptive policy. An INS internal recommendation to use the FBI's automated billing data-data that reflected whether an applicant's fingerprint card had been analyzed and, if so, whether a record had been found-was shelved until it could be fully incorporated in a new automated system for naturalization that was then only in its earliest planning stages.
INS regularly asserted that it was complying with most of the GAO recommendations.8 However, the evidence shows that it made no substantive changes in fingerprint processing procedures before launching CUSA.
As we noted above, the new case-processing strategies adopted by INS for CUSA emphasized accelerated naturalization production. Two CUSA strategies, the NDEC project and the transition to Direct Mail for naturalization applications, were implemented without thorough evaluation of their impact on the proper processing of applicant fingerprint checks.
INS used the NDEC to data-enter during a 6-week period 220,000 naturalization applications that had been backlogged for more than a year. Of the applications data-entered at NDEC, approximately 168,000 had not had their fingerprint cards previously stripped and sent to the FBI. Once those applications were data-entered, Los Angeles, for the first time, undertook a review of the fingerprint cards to ensure that they were suitable for comparison before sending them on to the FBI. For cards that were deemed unsuitable, Los Angeles employees contacted applicants and requested new fingerprint cards.
Although this review procedure was unprecedented and laudable given the problems INS had experienced with rejected cards, it failed to take account of when the applicant whose card was deemed unsuitable was scheduled for a naturalization interview. Many applicants received interview dates soon after the NDEC staff had data-entered their applications because Los Angeles INS had also opened a new interviewing office. In light of this accelerated interview scheduling, there was no guarantee that the applicant's fingerprint card had even been sent to the FBI for analysis.
When INS made the transition to Direct Mail, it failed to train its service center or contractor staff about fingerprint card processing, as noted above. The deficiencies in the Direct Mail plan and its execution are too many and too detailed to describe here, and are fully explored in our report. We note, however, that one service center director, upon realizing the degree of error in fingerprint card processing associated with the transition to Direct Mail, sent an e-mail message to Headquarters in July 1996 in which he advised, "we need to start over and design a process with integrity."
INS centralized its fingerprint card and rap sheet processing in one location beginning in June 1996 by establishing the Fingerprint Clearance Coordination Center (FCCC) in Lincoln, Nebraska. Contractor employees staffed the FCCC. The basic concept was that the FCCC would receive all fingerprint and bio-check responses from the FBI and distribute them to a particular contact in each district office. It would also immediately notify the district offices that a response had been received so that the district could delay interviews or ceremonies until the rap sheets arrived or the rejected cards were reprocessed. By centralizing this activity in one location, INS intended to more promptly alert adjudicators to applicants' criminal histories and to any fingerprint cards rejected by the FBI. In addition, INS hoped to increase accountability by designating specific employees in the Field to handle the fingerprint responses distributed by the FCCC.
As a result of understaffing, inadequate communication by Headquarters of the procedures and purpose of the FCCC, as well as the attempt to implement this process in the midst of the most hectic time of the busiest naturalization year ever, the FCCC actually increased confusion in the Field about fingerprint procedures.
First, INS Headquarters approved only half the number of personnel requested by the contracting managers and the INS staff officer in charge of implementing the FCCC. Headquarters had based its staffing estimates on an inaccurate assessment of the percentage of rejected cards and rap sheets in 1995 and failed to account for the increase in total cards that the CUSA initiative would create. Despite requests by contractors for at least double the personnel in order to handle the workload, a staffing increase was not approved until after CUSA. As a result, the FCCC was backlogged almost from the outset and notifications to field offices that a rap sheet or rejected card had been received were delayed, not accelerated. Timely processing, according to the contracting manager, was "near to impossible."
Even if the FCCC could have kept up with its workload, additional problems undermined its ability to succeed. The FCCC did not have the correct addresses for the "fingerprint specialists" in each district designated to handle the FCCC submissions. Those "fingerprint specialists" frequently had little or no experience with the fingerprint process and had little understanding of what their role entailed. The advantages of having a central location resubmit rejected fingerprint cards were undercut by the failure to communicate to the Field that cases for which new cards had been requested should be placed on "hold" and not adjudicated until 60 days after the new card was resubmitted.
In November 1996, when INS abolished the presumptive policy for fingerprint checks, FBI responses relating to naturalization applications were once again sent directly to the Field.9
Headquarters officials learned of bio-check processing problems in late 1995, after Records Division staff inadvertently discovered that certain bio-check responses were being destroyed. Shortly thereafter, INS implemented new procedures for processing bio-check responses, and eventually created the FCCC, as described above, to process both bio-check and fingerprint responses from the FBI.
However, INS was slow to restore integrity to its bio-check processing. When it was working to improve bio-check procedures in the spring of 1996, INS delayed the processing of more than 500,000 bio-check requests. Some of the delay was a function of the high volume of requests that were being generated at NDEC early in the CUSA program. Bio-check requests from NDEC applications were generated automatically by the computer system, and the massive input of information from more than 200,000 naturalization applications in a matter of weeks overloaded the data tapes and made them too cumbersome for the FBI to use without breaking down the data. Some of the delay was caused by INS' efforts to identify the reasons districts were submitting duplicate automated requests. Finally, some of the delay was purposeful, as INS requested that the FBI halt processing of bio-checks while it continued to work to solve the duplication problem. The net result was that the processing of bio-check requests generated electronically from October 1995 through April 1996, affecting more than 500,000 applicants, was not resumed until July 1996. As a result, many CUSA applicants did not have bio-checks completed before they naturalized.
In August 1996, prompted by media reports concerning the arrival in field offices of rap sheets pertaining to applicants who had already naturalized, INS Headquarters asked the Field to report on the extent of the problems they were having with late-arriving rap sheets. By August 19, 1996, Headquarters had already heard from the five Key City Districts that more than 100 cases were under review for possible denaturalization because of an applicant's failure to reveal criminal history of which INS had learned upon later receipt of the applicant's rap sheet or because post-naturalization review of a rap sheet uncovered a potentially disqualifying criminal history report (10 cases in Miami, 40 in New York, 61 in Los Angeles).10 In addition, INS Headquarters learned through the reports from the Field that rap sheets in more than 750 other cases still needed review.
The next month, the Subcommittee held its first hearings concerning CUSA and focused in large part on allegations of improper criminal history checking procedures. Among the INS employees who appeared as witnesses, two in particular-one from the Chicago District and the other from the Los Angeles District-testified about the naturalization of applicants whose disqualifying criminal history reports had not been reviewed. The INS response to these allegations was offered by Associate Commissioner Crocetti.
Despite the information INS Headquarters had gathered in August, Crocetti told the Subcommittee on September 24, 1996, that INS was aware of only 60 confirmed cases that would require denaturalization because of a disqualifying criminal history that had been reviewed after the applicant had naturalized. The testimony was misleading because by then INS Headquarters had learned from the Field that at least 850 cases were then under review. Nor was the apparent understatement explained by Crocetti's qualification that only 60 cases had been "confirmed," because at that time INS had not "confirmed" that any case would warrant denaturalization proceedings (Crocetti himself later told Congress that the number he had offered had been an estimate). While Crocetti's testimony, like other testimony by INS officials discussed in our report, was couched in concrete numbers and confident assurances, it was in fact based on little review of the information then available. Subsequent events further undermined the reliability of INS' early reporting about the consequences of CUSA's poor fingerprint processing procedures.
Immediately after the September hearings, a review team from INS Headquarters went to Chicago to learn more about the allegations that had been made by Chicago employees who testified before the Subcommittee. The team learned of more than 1,300 rap sheets pertaining to CUSA applicants that had not been reviewed before the applicants were naturalized. Of those, local officials had determined that approximately 300 would require further review to determine if denaturalization proceedings were warranted.
Work in other districts also revealed large numbers of unreviewed CUSA rap sheets. More than 1,000 had been identified by the Los Angeles District, 69 of which revealed criminal convictions that should have rendered the applicants ineligible for naturalization.11 This information was sent to INS Headquarters by e-mail on October 16, 1996. In addition, the Miami District identified more than 900 rap sheets that had either been received after naturalization or had been received in time but not reviewed.
In the meantime, INS Headquarters undertook a survey of the Field concerning late-arriving criminal history reports. The survey was limited to records processed by the FCCC (that is, received in the Field after July 1, 1996), and asked the Field to specify whether the late-arriving record reflected a criminal history that should have disqualified the applicant. Through a combination of bad reporting by the Field and bad compilation at INS Headquarters, the results of this survey listed only 415 late-arriving records nationwide, of which only 69 warranted further review for possible denaturalization proceedings.
The limitation of the survey to the period since the FCCC opened did not explain the discrepancy. In fact, in most districts the limitation to records processed by the FCCC had not been noticed. Some offices-like Los Angeles-incorrectly reported as "late-arriving" only those records that were both "late-arriving" and "disqualifying," or 69. Others-like Miami-accurately reported the number of late-arriving rap sheets since July 1 (116), but the report was not accurately recorded at Headquarters. Despite ample information that INS should not assure Congress about the integrity of its criminal history checking procedures, Commissioner Meissner wrote to then Subcommittee Chairman William Zeliff on October 15, 1996, and advised him that INS field offices had reported "the receipt of 415 late hits" out of 30,422 total rap sheets received since the FCCC became operational. She promised that INS would provide the Subcommittee with the results of its review of the 415 cases, but advised that "most of these cases" would not require denaturalization.
The Commissioner's letter was accurate in that 415 was the total number of late-arriving rap sheets that Headquarters staff had compiled based on reports from the Field. Her letter to Congress also specifically pointed out that this number was not the total number of unreviewed CUSA rap sheets, but rather was only the number for those "hits" that arrived post-naturalization since implementation of the FCCC. However, the Field had made numerous mistakes in compiling the survey responses, many of which should have been obvious to INS Headquarters officials who reviewed the incoming reports. Furthermore, given what Headquarters then knew about failures to review applicant rap sheets over the course of the entire CUSA program, Commissioner Meissner's letter omitted a fact that would have been responsive to congressional inquiries: that thousands of applicant rap sheets had sat unexamined in INS offices throughout the Key City Districts during CUSA, and INS had no method for determining whether they had arrived late or had simply been ignored.
Funds from INS' Examinations Fee Account, the account into which application fees are deposited, were used to help pay for CUSA. To access those funds, INS prepared "reprogramming" requests that were reviewed by the Attorney General and the Office of Management and Budget before being forwarded to House and Senate appropriations committees for concurrence. Although reprogramming requests technically require congressional notification and not approval, Commissioner Meissner and her staff regarded approved reprogramming requests as the equivalent of a statutory mandate concerning how such funds were to be spent.
INS submitted two reprogramming requests to Congress that were directly related to CUSA. Congress concurred in the first in June 1995, and the second in January 1996. Both requests sought funds to reduce INS backlogs in pending naturalization applications and in pending applications for adjustment of status-the method by which an applicant becomes a resident of the United States. The adjustment of status backlog had grown substantially after the passage in 1994 of a new adjustment of status provision, Section 245(i), of the Immigration and Nationality Act.
Section 245(i) expanded eligibility for adjustment of status to include aliens not in legal status at the time of application. The change meant that individuals seeking permanent residency status who previously had to leave the United States to apply for an immigrant visa at the U.S. embassy in their home country now could pay a penalty fee and apply directly to INS without leaving the United States. The law was passed as a temporary measure, and specified that applications submitted pursuant to this provision had to be adjudicated no later than October 1, 1997.
Congress made plain in response to both reprogramming requests that it expected INS to reduce its adjustment of status backlog. In response to the first reprogramming request, the House Appropriations Committee told INS that it had not devoted sufficient attention, in its request, to the processing of adjustment of status cases and directed INS to redirect funds it had planned to use on the promotion of naturalization to adjustment of status processing instead. With respect to the second reprogramming request, in January 1996 the House Appropriations Committee notified INS that it concurred, with the understanding that the funds would be used to eliminate the backlogs in both naturalization and in adjustments of status by the summer of 1996. Currency in adjustment of status processing was defined as four months' processing time from receipt of application to adjustment of status.
We found that INS Headquarters made few efforts to comply with its obligation to become current in adjustment of status processing during fiscal year 1996. We found that staff in the Key City Districts had little, if any, understanding of the reprogramming commitments INS made to the Congress. They uniformly understood that currency in naturalizations was a higher priority of INS Headquarters than work on adjustments of status. As a result, reprogramming funds made available to the Key City Districts were allocated primarily to naturalization processing, and adjustment of status processing suffered as a result of this disproportionate emphasis on naturalization.
By the end of fiscal year 1996, INS had achieved currency in naturalization applications, but failed to meet its 4-month adjustment processing goal by a huge margin. In fact, during the year the estimated time between filing and actual adjustment of status had increased to well over a year in the Western Region and to almost eight months in the Eastern Region. Overall, processing times for adjustment cases in the five CUSA Key City Districts were 28 percent higher than in INS as a whole. Pending caseloads in every Key City District were significantly larger than they had been at the end of fiscal year 1995. Although INS completed 421,000 adjustment of status cases nationwide in FY 1996almost twice the number it completed the year beforefewer than half of these applications were completed in the Key City Districts.
Headquarters officials told the OIG that it had not been their intention to allow adjustments of status processing to suffer because of the increased emphasis on naturalization. At the same time, they acknowledged that INS had done little to comply with the commitments made to Congress when INS obtained the reprogramming funds. In interviews with the OIG, most INS Headquarters managers said they were aware that INS had made a commitment to Congress to become current in processing both naturalization and adjustment applications, but none claimed that INS had ever planned a way of meeting the adjustment goals. None claimed that this commitment was communicated to or enforced in the Field. Deputy Commissioner Chris Sale told the OIG that, while INS made no deliberate decision to abandon the adjustments goals, the public focus on the CUSA project and the logistical attention necessary to get the five Key City Districts up and running gave the citizenship agenda preeminence. Headquarters relied on the Field to maintain an acceptable "threshold" of adjustments processing, but never specified what level would be acceptable or unacceptable. Although Headquarters occasionally reminded the Field not to neglect adjustments, District managers told the OIG that they perceived these messages as rhetorical reminders that carried little weight.
Unlike the naturalization effort, INS had no priorities, no plan, and no "project coordinator" for adjustments of status. INS did not open new adjustment-only offices, nor were adjustment interviews conducted in the new CUSA offices. Headquarters received frequent staffing and workload reports from the Field, and admittedly was aware of how the new employees hired with reprogramming funds were being used. Therefore, we found that INS failed to honor the agreements underpinning congressional approval of its reprogramming requests in 1995 and 1996 as a result of pursuing the production goals of the CUSA program.
Among the allegations we investigated were those concerning retaliation by INS management against INS employees who testified before congressional committees or otherwise provided information that was critical of the CUSA program. Six employees claimed that they had suffered adverse personnel actions in retaliation for their cooperation with congressional inquiries: three from the Los Angeles District, two from Chicago, and one from Dallas. The OIG conducted complete investigations in four of the six cases; the other two were primarily investigated by the Office of Special Counsel (OSC), which also has jurisdiction over such complaints. Of the six cases, one employee's allegations were sustained.
Because of Privacy Act concerns, our chapter concerning the investigation of these allegations of retaliation has been redacted. A complete copy of that section of the report is being provided to Congress under separate cover. To summarize our findings here, we note that INS did not generally retaliate against employees who cooperated with congressional inquiries. In the one instance where we found evidence supporting the claim of retaliation, there were other motives apart from the employee's role as a CUSA whistleblower that prompted the actions by INS. Although the existence of other motives does not exonerate INS for prohibited personnel practices, it does tend to indicate that agency officials did not, in general, take retaliatory action against the witnesses who brought to light many of CUSA's weaknesses.
In the wake of CUSA, INS has made some significant improvements and has asserted that it is less tolerant of error in the naturalization process. INS has made obvious improvements in its procedures for ensuring that applicants' fingerprints are checked by the FBI and that the results of those checks are available to adjudicators. It has also markedly improved its procedures for ordering and transferring applicant files so that they, too, are available at interview. Finally, it has implemented standardized checklists and other processing forms that allow it to monitor whether cases are adjudicated in a manner consistent with these new procedures. INS' "Naturalization Quality Procedures," first published in November 1996 and since revised several times, enhance the integrity of naturalization processing.
However, INS has not made progress toward developing and implementing adjudicative standards, including the standards for English testing and the evaluation of an applicant's "good moral character." There has been little progress toward ensuring that adjudicators, once they have the requisite tools (like the results of criminal history checks or the applicant's file), know how to use them. In short, INS has standardized its processing procedures, but has not improved the substantive aspect of the evaluation of an applicant's eligibility for naturalization. Our recommendations thus focus, overall, on steps to be taken to improve the quality of the naturalization adjudication itself.
Of greatest concern is that INS had not taken steps to identify and promulgate standards that INS itself had recognized as crucial but lacking even before CUSA. The absence of these standards influenced CUSA adjudications and continues to influence INS' current naturalization work. Although Commissioner Meissner announced as early as March 1997 that such standards would be published in an Adjudicator's Manual that would be available throughout the Field, INS has not drafted any portion of that Manual concerning the adjudication of naturalization applications. DAOs have no better guidance today than they did during CUSA concerning the appropriate evaluation of "good moral character" in the adjudication of a naturalization application. Furthermore, INS has still not developed a standard for English testing.
We recommend that INS develop these needed standards immediately. If those standards will be included in the anticipated "Adjudicator's Manual," then we recommend that INS prioritize the completion, publication, and dissemination of the Manual.
Concerning the evaluation of "good moral character," INS must provide guidance concerning:
In regard to the testing of English and Civics, INS is currently searching for a contractor to develop its testing standards and a contractor to administer its "reengineered" testing program.12 INS plans to create its own testing sites from which test results will be electronically transmitted to INS offices. That "reengineered" program is dependent on many variables, including the development of appropriate technology and the identification of a qualified contractor. The ambitious plan may be many months, if not years, from full execution, and in the meantime INS continues to adjudicate applications without uniform standards concerning these basic requirements of naturalization. Accordingly, we recommend that, with or without the services of an outside consultant, INS immediately develop a standard by which to evaluate an applicant's "ability to read, write and speak words in ordinary usage in the English language" as required by current law.
Other guidance that was absent during CUSA that undermined the quality of naturalization adjudications concerned how to evaluate an applicant's previous immigration history and in what circumstances that history might be relevant to the naturalization adjudication. Although INS has asserted that complete file revieweven if the file contains "confidential" information submitted pursuant to provisions of the Immigration Reform and Control Act of 1996is appropriate in conjunction with a naturalization adjudication, it is not clear that this emphasis has reached the rank-and-file. We found that misunderstandings about confidential IRCA information lingers to this day. Also, although INS, through NQP, has emphasized the importance of having the applicant's permanent file at the naturalization interview, it still has not offered concrete guidance concerning how to use the information in that file to inform the naturalization adjudication. Accordingly, we recommend that INS provide guidance concerning how to detect previous fraud by applicants for naturalization, particularly when the file contains information considered "confidential" under IRCA, and how to confirm or dispel during the naturalization interview such suspicions of fraud.
To improve its efforts at detecting fraud in the naturalization process, we recommend that INS improve the coordination between its investigative and adjudicative efforts so that information concerning fraudulent activity detected by INS investigators is timely available to and appropriately acted on by those who may be asked to adjudicate related benefits applications. We also recommend that INS more aggressively use and administer its computerized "flagging" procedures to alert adjudicators to the presence of cases in which fraud is suspected.
INS has made the greatest strides in the area of criminal history checking procedures, the area that was, as Commissioner Meissner called it in her interview with the OIG, CUSA's "fatal flaw." Definitive responses to fingerprint checks are now required for every naturalization applicant. However, INS had not completely eradicated the belief that thorough fingerprint processing in every case is worth the cost. Recently, citing the low risk and burden of other options, INS proposed a policy that permits the adjudication of a naturalization application without further criminal history checks of an applicant whose fingerprints have twice been rejected by the FBI as unclassifiable. We recommend that INS obtain classifiable prints from every naturalization applicant (excluding those exempted by age) unless it certifies that the applicant cannot-because of a physical condition-be fingerprinted in a manner that yields classifiable prints.13
We also note that INS continues to work with a "presumptive policy" in regard to applicant bio-checks. We recommend that INS confer with appropriating authorities and officials from other agencies to evaluate the risks associated with such an approach and the viability of alternative strategies.
Finally, among our most troubling findings were those concerning the reliability of the information INS provided in response to congressional inquiries. We recommend that INS ensure the accuracy of the information it provides to Congress, whether in the form of written responses or testimony. We also recommend that INS be vigilant to follow through on the agreements it enters into with Congress, whether the agreement is to pay particular concern to some aspect of its adjudication program-as was the case with its failed agreement to take appropriate action to respond to evidence of fraud during CUSA-or an understanding concerning appropriations. INS must continue to enhance the integrity not only of its naturalization processing but also of its relationship with Congress.
Robert L. Ashbaugh
Acting Inspector General