DCCUMENT RESUME 03451 - tA2E33751] Internal Revenue Service's Contrcs over the Use of Confidential Infczmants: Recent Improvements Not Adequate. GGD-77-46; B-147762. September 1, 1977. 54 pp. + 5 appendices (53 pp.). Repor; to Chairman, Joint Committee on Taxation; Vice Chairman; by Elmer E. Staats, Comptrcllez General. Issue Area: Law Enfor-ement dnd Crime Prevention (500); Tax Administraticn: Intelligence Activities to Insure Tax Law Compliance (2703). Contact: General Government Div. Budget Function: General Government: Central Fiscal Operations (803); Law Enforcement and Justice: Federal Law Enforcement and Prosecuticn ()51). Organization Concerned: Internal Revenue Service. Congressional Relevance: House Ccmmittee on Ways and Means; Senate Committee on Finance! Joint Committee on Taxation. Authority: 26 U.S.C. 7623. Internal Revenue Code of 1954, sec. 7623. Primarily because of inadequate management attention, the Internal Revenue Service (IRS) had not done all it could to get the most benefits with the least :isk in dealing with confidential informants. Findings/Conclusions: Informants were sometimes used in ill-defined and overly broad intelligence gathering efforts, procedures for evaluating their information were inadequate, and their use was not systematically reviewed by management. Since 1975, the IRS has strengthened its controls. but it could do more. Management needs to pay more attention to the fact that an informant is used only after the potential benefits and risk. have been properly assessed. Recommendations: The Commissioner of Internal Revenue should require that higher level management officials--preferably regional ccmrissicners--authorize the use of any informant who will be gathering information at the Service's request or encouragement after a determination that there is: reasonable cause to believe a tax law has been violated, no practical alternative for obtaining essential information, and. a specific limitation cn the time and scope of the informant's activities. The Commissioner should also require that requests to rse each such informant show, among other things, why the informant is needed, how be was developed and determined reliable, and how he will be used. In ord:!r to spur proper implementation of the guidelines and adequate management attention to informant activities, the Commissioner should review those activities at least annually. (Author/SC) REPORT TO THE JOINT COMMITTEE ON TAXATION CONGRESS OF THE UNITED STATES uo-~I BY THE COMJPTROLLER GENERA '. : ':it OF THE UNITED STATES Internal Revenue Service's Controls Over The Use Of Confidential Informants: Recent Improvements Not Adequate Department of the Treasury Primarl'y because of inadequate man igement attention, the Internal Revenue Service had not done all it could to get the most benefits with the least risk in dealing with confidential informants. Informants were sometimes used in ill-defined and overly broad intelligence gathering efforts, procedures for evaluating their information wt-e inadequate, and theii use was not systematically reviewed by management. Since 1975, the Internal Revenue Service has strengthened its controls but it could do more. Management needs to pay more at- tention to the fact that an informant is used only after the potential benefits and risk? have been properly assessed. GGD-77-46 SEPTEMBER 1, 1977 COMPTROLLER GENERAL OF THE UNITED STATES WASHINGTON. D.C. 20548 B-137762 To the Chairman and Vice Chairman Joint Committee on Taxation Congress of the United States In response to your Committee's request, we are addressing the Internal Revenue Service's management cont-ols ier the use of confidential informants. The Service agrees with most of our proposals for improving these controls; it does not agree, however, that regional commissioners should be responsible for approving the use of an informant. Inatead, the Service intends to delegate this authority to the district director who is the management official responsible for all Service activities in a district. We would consider, on a trial basis, delegating this responsibility to the district director as an acceptable alternative. As arranged with your Committee, we plan no fur- ther distribution until 30 days from the date of the report unless you publicly announce its contents earlier. At that time, we will send copies to interested parties and make copies available to others upon request. ptroller General of the United States COMPTROLLER GENERAL'S REPORT INTERNAL REVENUE SERVICE'S TO THE JOINT COMMITTEE ON CONTROLS OVER THE USE OF TAXATION CONFIDENTIAL INFORMANTS: CONGRESS OF THE UNITED STATES RECENT IMPROVEMENTS NOT ADEQUATE Department of the Treasury DIGEST The Intelligence Division of the Internal Revenue Service is responsible for enforcing the criminal provisions of the tax laws. Confidential informants have contributed to this mission. However, their use can endanger Service employees, jeopardize tax cases, and violate taxpayers' rights. Such serious risks reveal a need for strong management controls. In the past, such controls were inadequate. While the Service has taken steps recently to strengthen these controls, GAO recommends that more be done. No attempt is made in this report to deter- mine how many intelligence gathering efforts were gooc or bad in terms of results but GAO did look at results to see if confiden- tial informants are valuable to the Service. They are, but their use also entails risks. WHY CONTROLS ARE NEEDED The Intelligence Division's informant activity is small in terms of dollars--about $800,000 was paid out of operating expenditures totaling $261 million to confidential informants in fiscal years 1973 through 1975. It is not small, however, when one considers that those payments were made to about 2,000 informants while thousands more provided information for free. (See p. 3.) In dealing with confidential informants, the Internal Revenue Service is faced with trying to get the most benefits with the least risk-- a difficult task. If an informant is bent on violating the law, giving false information, or the like, there is little, if anything, the Service can do to prevent it. But with good management controls, Tuw Samt. Upon removal, thereort GGD-77-46 covr dte should be noted heon. i the Service can lessen the chances of such things happening and, if they do happen, can take quick corrective action, such as terminating an informant's services. At the same time, these controls can help the Service make the most of the benefits it might derive from an informant's services. (See p. 11.) HOW PAST CONTROLS WERE INADEQUATE It is essential that intelligence gathering efforts by informants have clear objectives and precise plans of action. Many of the intelligence gathering operations Gp_ reviewed were begun on the basis of assump- tions or allegations that an illegal acti- vity was being conducted in a geographical area or by a large group of people rather than sn specific allegations of tax law viiolations by specific individudls. CO:her intelligence gathering programs were started primarily beceuse informants were available and willing, if paid. They were sent out not for well-defined purposes but simply to gather whatever information they could on any individuals they might identify as pos- sibly being invnlved in illegal activities. (See p. 12.) Informants are valuable only if they provide information which the Service determines to be useful for further intelligence gathering; a criminal investigation, indict- ment, or prosecution; or tax assessments and collections and which otherwise would not be available to the agency. This requires timely, systematic analysis and evaluation to determine the information's tax implications. (See p. 26.) Internal Revenue Service procedures generally provided that an informant be paid only after his information was evaluated. In theory this procedure provided for systematic, timely eval- uation of information. In reality, it proved inadequate. ii If an informant approached the Service voluntarily with information that he wanted to exchange for money, the procedure worked well--the Service reviewed the information, evaluated its worth, and paid for it. However, if an informant were sent out to gather information, the procedure became inoperable because these informants were paid for their services rather than for information. (See p. 22.) In gathering intelligence about pezsons and groups who derive their income from illegal activities, the Service cannot avoid getting into sources of income such as gambling, loan sharking, political corruption, and narcotics trafficking--crimes which are out- side of its enforcement authority. To do otherwise would be remiss. It would be tantamount to granting immunity from com- pliance with the tax laws to a segment of the population having a high probability of willful nonccmpliance. (See p. 20.) To obtain information on illegal dealings, the Service used informants to infiltrate those activities, identify individuals involved, and gather information on them. The information they gathered frequently was general in nature, was not always clearly tax related, and contained names of numerous individuals. Even so, the infor- mation might still have been useful to the Internal Revenue Service and should have been systematically evaluated, but was not. (See pp. 24 and 26.) Finally, the Intelligence Division's use of confidential informants was not systematically reviewed at any level of management. Systematic review is necessary to insure that the use of informants is in line with the Service's mission and consistent with established procedures. (See p. 28.) Te-ar ~hn~iii WHAT THE INTERNAL REVENUE SERVICE HAS DONE TO IMPROVE ITS CONTROLS After adverse Publicity and controversy in 1975 over its use of confidential informants, the Internal Revenue Service took steps to strengthen its controls. New guidelines, issued in 1975 and 1977 -- require a written authorization before any Service employee gathers intelligence; -- require the chief of intelligence in each district office to approve the use of each paid informant and each unpaid informant who is encouraged to provide information and to maintain a record on each such informant used; -- spell out what informants can and cannot do in working for the Service, and provide instructions as to what Service employees should do when they become aware of illegal informant activity; -- specifically require managers to review and report on informant activities; and -- provide for the timely evaluation and processing of information received from informants. (See pp. 40 and 45.) WHAT MORE SHOULD BE DONE The Service's new guidelines do not provide adequately for management assessment of potential benefits and risks before an informant's use is authorized. rJnder the guidelines, primary responsibility for authorizing the use of an informant rests with the chief of intelligence. District, regional, and national office managers need not be involved unless confidential funds are to be expended. Involvement of these other managers is an exercise of fiscal responsibility. They may ask Questions about why and how an informant is to be used but their primary purpose is to iv authorize expenditures, not the use of an informant. (See p. 48.) Informants who gather information at the Service's request or encouragement present special risks. Therefore controls over them should be greater than controls over infor- mants who furnish information at their own initiative. Failure to clearly distinguish between these two types of informants results in too much control in some instances and not enough in others. Under the new guidelines, for example, a chief of intelligence can approve the use of an unpaid informant who is going to be gathering information at the Service's request or encouragement. Yet, under the procedures on confidential expenditures, the chief must obtain the district direc- tor's approval to pay even a few dollars to a person who voluntarily walks into all Internal Revenue Service office with valu- able tax'-related information. While controls over money are important and should be properly emphasized, control over informant activities should be based primarily on how an informant will be used--not on the amount of money to he spent--because the risks associated with using an informant do not vary depending on the dollars to be paid. (See p. 48.) Revised guidelines alone are not the answer. Past guidelines often were not complied with so there is no assurance that the revised guidelines will be unless management pays more attention to these activities than in the past. Although Intelligence Division officials directed and encouraged their people to use confidential informants, nei- ther they nor other Service managers were sufficiently concerned with how those infor- mants were being used. (See pp. 28 and 49.) Intelligence officials at the national, regional, and district offices tended to ignore or condone potential problems and weaknesses by not requiring compliance Tear Sheet v with existing guidelines and by not respond- ing to problems as they arose. Service managers outside the intelligence function appeared to have a "hands off" philosophy regarding informant activities. This attitude that intelligence personnel should be left to manage their own affairs appeared pervasive and generally accepted. (See p. 50.) This lack of management attention at all levels was the underlying cause of problems discussed in this report. It may have arisen because management failed to fully appreciate the problems that can arise when dealing with informants. Whatever the reason, it must change. GAO RECOMMENDATIONS Accordingly, GAO recommends that the Commis- sioner of Internal Revenue require that higher level management officials--preferably regional commissioners--authorize the use of any informant who will be gathering information at the Service's request or encouragement after a determination that there is -- reasonable cause to believe a tax law has been violated, --no practical alternative for obtaining essential information, and --a specific limitation on the time and scope of the informant's activities. The Commissioner should also require that requests to use each such informant show among other things, why the informant is needed, how he was developed and determined reliable, and how he will be used. To better spur proper implementation of the guidelines and adequate management attention to informant activities, the Commissioner should review those activities at least annually. One way this can be accomplished vi is through an inuependent third Party like the Service's Internal Audit Division. (See p. 50.) INTERNAL REVENUE SERVICE COMMENTS The Service did not agree that the use of informants should be approved by reqicnal commissioners in order to remove this decision from the intelligence function. It believes district directors should have this authority. Although GAO believes that regional commissioners would be in a better position to authorize the use of an informant it is willing to accept the Ser- vice's position, on a trial basis. The Service said that it does distinquish between informants who provide information on their own initiative and those who pro- vide information at the Service's request or encouragement and that its guidelines already call for information on why an informant is needed, how he was developed and determined reliable, and how he will be used. While the guidelines do contain bits and pieces of GAO's recommendation, the', fall short of providing the total control GPO is seeking. The Service agreed with the concept of GAO's last recommendation but noted that Internal Audit has always been involved in reviewing informant activities. GAO's recommendation, however, is not directed toward Internal Audit; it is directed toward the Commissioner. GAO's recommen- dation was intended to provide for more direct oversight of informant activities by the use of periodic reports to the Commissioner. (See p. 51.) vii Contents DIGEST CHAPTER 1 INTRODUCTION 1 IRS' intelligence mission and the role of confidential informants 1 History of IRS' involvement with confidential informants 3 Review objectives and scope 6 2 THE RISKS ASSOCIATED WITH CONFIDENTIAL INFORMANTS DICTATE THE NEED FOR EFFECTIVE MANAGEMENT CONTROL 8 How valuable are informants? 8 Conclusions 10 3 PREVIOUS MANAGEMENT CONTROLS WERE INADEQUATE TO MAXIMIZE BENEFITS AND MINIMIZE RISKS 12 Confidential informants used in ill-defined and overly broad intelligence gathering efforts 12 Conclus.ions 19 Procedures for accumulating, evaluating, and processing information provided by informants were inadequate 21 Conclusions 26 Management emphasized expanding-- not evaluating--informant activities 28 Conclusions 38 4 MANAGEMENT CONTROLS HAVE BEEN IMPROVED BUT MORE NEEDS TO BE DONE 40 Actions taken to correct weakne sss 40 Conclusions 48 RecommEndations 50 IRS con.vents and our evaluation 51 APPENDIX I Letter dated July 27, 1977, from the Commissioner of Internal Revenue 55 II Operation Leprechaun 60 III IRS guidelines on informants, confidential expenditures, and information gathering 80 IV Congressional hearings on IRS intelligence activities after publication of allegations about Operation Leprechaun 106 V Principal officials responsible for administering activities discussed in this report 107 ABBREVIATIONS GAO General Accounting Office IRS Internal Revenue Service CHAPTER 1 INTRODUCTION At tile request of the Joint Committee on Taxation, we reviewed the use of confidential informants by the Internal Revenue Service's (IRS') Intelligence Division. IRS' INTELLIGENCE MISSION AND THE ROLE OF CONFIDENTIAL INFORMANTS IRS' overall mission is to encourage and achieve the highest possible degree of voluntary compliance with the tax laws and regulations and to conduct itself so as to warrant the highest degree of public confidence in its integrity and efficiency. To successfully carry out this mission, IRS must seek out and prosecute persons who violate those laws. The most frequent violations prosecuted as crimes are willful (1) attempt. to evade tax, (2) failure to collect or pay over tax, and (3) failure to file returns. IRS' Intelligence Division is responsible for enforcing the criminal provisions of the tax laws. The Director of that division at the national office reports tc the Assistant Com- missioner for Compliance, who in turn reports to the Commis- sioner. Like IRS in general, the intelligence organization is highly decentralized among 7 regions and 58 districts. At the regional level, the assistant regional commis- sioner for intelligence acts as the principal assistant to the regional commissioner in planning, coordinating, and evaluating those intelligence activities under the juris- diction of the regional commissioner to insure that policies and programs are properly executed. At the district level, the chief of intelligence, under the direction of the district director, enforces the criminal statutes by developing information on alleged violations, evaluating allegations and indications of such violations to determine whether investigations should be undertaken, invest- igating suspected criminal violations, recommending prosecu- tions, and measuring the effectiveness of the investigation and prosecution processes. Thus, the responsibility for day- to-(ly intelligence operations rests with the chief of intelli- gence and the group managers and special agents 1/ assigned to him. 1/ Special agents are the Intelligence Division employees who actually gather intelligence on and investigate charges of criminal violations of the tax law. 1 The use of confidential informants is just one of many means that the Intelligence Division employs to identify investigate suspected criminal violators of the tax laws. and Special agents also gather information and coniduct investi- gations through surveillance of suspected violators, inter- views of witnesses and potential defendants, and reviews of public documents, third party records, and taxpayer records, if available. Special agents sometimes act in an undercover capacity. Who are informants? According to TRS, informants are persons who furnish information on their own initiative or as a result of being encouraged to do so by a special agent or other IRS employee. Such persons include those who furnish leads or bits of in- formation as well as those who submit detailed information regarding alleged violations. A confidential informant one who furnishes information on the expectation that hisis identity will be held in strict confidence. Section 7623 of the Internal Revenue Code of 1954 (26 U.S.C 7623) authorizes IRS to make payments which it determines are "necessary for detecting and bringing to trial and punishment" persons who violate tax laws. This provision was first enacted, in substantially the same form, as Section 7 of the act of March 2, 1867. Based on this general provision, it has been repeatedly affirmed through various decisions and actions by the courts, the Congress, and the Comptroller General of the United States that IRS has authority to make payments from appropriated funds to support and maintain informants tion which may lead to the prosecution of who gather informa- tax law violators. According to IRS, its informants usually fall into of three general groups: average citizens, tax violatorsone their associates, and disturbed persons. They are motivated and by such things as money, egotism, friendship, revenge, fear. and In general, informants are either (1) referred to IRS by other law enforcement agencies, (2) identified by special agents and requested to gather information because of their unique position, or (3) known to IRS intelligence personnel through prior intelligence gathering or investigative efforts. The Intelligence Division pays its confidential from imprest funds 1/ established for that purpose andinformants classifies 1/ Fixed sums of cash (currency or Government checks) advanced to duly authorized cashiers for specified purposes. 2 those payments as confidential expenditures. These expen- ditures include payments made to informants for specific infor- mation and for expenses and services which are not directly related to specific information. Confidential expenditures also include expenses incurred by special agents while con- ducting surveillance or operating in an undercover capacity. Intelligence Division expenditures for fiscal years 1973 through 1975 totaled $261.2 million--only a small portion of which went to confidential informants. Information IRS accum- ulated in March 1975 shows payments of about $800,000 to about 2,000 confidential informants during those 3 years. Not all confidential informants are paid, however. In reporting on its review of informant activities in 22 dis- tricts, IRS' Internal Audit Division showed the extent to which IRS used unpaid informants when it noted that: "The 22 districts prepared lists that identified 3,759 unpaid informants that had been utilized to provide information on a recurring basis during the period July 1, 1971 through March 31, 1975." IRS recognizes that many investigations cannot be success- fully completed without using informants and purchasing evidence particularly if the alleaed violator is engaged in an illegal activity. IRS, like other law enforcement agencies, considers the use of informants necessary in carrying out its enforcement mission because they can furnish information which otherwise might not oe available. HISTORY OF IRS' INVOLVEMENT WITH CONFIDENTIAL INFORMANTS IRS does not know, and we were unable to determine, when it first used a confidential informant but one may have been used as far back as the 1930s during IRS' investigation of Al Capone. Our research of available literature, review of intelligence gathering efforts, and interviews with intelli- gence personnel indicate that IRS' confidential informants are mostly involved in gathering information on organized crime. A history of IRS' involvement with organized crime, therefore, should provide a fairly accurate history of its involvement with confidential informants. Although !RS had been successful in jailing some infamous racketeers for income tax evasion, it did not have a well- organized racketeer investigation program until 1951, at which time it initiated its Tax Fraud Drive on Racketeers. This 3 drive, which was intended to force compliance with the Federal income tax laws by all members of the underworld, began after the Kefauver Committee 1/ had criticized IRS for failing to sufficiently enforce the tax laws against organized crime. From 1954 to 1957, the program against racketeers de- clined because, under the Eisenhower Administration, IRS was generally not permitted to request appropriations for addi- tional personnel to expand enforcement activities and because several Commissioners were not sympathetic to an emphasis on organized crime. In 1955, for example, the Commissioner of Internal Revenue, in a letter to the Attorney General, noted that IRS policy: "does not envision, in view of manpower limi- tations and the necessity for maintaining a balanced enforcement program, this Service engaging in a 'drive' against any particular segment of taxpayers." But the November 1957 Apalachin meeting of racketeers triggered a Justice Department drive on racketeers and increased IRS' interest. After President Kennedy's inauguration in 1961, the drive against organized crime became a more prominent Govern- ment-wide policy and all Government agencies were "enlisted and ordered" to joint in combating it. According to a former Director of the Intelligence Division, emphasis qiven this effort by IRS declined in 1965 and 1966 due to the departure of Robert Kennedy as Attorney General and hearinqs held by Senator Edward Long 2/ which disclosed instances of agents overstepping the bounds of constitutional restraint with respect to electronic surveillance. In November 1966, President Johnson voiced administration support of the organized crime program which led to the first Strike Force in January 1967. The objective of the Strike Force concept is to coordinate, through the Justice Department, the combined forces of Federal law enforcement agencies against the l/Special Committee of the United States Senate to Investigate Organized Crime in Interstate Commerce, established May 3, 1950. 2/Hearings before the Subcommittee on Administrative Practice and Procedure of the Committee on the Judiciary, United States Senate, 89th Cong., 1st Sess., pt. 3, pp. 1126-27, Hon. Edward Long, Chairman. 4 criminal element in our society. Support for this combined Federal effort was reemphasized in 1969 by President Nixon 1/. In 1971, President Nixon ordered all agencies to join in an effort to suppress narcotics trafficking--a facet of orga- nized crime. Because profit is the main motive for trafficking, and taxes are a consequence, the Treasury Department called on IRS to play a major role. The objective was to direct IRS' civil and criminal tax enforcement efforts against those who financed and profited from trafficking, as experience had shown such income was rarely reported for tax purposes. In June 1975, IRS discontinued its Narcotics Traffickers Program as a separate enforcemnent effort. The drive against traffickers was reemphasized in April 1976 when President Ford directed the Secretary of the Treasury and the Commissioner of IRS, in consultation with the Attorney General and the Admini- strator of the Drug Enforcement Administration, to develop a tax enforcement program aimed at major drug traffickers. IRS' recent reviews of confidential expenditures and informant activities In November 1974, Internal Audit began a review of the In- telligence Division's confidential imprest fund at the national office which involved inquiries into the payments to and the effectiveness and management of confidential informants. In February 1975, IRS initiated a joint Internal Audit/ Internal Security investigation of the intelligence gathering function in the Jacksonville, Florida, district office because of indications )f improper activities. A month later, an article in the Miami News alleged that IRS agents were paying informants to spy on the sex lives and drinking habits of prominent public officials as part of an intelligence gathering effort called "Operation Leprechaun"--a code name applied to certain intelligence operations in Miami. Three days later, on March 17, 1975, the Deputy Commissioner revoked IRS' guidelines on confidential expenditures and centralized approval of such expenditures at the national office pending a review and evaluation of those activities. The centralization remained in effect until December 10, 1975, when IRS issued new guidelines giving regional commissioners authority to approve confidential expenditures up to $2,000. 1/For a more complete discussion of the Federal Strike Force effort see our report entitled "War on Organized Crime Faltering--Federal Strike Forces Not Getting the Job Done" (GGD-77-17, March 17, 1977). 5 Because of these events and the results of Internal Audits' national office review (1) IRS expanded its joint Internal Audit/Internal Security investigation in the Jack- sonville district and (2) Internal Audit initiated a review of management and internal controls over information qather- ing, confidential funds, and informants in 22 other district offices--some in each of the 7 regions. Concurrently, IRS' Internal Security Division investigated alleged misconduct by Intelligence Division employees, including several who were in some way identified with Operation Leprechaun. REVIEW OBJECTIVES AND SCOPE Our review was directed toward evaluating the use of confidential informants by the Intelligence Division with attention to the -- role of confidential informants in IRS' intelliqence mission, --contributions made b- informants, -- procedures for initiating intelligence gathering efforts in which informants were used, -- procedures foL accumulating and processing informa- tion Provided by informants, and -- extent of management involvement in the use of informants. We reviewed the procedures and practices followed by the national office, two regional offices, and five district offices in those two regions for initiating and c&rrying out intelligence activities in which confidential informants had major roles 1/. The national office and the two regions we reviewed accounted for 33 percent of the 2,000 informants and 62 percent of the $800,000 paid them during fiscal years 1973, 1974, and 1975. As part of our review, we evaluated the work that IRS' Internal Audit Division had done since November 1974. IRS auditors bad done work at the national office and three of 1/Because infcrmant activities are highly sensitive, we are not identifying regions, districts, or intelligence gathering efforts by name except for Operation Leprechaun which is already a matter of public record. 6 the district offices we visited. Their work included determining the extent to which information was gathered on the personal habits of taxpayers, the adequacy of con- trols over confidential expenditures, and the propriety of methods used by IRS employees and confidential infor- mants to gather information. The one aspect of Internal Audit's work with which we differed was the classification of information gath- ered during Operation Leprechaun as sex related. (See p. 76). Because our overall evaluation showed that Internal Audit's work was adequate, however, we were able to limit the scope of our review by using their work to supplement ours. Although Internal Audit reviewed and reported on weaknesses in both accounting and management controls and although we also reviewed both types of con- trols to satisfy ourselves as to the adequacy of their work, we are limiting our report to a discussion of the more important management controls. The activities we reviewed were undertaken by IRS between July 1971 and March 1975. During that period written procedures were in effect which have since been discontinued or substantially revised. Since March 1975, IRS has reissued guidelines for practically all aspects of its intelligence gathering activities. We reviewed the new guidelines to determine if they adequately deal with the problems and weaknesses identified in our review and in IRS internal audits, studies, and investigations. Because of the publicity, interest, and contro- versv surrounding Operation Leprechaun, which involved exyensive use of informants, we reviewed that project in some depth and have summarized it in appendix II. 7 CHAPTER 2 THE RISKS ASSOCIATED WITH CONFIDENTIAL INFORMANTS DICTATE TIE MEED FOR EFFECTIVE MANAGEMENT CONTROL In discussing confidential informants, two factors must be considered--the benefits they can pros i.e and the problems they can cause. While these informants c- be vdiuable to IRS' intelligence mission, they also can . source of ad- verse publicity, criticism, and embarrass .t as vividly de- monstrated by Operation Leprechaun. More importantly, they can give false information, violate the law, encroach upon privileged communications, unlawfully inhibit free association of individuals and expression of ideas, ccmpromise the inves- tigation and prosecution of suspected tax violators, or attempt to further their own interests at IRS' expense. Such actions could endanger IRS employees, jeopardize tax cases, and violate taxpayers' rights. Such risks cannot be taken lightly, as indicated by the following excerpt from the April 1976 report by the Senate Select Committee to Study Governmental Operations with respect to Intelligence Activities (Book II, p. 183): "Although there are circumstances where these techniques [informants and others) if properly controlled, are legal and appropriate * * their very nature makes them a threat to the personal privacy and Constitutionally protected activities of both the targets and of persons who communicate with or associate with the targets." HOW VALUABLE ARE INFORMANTS? Intelligence Division personnel assigned to gather intelligence and conduct investigations can draw on a number of information sources outside IRS. These sources include other Federal agencies such as the Federal Bureau of Investi- gation, the Drug Enforcement Administration, the Bureau of Alcohol, Tobacco and Firearms, the Securities and Exchange Commission, and the Department of Labor; businesses including banks, security and commodity brokers, and other financial institutions; and state, county, and municipal agencies. IRS believes, however, that many criminal tax cases could not have been successfully completed except for the use of confi- dential informants who supplied information otherwise unavail- able, especially concerning taxpayers engaged in illegal activities. 8 It is difficult, however, and often misleading to measure the value of confidential informants. Sometimes it is fairly easy to identify the informant's contribution because his in- formation leads directly to a prosecution or an audit. Other times it is no_ so easy because the informant may be only a small, yet necessary, part of an intelligence gathering effort. An informant may provide information, for example, which by it- self means little but, when combined with information obtained from other sources, gives the special agent the evidence he needs to proceed with an investigation. The informant's contribution in such an instance is difficult to determine. We are unable, therefore, to assess in quantifiable terms the benefits derived from IRS' use of confidential in- formants. We do know, however, that in some intelligence gathering efforts confidential informants furnished information which was general in nature and of no apparent use to IRS while in other efforts informants provided information which was di- rectly tax related and useful int criminal prosecutions, criminal investigations, and assessments of additional tax. In one intelligence gathering effort, for example, IRS paid an informant to obtain information on organized gambling, loan sharking, fencing, and police corruption in a certain geo- graphical area. The informant was unsuccessful, however, and the district subsequently reduced the scope of the effort to focus on gambling. In all, this effort lasted over a year and the district paid about $1,520 to three informants. Yet, not enough speci- fic information was gathered to be useful to IRS. For the most part, the informants used in this intelligence gathering effort provided general information rather than specific tax informa- tion from which intelligence cases with good prosecution poten- tial could be developed. During another intelligence gathering effort, special agents purchased information from eight informants in areas where bootlegging, bookmaking, and lottery operations were .1.legedly taking place. The objective was to gather enough intelligence on individuals and others to warrant formal criminal investigations. Our review showed that the infor- mants provided -- information which resulted in an investigation of two taxpayers, -- leads on two taxpayers which were further developed by special agents before investigations were initiated, 9 -- information which was used by special aqents in two intelligence investigations that were already under- way, and -- information on one taxpayer which was referred to the Audit Division and used by that Division in auditing the taxpayer's tax return and recommending an additional tax assessment of $58,677. As a result of the investigations initiated or developed with information supplied by the informants, two tarpayers were convicted and two had been recommended to the Department of Justice for prosecution as of January 1977. Of the two cases involving information provided by informants after in- vestigations were underway, one resulted in the taxpayer's conviction and the other was dropped after the taxpayer's death. In a third intelligence gathering effort, informants contributed to 10 intelligence investigations. At the time of our review, 8 of the 10 investigations had been completed. Of these eight, five had resulted in recommendations of nonprosecution by the investigating agent, one was recommended for prosecution but declined by IRS' Regional Counsel, and two were forwarded to the Justice Department for prosecution. Of the latter two, one resulted in the taxpayer's conviction and the other was awaiting prosecution. In addition to their contributions to intelligence cases, informants contributed to audits of these same 10 taxpayers and 4 others, which resulted in recommended assessments of additional tax and penalties totaling about $595,000. Of this amount, $265 had been collected and about $543,000 was believed by Audit Division personnel to have good collection potential, should the recommended assessments stand. It should be noted that additional tax and penalties recommended by Audit Division examiners are subject to a variety of reviews and taxpayer appeals. Any of these reviews or appeals could result in actual assessments less than the amounts recommended. CONCLUSIONS Because informants' contributions are difficult to quantify, we have avoided getting into a discussion of how many intelligence gathering efforts were good or bad in terms of results. But, we did look at results to see if informants were valuable to IRS and we concluded that they were. 10 The use of confidential informants entails not only benefits but also significant risks. In dealing with confi- .- "'ial informants, therefore, IRS is faced with trying to r 'nize the benefits while minimizing the risks--a difficult If an informant is bent on violating the law, giving false information, or the like, there is little, if anything, IRS can do to prevent it. With sound managment controls, how- ever, IRS can minimize the chances of such things happening and, if they do happen, can take quick corrective action, such as terminating the informant's services. At the same time, these controls can help IRS maximize the benefits it might derive from the informant's services. Such controls should include (1) sound justification for using a particular informant with well-defined objectives and a specific plan of action so that everyone, including the informant, has a clear understanding of what IRS is trying to achieve and his role in that effort, (2) periodic evaluations of his information to facilitate its effective use, to deter- mine whether the informant's efforts are consistent with the agreed-on objectives, and to provide a basis for timely ter- mination of unproductive or improper informant activities, and (3) systematic management evaluations of informant acti- vities to insure that the Intelligence Division's use of informants is in line with IRS' mission and consistent with established procedures. As shown in the next chapter, sound controls did not exist in the past. 11 CHAPTER 3 PREVIOUS MANAGEMENT CONTROLS WERE INADEQUATE TO MAXIMIZE BENEFITS AND MINIMIZE RISKS IRS had not done all it could to maximize the benefits that informants can provide while minimizing the problems they can cause. Informants were used in intelligence gathering efforts that were ill-defined as to purpose and overly broad in scope, IRS did not have a workable system or procedure for evaluating and assuring effective use of informants' informa- tion, and the Intelligence Division's use of confidential in- formants was not subjected to systematic review at any level of management. These serious weaknesses in management control resulted from lack of attention at all levels. CONFIDENTIAL INFORMANTS USED IN ILL-DEFINED AND OVERLY BROAD INTELLIGENCE GATHERING EFFORTS IRS management did not adequately participate in initiating confidential informant activities. Identification, develop- ment, and use of informants were left up to the special agents and other intelligence personnel who would be having direct contact with informants. Intelligence personnel were not re- quired to obtain higher level approval before using an inform- ant or before initiating an intelligence gathering effort in- volving informants unJess confidential funds were to be used; even then established procedures were often disregarded. Also, the procedures for requesting approval to expend confidential funds did not require the requester to submit information snowing how informants would be used, who they were, and how they were determined to be reliable. This lack of management participation and noncompliance with estab- lished procedures resulted in the use of confidential in- formants in intelligence gathering efforts which were ill- defined as to purpose and overly broad in scope. Need for more adequate justifications before approving the use of informants The only formal means that IRS district, regional, and na- tional management had of becoming aware of and approving the use of confidential informants was through confidential imprest fund procedures. II confidential expenditures, including payments to be made to an informant, were expected to exceed $500, a written request to gt the funds authorized had to be made to the district director, the regional commissioner, or the Director or Assistant Director of the Intelligence Division, depending on the amount of money needed. Guidelines in effect 12 until March 1975 required that n* * * written requests for authority to make confidential expenditures must contain a clear and concise statement of the objective sought, the amount requested, the pro- posed plan of action, the period of time over which the expenditure will be made, the anticipated results, and any other information of value to the official who will act upon the request." These requirements were usually not complied with. We reviewed 52 requests for approval to expend confidential funds in intelligence gathering efforts invol- ving informants. These were all the requests for money in excess of $500 on file for fiscal years 1972 through 1975 in the two regions selected for review, except for requests submitted in conjunction with the Narcotics Traffickers Pro- qram. We did not include these latter requests because IRS' national office initiated that program with formal guide- lines stating the objectives and establishing a committee at the national office to select specific targets. Of the 52 requests, 35 did not contain all of the re- quired information but were approved nonetheless. Of the 35 --22 did not contain a statement of the objectives sought, -- 26 did not show the proposed plan of action, --22 did not document the anticipated results, and -- 14 did not st ce the time period for making the expenditures. The following examples of unclear and incomplete re- quests to expend confidential funds show the type of ill- defined and broad-based intelligence gathering efforts in which confidential informants were involved. Example A The national office approved two related intelligence gathering efforts, involving payments of about $10,000 to one informant, to determine the extent of organized crime in a certain area. In 1973, a difference of opinion surfaced within IRS as to the extent of organized crime and its control of illegal activities in a particular IRS district. 13 District officials felt that illegal activities were not or- ganized; national office officials felt otherwise. To resolve this controversy, the national office initiated an intelligence gathering effort without the knowledge of regional or district officials. The request for authorization to expend confidential funds during this effort read as follows: "For approximately three months during the latter part of 1971 we utilized the services of a national type informant in [a specific geographical area] for intelligence purposes. This individual has been avail- able to us for the past eight years and has always proven reliable and industrious. "As a result of his numerous and varied contacts, he is able to obtain tax information which has been very useful to the Service wherever he may find himself. "This informant has advised me that he will be in the * * * area again in the very near future and would like to make available his services. "We are requesting authority to expend confidential funds in the amount of $6,000 for the period January 14, 1974 through March 31, 1974 so that we may purchase information from this informant." This request was approved by the Director, Intelligence Division, in January 1974. The informant was paid $5,345 during the term of this intelligence gathering effort. His information did not result in any investigations, indictments, prosecutions, convictions, or additional tax assessments. A national office official told us, nevertheless, that the then Director, Intelligence Division, considered the intelligence gathering effort successful because it indicated that illegal activities in the district were con- trolled by organized crime. About 2-1/2 months after the informant ceased operations, the district and region, unaware of those operations, requested national office assistance in resolving the question about the extent of organized crime in the district. Rather than use an undercover special agent, as the district and region had originally requested, it was decided to use the same in- formant as before because he was experienced and could save time, rm.oney, and effort. Accordingly, the Director, Intelli- gence Division, approved a request for authority to expend $12,000 for the purchase of information from this informant for the 6 months ending January 31, 1975. 14 We were told that the Director expected this second intelligence gathering effort to result in the identification of targets for investigation since the first effort had con- vinced him that illegal activities in the district were organized. The request for authority to expend $12,000 was not that specific, however. It merely stated that the informant would be used to resolve the controversy over organized crime's involvement in the district; to identify evidence relating to loan sharking, prostitution, bookmaking, wagering, and public corruption; and to identify large-scale tax frauds of any other nature. The informant was paid $5,041 during this second project. His information did not result in any investigations, indict- ments, prosecutions, convictions, or additional tax assess- ments. We were told that the Director, Intelligence Division, considered the intelligence gathering effort a failure. The Assistant Regional Commissioner for Intelligence, however, considered the informant's activity a success because it showed that gambling and loan sharking operations in the district were not organized. Example B IRS regional intelligence officials initiated an effort to gather information on political corruption because (1) an in- formant was available, (2) the district needed work, and (3) district and regional officials believed there was potential for developing cases in the area of political corruption. On June 3, 1974, the Acting Regional Commissioner approved a request submitted on that date by the Acting Assistant Re- gional Commissioner for Intelligence to expend not more than $5,000 in confidential funds through December 31, 1974. The request stated that an informant who had been proven reliable in the past by another Federal law enforcement agency "* * * has many connections and contacts in other districts in [this] Region and has expressed a willingness to conduct covert gatherings of in- telligence in areas of interest to the * * * Dis- trict. The Chief [of district intelligence] is interested in developing intelligence in the field of public corruption and the area of kickbacks from contractors to public officials." The Chief of Intelligence told us that he thought the effort had merit because the local news media had made numer- ous charges of political corruption against elected and ap- pointed State and local officials. He also told us, however, that the district had done very little preliminary work before the informant's arrival, and that the region considered 15 the effort a "shot in the dark." The district's work had consisted primarily of compiling a list of gubernatorial candidates which contained such information as their ages, previous employment, and current political status. The informant was to take the names of the candidates and attempt to identify the major contributors to their campaigns who might expect favors in return for their contributions. Then he was to attempt to develop information on candidates who IRS believed would accept payoffs if elected. The informant gathered information in the district for about 7 weeks and was paid $3,288. On July 26, 1974, the dis- trict suspended the effort pending an evaluation. It was not resumed primarily because the informant's information indi- cated that the objectives and scope of the effort were too broad. The special agent assigned to work with the informant evaluated the effort and recommended to the Chief of Intelli- gence that it be discontinued because "* * * it appears that the undercover assignment was an attempt to cover too broad an area. * * * If a specific entity, individual or a group of individuals can be isolated, it is possible that an undercover assignment of this type might prove effective." Regional intelligence officials who had sent the informant to the district subsequently evaluated the effort and also con- cluded that it should not be continued. One of them told us that (1) widespread political corruption was thought to exist in the area, (2) most of the special agents were known locally and therefore an informant was needed to gather information, and (3) he had determined that the informant was reliable. He said the objectives were intentionally broad so as to not restrict the informant's intelligence gathering activities. Example C On September 27, 1973, the Chief of Intelligence in one district asked the Assistant Regional Commissioner for Intelli- gence to authorize the expenditure of up to $2,500 in confi- dential funds through June 30, 1974. This was a continuation of an effort begun under a $500 authorization approved by the Chief of Intelligence in April 1973. The September 1973 request is quoted below: "Tne * * * District is probing for intelligence on [certain illegal activities]. We are also probing for intelligence in the general program. On October 1, 1973, we will begin paying an informant on a regular monthly basis. We are considering two others for 16 similar arrangements. To maintain our activities we will need an authorization of $2,500 for the remainder of fiscal year 1974, with $1,000 set up in a local imprest fund. [The Chief of Intelligence] will be cashier with [a group manager] an alternate." The Acting Regional Commissioner approved the request on October 2, 1973. On June 19, 1974, the Chief of Intelligence asked the region for an additional $3,000 to continue the effort through June 30, 1975, and this request was even more vague. The ob- jective, plan of action, and anticipated results were alluded to in only one sentence: "The * * * District is continuing to expand its program of developing informants and making payments for the production of evidence for use in the program." This request was also approved by the Acting Regional Commissioner. Example D On July 8, 1974, a special agent in another district sub- mitted a request, which was approved by the District Director, for authority to expend up to $2,000 in confidential funds. The request did not specify objectives, a plan of action, anti- cipated results, or the estimated time frame; it stated that: "The Intelligence Division currently has a confidential informant working in close association with an indivi- dual who is * * * involved in narcotics, gambling, pro- stitution, and possible loan-sharking. He is respected and acknowledged by persons within the [district] having strong organized crime connections." Example E On June 11, 1973, a District Director authorized the Chief of Intelligence to expend confidential funds up to $2,000. District officials were unable to locate a written request for the funds; the Chief of Intelligence told us the request was probably an oral one. In authorizing the funds, the District Director described generally the purpose of the funds as follows: "In accordance with our previous meeting in this regard, I understand that any authorized expenditures will be for the purpose of obtaining information valuable to the Internal Revenue Service and that the amounts expended will be consistent with your evaluation of the information received." 17 The authorization did not set out a time limit for expending the funds. The preceding examples are typical of most of the requests we reviewed--devoid of the minimum information specifically re- quired by IRS guidelines. The following, however, is an exam- ple of a request that contained more information than required; namely, information on why an informant was needed, how he was developed and determined to be reliable, and what his role would be in the intelligence gathering effort. Example F On November 1, 1973, the Chief of Intelligence requested the District Director to approve confidential expenditures up to $2,000 to pay an informant for a proposed 2- to 3- week effort. The request contained the following background information: "We [the district] currently have a Strike Force probe underway * * * with regard to allegations of illegal activities on the part of certain [county] officials. We have received information to the effect that the [sheriff and mayor] have been receiving payoffs to allow prostitution, pornography, and coin-operated gaming devices to flourish in [the city and county]. [Another individual] who owns a number of taverns and bars in the county and controls most of the coin- operated gaming devices, is alleged to be in control of most of the illegal activities in the county. * * * [This individual! is also alleged to have made payoffs to [state liquor control commission] agents in order to obtain liquor licenses for the various taverns and bars. "* * * We [the district] have exhausted our sources of information and are now seeking other ways to determine the truth of the allegations and the extent of the alleged corruption if the allegations are true. "[Another district] utilized an undercover operator on a situation somewhat similar to this [in another location] this past year. * * * The information provided by this undercover operator was of great value to the [other district] and assisted them in developing several criminal cases. [A district em- ployee has contacted this undercover operator] and has determined that the undercover operator would be available for [this] assignment. "The objective in placing an undercover operator in [the city] would be to determine whether or not an inter- 18 state prostitution ring is operating [in the county]; to try to determine whether allegations that possible organized crime money is being invested in [targeted individual's] coin-operated game devices operation, are true; to determine if, in fact, [state liquor control commission] employees are being paid off in order to obtain liquor licenses; and to determine whether or not the [sheriff and mayor] are, in fact, receiving moneys to allow prostitution, pornography, and gambling to occur in their city and county. "The exact approach the undercover operator will take to try to gather this evidence will be left up to him. His probable approach will be to try to go in and set up a house of prostitution and/or offer to buy a tavern or bar in order to determine what approaches he would have to make to the local officials and [the other target] in order to be allowed to [operate]. "The undercover informant feels that he could ac- complish this within a two to three week period of time. He wants living expenses plus $20 to $30 per day to purchase drinks and information. He also wants mileage for his automobile at the rate of 11 cents per mile. He also expects to receive payment for the information developed, if such information is of value to the Intelligence Division. * * *" The District Director approved the request on November 6, 1973. CONCLUSIONS To maximize the benefits confidential informants can pro- vide while minimizing risks and potentially bad effects, it is essential that intelligence gathering efforts involving inform- ants have clear objectives and definitive plans of action. Many of the intelligence gathering efforts we reviewed were initiated on the basis of assumptions or allegations that an illegal activity was being conducted in a geographical area or by a large group of people rather than on specific allega- tions of tax law violations by specific individuals. Other intelligence gathering efforts were initiated primarily be- cause informants were available and were willing to gather intelligence if paid. They were sent to localities not for any well-defined purpose but simply to gather whatever infor- mation they could on any individuals they might indentify as possibly being involved in illegal activities. In these efforts, !RS sent confidential informants out to gather and turn over information. These individuals therefore 19 served essentially in the capacity of undercover agents who, in effect, furnished their services to IRS, as distinguished from other confidential informants who furnished information on their own initiative and received no instructions or encourage- ment from IRS to obtain that information. IRS considers it necessary to use informants in undercover roles to gather intelligence about persons and groups who de- rive their income from illegal activities. But should IRS be involved in gathering intelligence on these type activities? This question looms important because in gathering such in- telligence IRS cannot avoid getting into the illegal sources of that income such as gambling, loan sharking, political cor- ruption, and narcotics trafficking, which are crimes outside of IRS' enforcement authority. The Intelligence Division's mission is to identify the existence of willful noncompliance by taxpayers and the de- vious and complex methods employed by them to evade the tax laws. In furtherance of this mission, the Intelligence Di- vision tries to identify and prove noncompliance by many individuals that derive their income from legal activities, such as tax return preparers, landscapers, horse trainers, jockeys, doctors, nurses, funeral directors, accountants, at- torneys, golf professionals, interior decorators, scrap metal dealers, dairy farmers, moonlighting policemen, and sales- men. In light of its mission and in the interest of equity, IRS would be remiss to ignore persons and organizations who derive their income from illegal activities and also vio- late the tax laws. To do so would be tantamount to granting immunity from compliance with the tax laws to a segment of the population--a segment considered by IRS as having a high probability of willful noncompliance. As stated by the Com- missioner of Internal Revenue in hearings before the Subcom- mittee on Oversight of the House Ways and Means Committee in 1975: "* * * those who make their livelihood from violations of the laws of this Nation other than [the tax laws] are some- what unlikely to comply with the tax laws." Accepting the premise that IRS cannot iqnore the tax impact of illegal activities, the question then centers around the use of confidential informants in efforts directed toward those activities. Intelligence Division personnel at all levels believe that confidential informants are essential because (1) they can sometimes provide information not other- wise available to IRS because of their association or relation- ship with individuals suspected of violating the tax laws, (2) they sometimes have special experience and knowledge which a special agent does not have and could not acquire, and (3) they are sometimes already in a position to provide information. 20 Although confidential informants may sometimes be essential, they can also cause serious problems. Thus IRS must exercise caution in deciding when and how to use their services. By using confidential informants in undercover efforts that were ill-defined as to purpose and overly broad in scope, IRS was not exercising sufficient caution. Indeed, even exploratory efforts can and should be properly focused. Even though many requests to expend confidential funds did not contain information required by IRS procedures, they were approved by officials at all levels of management--dis- trict, regional, and national. By approving the requests these officials were not exercising sound management control over intelligence gathering activities but were, in effect, condoning and contributing to noncompliance with established procedures and controls. Not only were established procedures not followed but the procedures themselves were inadequate because they did not require all the information necessary for adequate manage- ment control of informant activities. Because of the risks involved, informants should be used only if the information needed cannot be obtained through regular enforcement tech- niques. Therefore, approving officials should have information available to enable them to determine whether the objectives could be achieved through more desirable and less risky means. IRS procedures, however, did not require inclusion of this type of information in the requests. Specifically, the proce- dures did not require an explanation of how informants were going to be developed and used, how they were determined to be reliable, and why it was necessary to use them rather than other enforcement tools, such as undercover special agents. Thus, the approving officials did not always know whether an inform- ant was really needed or what role the informant would play in the intelligence gathering effort. PROCEDURES FOR ACCUMULATING, EVALUATING, AND PROCESSING INFORMATION PROVIDED BY INFORMANTS WERE INADEQUATE The quality of a confidential informant's information and the way it is used determine the extent to which the Intelli- gence Division is successful in using confidential informants to identify individuals and organizations not complying with the tax laws. Intelligence officials in one IRS region said: "An effective Intelligence operation is a direct function of its capacity to develop sources of information and to act intelligently on the information at the earliest possible time." 21 IRS intelligence officials directed and encouraged special agents to develop and use confidential informants. About 2,000 such informants furnished information for pay during fiscal years 1973 through 1975 while thousands more furnished information free. This information supplemented that developed through other means by the Intelligence Division's 2,600 special agents. Although the Intelligence Division accumulated a mass of information from informants and other sources, it did not have a workable system Jr procedure for evaluating that information --either when it was obtained or at periodic intervals--or for assuring its effective use. Such a system or procedure could have not only improved the tax consequences of informants' information but also provided a means of controlling informants' activities and intelligence gathering in general. Procedures for evaluatinq information q for it were ineffective before_pay__g IRS' policy on confidential expenditures stated that, as a rule, informants should be paid "* * * only after the infor- mation or evidence has been obtained, evaluated, and determined worthy of compensation." IRS recognized, however, that situa- tions might arise where this would not be practicable. There- fore, IRS permitted payments to be made to "* * * lay the groundwork for the procurement of information." This would include undercover work and other activities requiring the ex- penditure of confidential funds to obtain information. It may be necessary, for example, to rent an apartment that the informant could use as a base for gathering information. The policy requiring information to be evaluated before it was paid for had little effect because the Intelligence Division generally did not purchase specific information. Instead, the Division usually paid informants to gather infor- mation. Although this practice was permitted, it nullified the basic requirement that information be evaluated before any payments were made. Intelligence personnel usually made fixed, periodic pay- ments to informants for their efforts in gathering information. District intelligence officials told us this was often the only practical way to pay informants because they devoted much of their time to IRS. In other words, if IRS expected an informant to devote a good deal of time gathering information, IRS was going to have to pay him periodically--much like a regular employee. Obviously, there was no intent to receive and evaluate information before paying informants, and the payments did not necessarily bear any direct relationship to the value of the information received. 22 For example, special agents in one district paid an in- formant $100 a week for approximately 7 months to identify and covertly gather information on individuals thought to be involved in narcotics trafficking. The Chief of Intelligence in the district told us the informant had proved himself in a prior investigation in another district and had the ability to gather valuable tax information. The Chief said that the first few hundred dollars were paid to get the informant set up in an apartment and that Intelligence personnel did not evaluate each item of information before paying the in- formant but that he believed the information was worth at least $100 a week. Intelligence officials at IRS' national office also made fixed, periodic payments to informants. For example, one informant received 17 monthly payments between January 1972 and July 1973--15 of $240 each, 1 of $200, and 1 of $300--while another informant received 8 weekly payments of $375 each and a total of $493 for expenses. An intelligence official in one IRS regional office paid an informant $462 to gather information about alleged skimming (taking money before it is reported in the records as received) in the textile industry. The official told us that the infor- mation the informant was gathering was not very good, but the payments were made to retain him in case a situation might arise when he could obtain gocd information. This same offi- cial made two payments to another informant totaling $1,660 "to keep the informant happy and willing to co-operate" rather than for specific information he was providing. Because of the procedures followed in using and paying in- formants, we could not always relate payments with information received, if any, and could not always determine whether the in- formation was evaluated. IRS internal auditors reported that in 11 of the 22 districts they visited, they could not effectively evaluate the type and quality of information obtained from informants because a written record of the information was not always prepared. They reported further that: "* * * information obtained from informants was not always evaluated prior to incurring confidential expenditures as required by [IRS policy]. For example, in 9 of the 22 districts, 13 confidential informants were paid a total of $206,585 on a weekly or other recurring (i.e., biweekly or monthly) basis. As a result, the information from these informants was generally purchased without any evidence of evalu- ation prior to payment and there was little if any direct control exercised over the type of information provided." 23 Informants' information was not periodically reviewed and evaluated Special agents were given broad discretion as to the ex- tent and type of information they could gather through inform- ants and the disposition of that information. Information was usually kept in background files which were set up and main- tained by the special agents or group managers having contact with informants. These files were not a part of any formal IRS information system, were not subject to formal periodic reviews, and sometimes contained personal taxpayer information that did not appear needed to enforce the laws. Agents and group managers were encouraged by intelligence officials to build their own information files on individuals who might be candidates for formal IRS investigations. These files were arranged by informant or taxpayer name, by professional group, or in any other convenient manner. Agents and group ma agers were not required to tell anyone about the files until they filt enough information had been accumulated to initiate a forrial investigation. One Chief of Intelligence, at a meeting of special agents and group managers, described the procedure as follows: "* * * Each Group Manager should have a drawer full of back- ground items that he feels would be adequate to assign for criminal investigations as needed. He should ac- cumulate this information by requiring each Spkecial Agent to have at least three background items in his file, developing them to a point where they will be ready to go for a criminal investigation. The Group Manager might also assign an agent in his group to Intelligence Gathering for whatever percentage of time is necessary to accumulate and maintain this inventory of background items worthy of assignment. (These should not be made into Information Items but get copies of returns, public records, etc., and hold for assignment.) Once the Manager has this 'Bank', he can evaluate these items and pick the better ones to assign as needed." This method of accumulating and maintaining taxpayer information did not provide fo. any formal evaluation until the special agent requested it. Once the agent decided an evaluation was needed, he would prepare an intelligence infor- mation item showing, among other things, the taxpayer's name and a description of an alleged tax violation. At that point, the information would be formally evaluated by the group mana- ger and a decision made as to whether specific enforcement action was warranted. 24 IRS guidelines did not adequately define tax-related information, and IRS had no procedure for identifying and properly controlling information which did not relate to possible tax violations or other Federal crimes. Because of this and because of the absence of systematic reviews of information accumulated, the background files maintained by special agents contained personal information on taxpayers which did not appear essential to enforcing the tax laws. IRS auditors reported. fnr example, that of 4,004 documents prepared from infori; !. ' information in 22 districts, 562, or 14 percent, contair - i isically non-tax-related information. Of the 562 documentb. 4 or 1.3 percent contained information of a derogatory nature such as references to taxpayers' sex and/or drinking activities. In reviewing background files we, too, noticed that special agents occasionally obtained information from informants that related to individuals' social relationships and activi- ties and that did not appear necessary for enforcing the tax laws. Intelligence officials told us, however, that this type of information may be useful. The Chief of Intelligence in one district, for example, told us that special agents were trained to document all information provided by informants and that information about a taxpayer's girlfriend, spouse, and spending habits could be useful in developing further information on the taxpayer's taxable income. In any event, IRS had not attempted to define or limit the type of information which could be gathered and retained by special agents, and the information gathered was not sys- tematically reviewed. Problems in the design and implementation of an overall intelligence gathering and retrieval system Because it recognized the need for a formal system of gathering and using information obtained from confidential in- formants and other sources, IRS developed a computerized in- telligence gathering and retrieval system. The system was tested in five districts from May 1972 to May 1973, when it was formally adopted for implementation nationwide. IRS intended the system to provide an effective, uniform means of gathering, evaluating, cross-indexing, and retrieving intelligence data. A group or an individual in each district was responsible for evaluating newly received information, ar- ranging for it to be entered into the system, and periodically reviewing it. Guidelines for the system did not define tax- related information but did provide (1) criteria for including information in the system and (2) examples of documents which could be included. The system provided periodic listings 25 of national and district intelligence data in several formats. In general, the system was designed to provide better control over a'd more effective use of intelligence data. In December 1974, IRS began a study to determine whether only directly tax-related information was being entered into the system and in January 1975 the Deputy Commissioner directed that all activities related to the system be suspended until completion of the study. In a Juie 20, 1975, report on a test of the system in four districts, IRS internal auditors reported problems relating to supervisory control over information entering the system, re- trival and evaluation of information, and compliance with in- structions. They also reported a need for more specific in- structiors an the type of information to be entered into the system and the objectives of accumulating information. The system was officially discontinued on June 23, 1975, when IRS issued new information gathering guidelines. CONCLUSIONi. Informants are valuable to IRS only if they provide in- formation hiich is determined to be useful in terms of further intelligence gathering; a criminal investigation, indictment, or prosecution; or tax assessments and collections and which otherwise would not be available to the agency. This requires a timely, systematic analysis and evaluation to determine the information's tax implications. IRS' procedures generally provided that an informant be paid only after his information was evaluated. This procedure, in theory, provided for systematic, timely evaluation of information. In reality, however, it proved inadequate. If an informant would approach IRS voluntarily with information that he wanted to exchange for money, the procedure would work perfectly--review the information, evaluate its worth, and pay for it. If an informant was sent out to gather information, however, the procedure would become inoperable because the Intelligence Division tended to pay such informants for their services rather than for specific information. In such cases, IRS should have had some other procedure to insure timely and systematic evaluation of information, but it did not. To obtain information on narcotics trafficking, gambling, political corruption, and other illegal activities, IRS used informants to infiltrate those activities, identify individuals involved, and gather information on them. In these situations, the informants' efforts were not (1) directed toward obtaining specific information, (2) limited to specific individuals, and (3) restricted to specific time periods. As a result, they frequently gathered information over long periods of time-- 26 information that was general in nature, that was not always clearly tax related, and that contained the names of numerous individuals. Even so, the information might still have been usefu. to IRS and therefore should have been systematically evaluated. IRS did not adequately define what information could be gathered and retained; after information was accumulated, it was stored by special agents in their background files and not subject to formal, periodic evaluation. We could not quantify the effects of this practice in terms of prosecutions lost, taxes not assessed, or the like. It is reasonable to assume, however, that a procedure requiring thorough evaluation and effective processing of information received from inform- ants would have enhanced IRS' chances of achieving meaningful results from its intelligence gathering efforts. Its attempt to develop a uniform system to gather and retrieve information was clear recognition of this. Without periodic reviews of information gathered by special agents, IRS also has little assurance that agents are using in- formants to gather information related to tax law enforcement and in ways which will not embarrass IRS and violate taxpayers' rights. Although reviews may not prevent these things from happening, they would provide better assurance of early detec- tion and correction. IRS comments and our evaluation In a July 27, 1977, letter (see app. I), thc Commissioner of Internal Revenue strongly disagreed with our contention that IRS did not have a workable system or procedure for evaluating and assuring use of the information received from informants. The Commissioner noted that "* * * Until March, 1976. it had been a long- standing procedure that when the Intelligence Divi- sion received an information item, it would be evaluated by the Chief, intelligence Division, or his/her delegate for possible criminal potential. If the information item did not have criminal po- tential, but did have Audit or Collection potential, the Chief, Intelligence, would forward the informa- tion item to the appropriate function for evaluation. In March, 1976, the evaluation process was trans- ferred to the Service Centers. "We believe that the probler-, that GAO discovered were related to the handling of information items generated from long-term informants, which were not prepared and evaluated as expeditiously as would 27 have been expected. However, this was not completely a problem with the procedures, but rather in the fact that a few Chiefs of the Intelligence Division did not follow existing procedures. * * *." The Commissioner is making the point that IRS' procedures were good once an information item was prepared. We do not disagree. As we said on p. 24: "This method of accumulating and maintaining tax- payer information did not provide for any formal evaluation until the special agent requested it. Once the agent decided an evaluation was needed, he would prepare an intelligence information item * * *. At that point, the information would be formally evaluated by the group manager and a de- cision made as to whether specific enforcement action was warranted." Thus, the problem noted during our review and confirmed by IRS' comments, related to the procedures for accumulating, evaluating, and processing informant information before an information item was prepared. IRS recognized that its pro- cedures were inadequate when it developed the computerized intelligence gathering and retrieval system ard again later when it discontinued that system. We believe the results of our review and the actions taken by IRS clearly demonstrate that IRS' system or procedure for evaluating and assuring effective use of informant information was unworkable. MANAGEMENT EMPHASIZED EXPANDING--NOT EVALUATING--INFORMANT ACTIVITIES Intelligence officials at all levels emphasized and en- couraged developing and using informants as sources of in- formation--an emphasis that managers outside the intelligence function were aware of and approved. Managers within and outside the intelligence function, however, exhibited in- adequate concern over the risks and potential problems of this aspect of intelligence operations and did not provide for systematic evaluations of it. When intelligence operations were reviewed, managers were generally more concerned with the number of cases generated and individuals prosecuted than with the methods used to achieve those results. Management emphasis on greater use of informants IRS management, at all levels, continually emphasized improving intelligence gathering through greater use of informants. 28 National, regional, and district offices prepared annual plans establishing priorities and goals for intelligence operations. The plans were approved by the Assistant Commis- sioners for Compliance, regional commissioners, and district directors respectively. In its plans, the national office established goals and objectives in such areas as intelligence gathering, allocation of resources, case selection, and case management. For ex- ample, plans for fiscal years 1973, 1974, and 1975 required each region and district to spend 5 to 10 percent of direct investigative time on identifying noncompliance. The fiscal year 1974 plan stated that the primary objective of the intel- ligence gathering activity would continue to be the discovery of willful noncompliance situations that have widespread implications and that: "* * * Attainment of this objective requires emphasis on: * * * Development of, and regular contact with individuals who have access to significant information pertinent to tax law violations." The plan established a general goal for allocating resources to intelligence gathering as follows: "Not less than 10 percent of direct investigative time must be allocated by each region for identifying non- compliance." The regional offices developed plans in line with the general goals established by the national office but refined them to meet the particular needs of the districts. The re- gional plans usually contained an assessment of the districts' past efforts in meeting established goals and were more speci- fic in establishing goals and identifying areas to be empha- sized in the upcoming year. In particular, the regional plans were much more emphatic about instructing districts to expand information gathering activities through the use of informants. For example, the fiscal year 1972 plan prepared by one region emphasized the need to develop sources of information and stated: "There has been too little effort and money going into the development of hard information." "Many special agents have not been fully attuned and devote too little attention to developing meaningful contacts." "There are unfortunate 'hang-ups' about purchasing information." 29 The plan also stated that the region would look at "develop- ment of information s.urces" in evaluating district office performance for fiscal year 1972. The fiscal year 1973 plan for the same region reempha- sized the need to gather intelligence through informant' and stated: "Each district must do more to develop current informa- tion in Intelligence on noncompliance." "There should be a significant increase in cases orig- inating from special agent sources in FY 1973." "A significant increase should occur in information arising from paid informants." The fiscal years' 1974 and 1975 plans for this region continued to emphasize intelligence gathering and the use of informants. For example, the fiscal year 1974 plan stated: "FY 1973 marked the beginning of real progress in the development and use of Intelligence information sources. However, there are marked differences among the districts and much remains to be done in the current fiscal year to broaden the scope of our activities, to develop a large group of informants, and to refine existing infor- mation retrieval systems in the districts." District offices prepared work plans which reflected the emphasis placed on information gathering and the development of informants as sources of information. For example, the fiscal year 1973 plan for one district stated: "* * * Our prime purpose will be the development of a true sense of urgency in everything we do. Our broad objectives will relate to fostering an effort which will result in the identification of areas of noncom- pliance." "The objective of each manager and special agent will be to know his geographical area of consideration for the purpose of identifying noncompliance in industries and occupations by the development of informants and [other] sources of information." This plan also established a specific informant goal for each special agent: " * * one of the prime functions of Intelligence is the gathering of meaningful data or information rela- tive to areas of noncompliance. In chis regard, each 30 agent will be responsible for the development of at least one workable informant. 10% of our manpower will be expended in [the intelligence gathering] area." The fiscal year 1974 plan prepared by another district emphasized intelligence gathering and the use of informants as follows: "* * * Each special agent has shown an active interest in Intelligence gathering activities and as a result we have developed many of the projects and cases presently in inventory. During Fiscal Year 1974 we will continue with the high degree of activity in this program." "* * * each special agent will continue to devote a portion of his work time to case development activities. Except under unusual circumstances this should not ex- ceed 13% of each agent's total duty time." "Each agent will develop sources of information and cultivate informants who can provide us valuable information." Informant activities not systematically reviewed by national, regional, and Tistrict office management Even though IRS man-gers repeatedly emphasized greater use of informants in intelligence gathering activities, they did not focus on how informants were being used in their subsequent reviews of intelligence operations. National and regional management periodically evaluated overall Intelligence Division operations. In addition, re- gional intelligence officials evaluated specific projects or programs to resolve problems or answer specific questions as they arose, such as those related to the need for additional funds. None of these evaluations, however, routinely included a review of informant activities. Moreover, district manage- ment did not periodically review Intelligence Division opera- tions. National office evaluations The national office intelligence staff usually visited each region each fiscal year under a National Office Review Program. The overall purpose of the program was to deter- mine the extent of the regions' familiarity with district operations and determine whether the regions were providing adequate leadership and guidance to district offices. 31 Reviews of intelligence operations were made in conjunc- tion with reviews of appellate and audit operations. There were no formal guidelines setting out the review procedures to be followed. Usually, the national office representative who conducted the intelligence review would hold discussions at the regional office on problems identified in previous reviews, problems identified by Internal Audit, and specific programs of interest. The national office representative would normally visit oie or more districts in each region to review such things as management effectiveness, staff problems, and the responsive- ness of the districts to priority programs such as narcotics trafficking and Strike Force. The reviewing official would prepare a report and submit it to the Assistant Commissioner for Compliance. The Deputy Commissioner would be briefed on the results of the reviews and would attend the closeout con- ferences held with the regional commissioners. We reviewed the reports of the national office reviews made during fiscal years 1973 through 1975 in two regions. The format of the reports varied somewhat, but the content was generally the same. The fiscal year 1974 reports for these regions assessed the impact of the office of the as- sistant regional commissioner for intelligence on regional intelligence operations in terms of program direction and support provided to the districts and generally summed up the region's statistical accomplishments in terms of recom- mended prosecutions. None of the reports ,we reviewed contained an assess- ment of informant activities as such. The fiscal year 1974 report on one region, however, addressed a problem relating to the lack of direction and uniformity in information gathering. The problem was identified as a result of an analysis of fiscal year 1973 information gathering efforts in selected districts in that region. This report stated: "There was a great disparity among districts in the percentage of time applied and the manner of collecting, storing and evaluating information. In the three dis- tricts visited * * * there appeared to be a need for greater direction and more specific objectives in their infcrmation gathering efforts." "The [Assistant Regional Commissioner for Intelligence] is aware of the situations described above and recently conducted a seminar of district information gathering representatives. The purpose of the meeting was to exchange ideas and to discuss common problems in this area. The meeting enabled the smaller districts, in 32 particular, to learn from the larger districts that had more experience in this activity. The [Assistant Regional Commissioner for Intelligence] plans to monitor the future direction of this program to ensure that districts have specific objectives outlined [for information gathering efforts] * * *." (Emphasis added.) The national office reviewing officials concluded that the districts should have specific objectives for information gathering and that this would result in the districts being able to limit the allocation of their resources for informa- tion gathering to the 10-percent goal established by the na- tional office for fiscal year 1974. National office officials apparently did not attempt to determine whether this problem was widespread so as to identify underlying causes and provide nationwide guidance and control. Instead, when national office reviews identified problems with intelligence operations, the approach seemed to be one of dealing with the specific problem rather than making comprehensive evaluations. For example, the report on the fiscal year 1975 visit to the region discussed earlier did not mention the informa- tion gathering problem identified in the prior year even though the problem seeded unabated. We analyzed 13 requests for confidential funds prepared by districts in that region after fiscal year 1973; 7 contained vague objectives--the same problem identified in 1973. Of the national office reports that we reviewed, only one directly referred to informant activities. The report empha- sized the need for one of the districts in the region covered by the review to develop and use informants. This fiscal year 1973 report stated: " * * * They [the district] are aggressive in developing intelligence regarding organizations. However, no real 'street work' is being done in intelligence gathering. For instance, the Narcotics Traffickers Program Group Supervisor expressed a negative attitude toward developing informants in this area. This attitude was derived from his own unsuccessful experiences in obtaining useful in- formation from informants when he was an agent and does not reflect any lack of enthusiasm for his present duties." "Concerning the Strike Force Program, the Chief expressed the opinion that there is no identified organized crime in the * * * area. While we recognize that this opinion is expressed by many people in the * * * area, we believe that Intelligence should make an effort to develop in- formants and gather street intelligence to identify the 33 scope and extent of organized illegal activity in this large permissive city where prostitution, pornography, gambling, and narcotics are ever present." "The Chief also expressed the opinion that in the * * * area there is little to gain from informant development; that none of the enforcement agencies have developed informants. [The] NTP Group Supervisor, shares this negative view. [The supervisor] said that he does not encourage the use of informants and that his personal experiences with informants were less than gratifying." "The * * * Intelligence Division is now developing a program designed to identify the nature and scope of organized crime through investigative techniques in- cluding surveillance and informant development." The report praised another district in the same region as follows: "[the district] - Very effective in investigating targets. * * * Agents are developing informants and gathering information on all types of criminal activity." Regional office evaluations IRS required each assistant regional commissioner for intelligence or his staff to visit each district at least twice each fiscal year and the assistant regional commissioner had to personally participate in one visit to each district or in the closing conference with the district director. The assistant regional commissioner was to use the visits as a means to keep the district directors and regional commissioner informed of district intelligence operations. The major purposes of the visits were to -- evaluate the effectiveness of district intelligence programs, activities, and supervisory personnel; -- discuss intelligence problems with district officials and provide on-the-spot suggestions and assistance for improving intelligence operations; and -- follow up on actions taken to correct deficiencies noted in internal audit reports. The guidelines required the assistant regional r':mmissioner to submit a written report on each visit to the -; ,;.al com- missioner who in turn was required to furnish a copy to the district director. A copy of each report also was to be 34 furnished to the Director of Intelligence at the national office. The reports were to cite important observations made during the visit, deficiencies noted, and corrective actions the district director had taken or planned to take. Intelligence officials made these visits, but they did not specifically evaluate informant activities. Instead, they were concerned with such things as case management, case inventory, referrals from the Audit Division, special agent caseload, overall staffing levels, and effectiveness of district intelligence managers. One regional official told us that review teams were mainly concerned with assess- ing case management and were not concerned with specifically assessing informant activities. This official told us that the region assumed an operational view; that is, the visits attempted to answer such questions as: Are the districts using their people to the best advantage? Is a specific program or effort having an impact on compliance? Are the districts following established procedures? According to reports issued on these visits, regional officials sometimes reviewed areas directly related to in- formants, such as the effectiveness of a district's overall intelligence gathering program, but they did not focus on how informants were being used, the type of information they were providing, or the procedures being used to pay them. Generally, when a report mentioned intelligence gathering, it was an expression of concern over too little intelligence gathering or a statement emphasizing its importance. For example, in a report prepared on one district in fiscal year 1972, the Assistant Regional Commissioner for Intelligence stated: "Your [the district's] case development and [certain other efforts] have consumed 10.3% of your direct in- vestigative time in the first eight months of FY 1972. Other priorities might dictate a temporary reduction of [this] effort going into FY 1973. However, we 1the region] ask that you resist any such temptation. In fact, there is a great need to expand intelligence gathering relative to organized crime and corruption in [the State] generally." "We [the region] know that you [the district] will continue to give priority to your Narcotics and Strike Force Program commitments. You have done well in both program areas and future potential looks even better. We only ask that you take two additional steps as soon as possible." 35 "(2) Additional 'street' information is needed con- cerning organized crime and corruption particularly in the * * * area." A report on another district indicated concern that intelligence personnel were depending too much on Audit Division referrals and stressed the need for special agents to become more involved in intelligence gathering. Special reviews of intelligence operations Regional intelligence officials made unscheduled visits to review specific information gathering efforts or programs. Officials in one region told us that such visits were not made to evaluate informant activities. Instead, they were made either (1) to investigate and resolve specific problems or (2) because of interest in a particular information gathering effort or intelligence program. We reviewed the reports on unscheduled visits in one region. Informant activities were reviewed in detail in only one instance. In November 1973 the Director of Intelli- gence at the national office requested the regional office to review the intelligence gathering effort known as Operation Leprechaun before a decision was made on whether to authorize additional confidential expenditures. A member of the regional intelligence staff reviewed the effort and prepared a report that included information on how informants were being used and references to the type of information being obtained from informants. A copy of this report was sent to the Director of Intelligence. (The results of this visit are discussed in more detail in app. II.) Although other visits included reviews of specific information gathering efforts, they dealt primarily with the expected results of the effort and problems related to staffing. District office evaluations Consistent with management at all other levels, district officials did not periodically evaluate and report on inform- ant activities. District directors authorized confidential expenditures for amounts between $500 and $2,000; however, they did not conduct, or provide for, systematic reviews of the information gathering efforts for which the funds were used. The chief of intelligence in each district could authorize confidential expenditures up to $500 but did not have to submit a report showing how the money was used or the specific re- sults achieved. In addition, the chiefs did not periodically 36 review the information obtained from informants or require group managers to perform such reviews. District management generally did not feel that inform- ant activities warranted much of their attention. The Dis- trict Director in one district told us, for example, that he wanted the Chief of Intelligence to keep him apprised of pro- blems but that he did not want to become involved in the de- tails of informant activities. A Chief of Intelligence in another district told us that, under the procedures in effect prior to March 1975, an information gathering effort was reviewed "when someone decided it was necessary" and the success of an effort was measured by results, such as the number of investigations and prosecutions. Informant activities not evaluated by Internal Audit IRS internal auditors periodically reviewed specific Intelligence Division programs such as those involving narcotics traffickers and the Strike Force and the reports resulting from those reviews occasionally contained comments pertaining to informant activities. As a matter of policy, however, Internal Audit did not specifically review informant activities during its regular audits. Internal auditors in the southeast region told us that their reviews sometimes included intelligence gathering activities, but that this was only incidental to their work on specific intelligence programs and was not an area specifically designated for review. In 1974, for example, IRS auditors rtviewed the Strike Force program in two districts. Although their review covered certain aspects of intelligence gathering, it did not specifically deal with informants. In addition to audits of selected Intelligence Division programs, internal auditors annually reviewed confidential imprest funds. The auditors, however, did not evaluate the activities the funds were being used for or the information received to identify management controls over the use of informants. Internal Audit officials at the national and southeast regional offices told us they had not reviewed informant activities because they did not have full access to Intelligence Division records, primarily those showing informants' names. Internal Audit officials in the region told us that intelligence officials resisted auditors' attempts to review confidential payments made to, and information supplied by, informants. 37 The Director of Internal Audit told us that internal audit coverage of informant activities in the southeast region had been inadequate because his auditors continually met resistance whenever they delved into intelligence mat- ters requiring access to informant records. According to the Director, this frustration finally led to less than adequate audits of the imprest funds. For example, the auditors would do their work at the regional office where the imprest fund was located but ignore supporting docu- mentation in the affected district. The Director noted, however, that this was not the case in all regions. The Director of Intelligence, on the other hand, told us that Internal Audit had access to any records it wanted except for the true identities of confidential informants. IRS apparently resolved the question of access by issuing guidelines in February 1977 which specifically provide for Internal Audit access to informants' names. CONCLUSIONS Although informant activities represent a relatively small facet of Intelligence Divison operations, they require special management attention because of the unique risks they create. It is not enough to review informant activities when problems or questions relating to those activities arise because that fails to provide the continuous feedback that IRS management needs to assess the risks of using informants versus the contri- butions they may provide. While IRS managers encouraged development and use of informants as an investigative tool, particularly in intelligence efforts directed toward illegal sources of income, they did not acknowledge the need for special management attention to how those informants were being used. This failure to appropriately recognize the inher- ent risks involved in using informants was due in part to the apparent sense of urgency on the part of intelligence officials to "get the job done." While management exhibited a great deal of concern for establishing goals and producing results, it gave little attention to how those results were achieved. 38 By independently examining and appraising operations, Internal Audit can provide a valuable source of information on how the Intelligence Division conducts its operations, thereby complementing all other elements of management con- trol. To be of maximum usefulness, however, the scope of Internal Audits' activity cannot be restricted. It should, instead, extend to all agency activities and related man- agement controls. By not adequately reviewing informant activities IRS internal auditors did not effectively contribute to management control of those activities. 39 CHAPTER 4 MANAGEMENT CONTROLS HAVE BEEN IMPROVED BUT MORE NEEDS TO BE DONE Between 1975 and 1977 IRS issued new guidelines which addressed many of the weaknesses we identified concerning the use of confidential informants. IRS could do more, however, to better insure that an informant is used only after the potential benefits and risks have been properly assessed and that management is sufficiently attentive to how informants are being used. ACTIONS TAKEN TO CORRECT WEAKNESSES When Operation Leprechaun was publicized in 1975, the Intelligence Division's management controls over the use of confidential informants were inadequate. The only formal guidelines IRS had relating to informants were those dealing with confidential expenditures. In 1975, these guidelines were substantially revised to strengthen controls over paid confidential informants. Since 1975 IR3 has developed new guidelines for many other aspects of the Intelligence Divi- sion's operations including the use of informants and infor- mation gathering. (See app. III.) The guidelines provide needed procedures and controls for initiating intelligence gathering efforts and controlling informant activities. They -- require a written authorization before any IRS employee gathers intelligence, -- prescribe controls over the use of all paid informants and those unpaid informants who are encouraged to furnish information to the Intelligence Division by requiring the chief of intelligence in each district to approve their use and maintain a record on each one used, -- spell out what informants can and cannot do in gathering information for IRS and provide instructions as to what IRS employees should do when they become aware of il- legal informant activity, -- specifically require managers to review and report on informant activities, and -- raise the management level at which confidential expenditures may be authorized. 40 The guidelines require a timely evaluation of information provided by informants and provide clear and specific proce- dures for processing the information. The guidelines also (1) state that only "directly tax related information" may be gathered, (2) define "directly tax related," and (3) provide examples. Intelligence and other IRS officials are now re- quired to periodically evaluate those aspects of intelligence operations directly related to confidential informants. Implementing the guidelines will be a step toward over- coming past problems and weaknesses in managing confidential informant activities. To help achieve effective manegement control over informant activities, however, several aspects of the guidelines need to be strengthened. Specifically, the guidelines do not -- provide enough control over the use of informants with the result that informants could be used to covertly gather information without approval above the level of the chief of intelligence or -- provide for adequate consideration of such factors as how an informant was developed and determined reliable, how he will be used, and why he is needed before au- thorizing the use of an informant, the gathering of in- formation, or the expenditure of confidential funds. Initiating intelligence gathering efforts and using informants Under IRS procedures in effect until March 1975, intelli- gence gathering efforts involving informants had to be approved only if confidential funds were going to be expended. Approval above the level of chief of intelligence was required only if the total amount of the payments to informants and other confi- dential expenditures was expected to exceed $500. In 1975 IRS, for the first time, issued forical guidelines which required written authorization for information gathering. The guidelines distinguish between (1) a study, survey, or can- vassing activity on a group of taxpayers (who may or may not be identified by name) within such categories as an occupation, an industry, a geographic area, or a specific economic activity and (2) an information gathering effort directed at one or more specifically identified taxpayers. In connection with this dis- tinction, the guidelines require studies, surveys, or canvassing activities to be approved by the district director, while the chief of intelligence can approve information gathering on spe- cific taxpayers. Intelligence gathering efforts to be conducted by the regional or national offices require the approval of the regional commissioner or Assistant Commissioner for Compliance, respectively. 41 A request for authorization to conduct a study, survey, or canvassing activity must state the purposes, define the scope, and specify the est dlated life of the effort and the type of information to be 5(,thered. In authorizing informa- tion gathering on individuals, the chief of intelligence must specify the known or assumed name of the taxpayer and the reason information gathering has been authorized. In December 1976, the Commissioner of Internal Revenue asked for our views on IRS' proposed informant guidelines. We provided comments in January 1977, and IRS issued the guidelines in February after making some of the changes we suggested. The guidelines require the use of each "controlled informant" to be authorized by the chief of intelligence. Controlled informants are those who receive payment or who, although not paid, are encouraged to furnish information to the Intelligence Division. The chief is to consider cer- tain factors in evaluating the potential use of each con- trolled informant and is to prepare for the file a written summary ot the factors considered. The factors include -- the available background information on the informant, such as his address, criminal record, and associates; -- the reliability of information provided to IRS and other agencies, if known; and --the informant's source and means of securing informa- tion. Special agents who use controlled informants must instruct them to not use unlawful techniques (such as breaking and en- tering, electronic surveillance, and opening or otherwise tam- pering withl the mail) in obtaining information. In addition, the guidelines prohibit using informants to do things which IRS could not authorize its undercover special agents to do. The chief of intelligence must also authorize the use of a "restricted source." This is an unpaid informant who is known to an IRS employee but who refuses to provide informa- tion without assurance that his identity will not be included in any IRS record. 1/ The chief may authorize his use only if the following conditions are met. l/This type of informant differs from an anonymous inffrmant-- a person whose identity is unknown to anyone in IRS, such as a person who provides information by means of an anony- mous telephone call or unsigned letter. 42 --The information is represented by the informant as being of such significance and IRS feels that it will be, that failure to obtain it--even under such re- strictions--would likely be neglect of duty. -- Reasons exist for believing the informant to be highly reliable. -- There is no reason to believe that the informant is engaged in an illegal activity. -- Little likelihood exists that IRS can obtain the information from other sources. In our January 1977 comments on the proposed guidelines, we questioned the appropriateness of classifying some inform- ants as "restricted sources" because we felt it important, for control purposes, that IRS have a record of the true identity of every individual who provided information at IRS' request or encouragement. Although IRS did not delete that classifica- tion from its guidelines, it did revise them to require each regional commissioner to evaluate the use of restricted sources fcr a 6-month trial period beginning February 3, 1977. The evaluation is to show the number of times each district used a restricted source informant and is to include a determination of whether the value of the information was such that IRS would have been negligent not to accept it. Each regional commissioner is to forward his evaluation along with recom- mendations to the Deputy Commissioner who will then deter- mine whether IRS will continue to use restricted source in- formants. Also pursuant to our suggestion, the proposed guidelines were revised to provide that IRS, under no circumstances, will take any action to conceal a crire committed by one of its informants and to establish specific procedures for employees to follow when they suspect, or know, that an informant has committed a crime. If it appears that the informant illegally obtained in- formation, or committed a serious crime not related to infor- mation gathering, the employee who has knowledge of the vio- lation must notify the chief of intelligence. The chief must report the violation to the appropriate law enforcement agency. If this is inadvisable, he is to immediately notify the Director, Intelligence Division, of the facts and circumstances con- cerning the violation, including his recommendations on reporting the violation and on continuing to use the informant. After the Director reviews the matter, he is to notify the Deputy Commissioner and provide him with his comments on the 43 chief's recommendations. The Deputy Commissioner is then to determine -- when the appropriate law enforcement authorities shall be notified of the violation, -- what to do with the information gathered through violation of the law, and -- whether IRS should continue to use the informant. In determining the advisability of notifying appropriate law enforcement authorities of an informant's criminal activity the responsible officials are supposed to consider -- whether the crime is completed, imminent, or in progress; -- the seriousness of the crime in terms of danger to life and property; -- whether the crime is a violation of a Federal or State law and whether it is a felony, misdemeanor, or lesser offense; -- the degree of certainty regarding the criminal activity; -- whether the appropriate authorities already know of the criminal activity and the informant's identify; and -- the significance of information provided or to be provided by the informant, and the effect on IRS' investigation if the other law enforcement agency is notified. If informants are to be paid from confidential funds, approval to expend those funds must be obtained from the dis- trict director, the regional commissioner, or the Assistant Commissioner for Compliance, depending on the amount involved (district directors up to $1,000, regional commissioners up to $2,000, and the Assistant Commissioner in excess of $2,000). The requests must be in writing and, to the extent possible, must show -- the name of the taxpayer or entity, --the case or project number if applicable, -- the specific amount being requested, --the location of the imprest fund to be used, 44 -- the proposed plan of action, -- the period over which the expenditures will be made, and -- a complete description of the purpose of the expenditures with a separate statement explaining proposed expenditures for purposes other than purchasing information. Accumulating, evaluating, and processing information Before March 1975, IRS procedures for accumulating, evaluating, and processing information obtained from infor- mants were weak. Specifically, the type of information IRS employees could gather using informants was not defined and systematic evaluations of potentially valuable information obtained from informants and removal and separate handling of non-tax-related information supplied by informants were not required. The new procedures are designed to correct these weaknesses. The guidelines define directly tax-related information and also include examples of information which would be con- sidered directly tax-related, such as -- personal expenditures or investments not commensurate with known income and assets; -- receipt of unreported income; -- overstatement of itemized deductions, business expenses, cost of sales, and tax credits; -- improper deduction of capital or personal and living expenses; -- failure to file required returns or pay tax due; -- omission of assets of improper deduction or exclusion of items from real estate and gift tax returns; and -- violation of conditions and requirements relating to tax exempt status of organizations. The guidelines also provide for the timely evaluation and processing of information received from informants. Special agents are prohibited from maintaining information in background files unless the information pertains to an assigned investi- gation or an authorized information gathering effort. The 45 guidelines also prescribe procedures for evaluating and processing any information received from informants that does not meet the criteria for inclusion in background files. The guidelines require various reviews and evaluations of the information gathered. A special agent must periodically review information gathered on each taxpayer to determine if he should (1) recommend to the chief of intelligence that an investigation be opened on a taxpayer or (2) submit the information, no longer of interest to the Intelligence Division, to the chief to be purged from the files. The chief or assistant chief of intelligence must personally conduct a quarterly review of samples of infor- mation gathered to insure that only directly tax-related information is being retained. The reviewing official must forward a record of the review to the district director. Managenent involvement in informant activities The new guidelines require more management review of activities directly related to the use of informants than did the procedures in effect before March 1975. In addition Internal Audit has broadened its reviews of Intelligence Division programs and its audits of confidential imprest funds to concentrate more on informant activities. Management reviews The new guidelines require managers at all levels to re- view the information IRS employees are gathering and retaining. District directors must provide for quarterly reviews of all authorized information gathering efforts to insure compliance with policy and procedures. Regional commissioners are re- quired to conduct semiannual reviews of each district's infor- mation index system to determine that only directly tax-related information is entered and that unnecessary information is de- stroyed or retired. The Assistant Commissioner for Compliance must include an annual review of information gathering activi- ties as part of the National Office Review Program. At our suggestion, the new guidelines on informants were revised to require assistant commissioners and regional com- missioners to examine informant activities during regularly scheduled reviews to insure compliance with IRS policy and procedures. Each regional commissioner and district director is re- sponsible for closely monitoring and accounting for the 46 expenditure of confidential funds. Each regional commissioner or his delegate is to maintain a cumulative file of confi- dential expenditures, by district, for each assigned case or authorized information gathering effort to insure that authorized amounts are not exceeded. During their regularly scheduled visits to regions and districts, assistant commis- sioners and regional commissioners are required to review confidential expenditures to ascertain compliance with the guidelines. Within 30 days after the end of each fiscal year, each chief of intelligence must prepare a report summarizing the cost and results of transactions with all informants. The report is to include data on the number of information items for which informants were the source, cases opened as a result of information received from them, and the status of those cases. The district director must concur in the report and the assistant regional commissioner for intelligence must provide a copy of the report to the regional commissioner and the Director, Intelligence Division. This is the first time IRS has required such a report. It should greatly assist IRS management in evaluating and quantifying the benefits of using informants. Internal audit coverage IRS' Internal Audit Division has revised its audit plans to concentrate more on informant activitie's. In auditing confidential imprest funds, internal auditors will (1) test the fiscal controls over the transfer and custody of funds and the payments made to informants, (2) verify the existence of paid informants, (3) determine what information they pro- vided for the payment, and (4) determine how that information relates to the administration of the tax laws. Internal audit officials in the southeast region told us that they intend to include evaluations of informant activities in their revievs of intelligence programs such as those relating to narcotics trafficking and Strike Force activities. In line with this, the internal auditors told us that they would review specific information gathering efforts in these programs and determine, among other things, (1) the type of information informants were providing, (2) how they were being used, and (3) the contributions they were making to tax law administra- tion. The new guidelines specifically provide the internal audi- tors access to informants' names which will permit them to make some of the tests necessary for a full review of a district's transactions with informants. 47 CONCLUSIONS IRS has taken some important steps to strengthen its controls over the use of confidential informants. But guide- lines it has issued in the past 2 years do not adequately provide for management assessment of potential benefits and risks before an informant's use is authorized. Under the new guidelines, primary responsibility for authorizing the use of an informant rests with the chief of intelligence. Other district, regional, and national office managers do not have to get involved unless confi- dential funds are to be expended. The involvement of these other managers, however, is an exercise of fiscal responsi- bility, Although they may ask questions about why and how an informant is going to be used, their primary purpose is to authorize the expenditure of confidential funds, not to authorize the use of an informant. IRS should distinguish between two types of informants-- those who provide information to IRS on their own initiative and those who gather information at IRS' request or encourage- ment--and should prescribe controls accordingly. The failure to clearly distinguish between types of informants results in too much control in some instances and not enough in others. Under the new guidelines, for example, a chief of intelligence can approve the use of an unpaid in- formant who is going to be gathering information at IRS' request or encouragement. Yet, under the procedures on confidential expenditures, the chief must obtain the district director's approval to pay even a few dollars to a person who voluntarily walks into an IRS office with valuable tax-related information. Controls over money are important and should not be de- emphasized, but controls over informant activities should be based primarily on how the informant will be used and not on the amount of money that will be spent because the risks as- sociated with using an informant do not vary depending on the dollars involved. Because informants who gather information at IRS' request or encouragement present special risks, the controls over them should be greater than those over informants who furnish information on their own initiative. Thus, the approval to use such an informant, whether he is to be paid or not, should come at a high enough level to permit a thor- ough and objective assessment of what might be gained in using the informant relative to what might be lost. In this regard, before the Intelligence Division can use a highly trained special agent to gather information in an undercover capacity, approval has to be obtained from the 48 chief of intelligence, the district director, the assistant regional commissioner for intelligence, and the Director of the Intelligence Division at the national office. Yet, IRS gives the chief of intelligence sole responsibility for authorizing the use of a less reliable informant to do the same thing. We find that inappropriate. We believe that a regional commissioner is in a better position than a chief of intelligence to objectively weigh the need for an informant against the potential risks in using him. A chief of intelligence is primarily concerned with IRS' intelligence activity which could easily affect his decision to authorize use of an informant. A regional commissioner is responsible for and concerned with all IRS activities and thus would be better qualified to assess the benefits and risks of using an informant in the context of IRS' overall mission. We recognize that district directors are also responsible for and concerned with all IRS activities and that they, too, could be expected to objectively assess the need for an in- formant. We believe that a regional commissioner should make this assessment, however, because (1) he is more detached from the operating level and thus less likely to base his decision solely on a desire to achieve results and (2) it would provide for more uniformity. Instead of 58 district directors making decisions, there would be just 7 regional commissioners. It is important that the approving official have suffic- ient information to properly assess the need for an informant and the potential benefits and risks. Accordingly, any request for approval to use an informant to gather information at IRS' request or encouragement should state why the informant is needed, how he was identified and determined reliable, and how he will be used. IRS guidelines do not require this. The new guidelines require (1) periodic evaluations of confidential informant activities, (2) a summary report on the results of using informants, (3) a review of information collected, and (4) close monitoring of confidential funds. The guidelines pinpoint managers who will make the evaluations and show how the evaluations will be made and what will be covered and reported on. But revised guidelines alone are not the answer. Past guidelines often were not complied with and there can be no assurance that the revised guidelines will be unless manage- ment pays more attention to these activities than it has in the past. 49 Intelligence officials at the national, regional, and district offices tended to ignore or condone potential prob- lems and weaknesses by not requiring compliance with existing guidelines and by not responding to problems as they arose. IRS managers outside the intelligence function appeared to have a "hands off" philosophy regarding informant activ- ities. This attitude that intelligence personnel should be left to manage their own affairs appeared pervasive and generally accepted. The lack of management attention at all levels was the underlying cause of the problems discussed in this report. It may have arisen because management failed to fully appre- ciate the problems that can arise when dealing with informants but, whatever the reason, it must change. RECOMMENDATIONS We proposed that the Commissioner of Internal Revenue revise IRS' guidelines on informants to require -- that regional commissioners authorize the use of any informant who will be gathering information at IRS' request or encouragement after determining that (1) there is reasonable cause to believe a tax law has been violated, (2) there is no practical alternative for obtaining essential information, and (3) there is a specific limitation on the time and scope of the informant's activities; and -- that requests to use each such informant show, among other things, why the informant is needed, how he was developed and determined reliable, and how he will be used. Also, to encourage proper implementation of these guidelines and adequate management attention to informant activities, we proposed that the Commissioner review IRS' informant activities at least annually. One way this could be accomplished is through an independent third party like IRS' Internal Audit Division. 50 IRS COMMENTS AND OUR EVALUATI", IRS did not agree that regional commissioners should authorize the use of informants. In commenting on our proposals, the Commissioner of Internal Revenue said: "We believe that this authority should be maintained within the District. The District Director should have the authority to authorize the use of informants, as this is the management official responsible for all of the IRS activities within a district. The control of the use of informants by the District Director is consistent with this overall responsibility. Our current procedures will be revised to include this provision and will be issued within a month." "We feel that our current procedures with the modification for the authorization to approve the use of informants will be adequate and proper and should be given a chance to work. Therefore, we would welcome GAO to return and review our procedures after they have been in effect for a year or so." "The Deputy Commissioner is issuing a memorandum to all Regional Commissioners advising them that our procedures are being amended to require District Dir- ector approval on the use of informants, with the pro- vision that this authority may not be redelegated. In this memorandum the Deputy Commissioner will emphasize the need for close adherence to our guidelines in this sensitive area and the necessity for close management monitoring. Also, the Assistant Commissioner (Compliance) has requested the Assistant Commissioner (Inspection) to give high priority to this activity in Internal Audit program reviews and to require an audit of this area in the Internal Audit conducted of every District Intel- ligence Division. In addition, each Regional Commis- sioner will review this area as a part of his management responsibilities of each district office. The National Office Review Program, which evaluates the effectiveness of each region, also will continue to follow-up and review this program area." "We feel that the combination of [Internal Audit] reviews and our National Office and reginnal reviews will ensure that the procedures relating to the use of informants will be properly followed by our field officials. We also feel that such reviews will compel the involvement of managers with regard to the control and use of informants, as well as to the value of the 51 results obtained from the use of informants, which may have been lacking in the past." Although we believe that regional commissioners are in a better position to authorize the use of an informant, IRS' intention to give that authority to the district directors is not inconsistent with the main thrust of our proposal--to get more management involvement from outside the intelligence function before an informant is sent out to gather informa- tion. Accordingly, we cnnsider IRS' position acceptable, on a trial basis, IRS said also that -- it does distinguish between informants who provide information on their own initiative and those who provide information at IRS' request or encouragement and -- information on why the informant is needed, how he was developed and determined reliable, and how he will be used is already called for in its guidelines. Although IRS does recognize the distinction between the two types of informants, its controls do not adequately flect that distinction. re- As we said before, a chief of intel- ligence can approve the use of an unpaid informant who is going to be gathering information at IRS' request or encouragement, but must obtain the district director's approval before even a few dollars to a person who voluntarily walks paying into an IRS office with valuable tax-related information. With regard to the latter, IRS noted that "Cases in which informants pro- vide IRS with information on their own initiative generally involve a one-time, non-paid situation * * *." However, as indicated by IRS' use of the word "generally," there is nothing to preclude such an informant from being paid. In concluding that IRS should distinguish between types of informants and establish controls accordingly, two we were prima ly influenced by the fact that IRS' existing con- trols are money oriented. In establishing controls, IRS seemed more concerned with whether or not an informant was going to be paid than it was with how the informant was go- ing to be used. Indeed the level of management involvement is predicated on the amount of money to be expended. Our purpose was to redirect IRS' concerns. In dealing with an informant, the important question is not "How much are we going to pay him?" but rather "What is the informant going to do and how?" 52 Likewise, although IRS' current guidelines do call for some of the information we ask for, they do not call for all of it. Specifically, as noted by IRS, the guidelines do re- quire the chief of intelligence to consider the following factors in evaluating the potential use of an informant: available background information on the informant, such as his address, criminal record, and associates; the reliability of information provided to IRS and to other agencies, if known; and the informants' source and means of securing information. The guidelines do not require the chief to consider why the informant is needed and how he will be used--information that we think is vital before making any decision to use an infor- mant. In its response, IRS also referred to its guidelines on confidential expenditures. Although, ; discussed earlier, these guidelines do call for including certain information, such as the proposed plan of action, in requestrin- author; zation to expend confidential funds, they do not specifJ_dlly call for all the information we think necessary. It should be remembered also that these guidelines apply to approving the expenditure of confidential funds; they do not apply to authorizing the use of an informant. In summation, therefore, IPS' guidelines do not, in total, provide the control that we are striving for. The most important element of that control--the involvement of someone outside the intelligence function in deciding whether to use an informant--is absent from the guidelines. Because the guidelines are already confusing, we felt it would be easier to recommend a total control package rather than try to recommend piecemeal changes to existing guidelines. IRS agreed that Internal Audit should review informant activities but noted that this would be nothing new. We agree that Internal Audit has always been responsible for reviewing these activities although, as we said, it had not done as much as it should have. Our recommendation, however, is not directed at Internal Audit; it is directed toward the Commissioner. IRS' new guidelines and its response to our report call for increased management involvement in approving and review- ing informant activities. This increased involvement is important and should lead to better control over the use of informants provided all levels of management are committed to the need for such involvement. Without this commitment, review and approval of informant activities can easily deteriorate into a "rubber stamping" exercise. To guard against this, oversight by the Commissioner is essential. 53 We are giving the Commissioner the option of using Internal Audit, or any other third party, to assist him in providing that oversight. If the Commissioner were to choose Internal Audit, we would anticipate his receiv- ing a report, at least annually, on the scope and results of its reviews and taking any action necessary to correct management deficiencies. While Internal Audit's responsi- bility for reviewing these activies would be nothing new, its formalized reporting to the Commissioner would be. 54 APPENDIX I APPENDIX I Internal Revenue Service Department of the Treasury Commissioner Washington, DC 20224 JUL 2; > Mr. Victor L. Lowe Director, General Government Division United States General Accounting Office Washington, D. C. 20548 Dear Mr. Lowe: We appreciate the opportunity to review your recently prepared draft report entitled "IRS Management Controls Over the Use of Confidential Informants." The findings and recommendations have been given much thoughtful consideration, and we have had--as you suggested in your letter requesting our review--informal sessions with your staff to work out technical issues. As a result, agreement on certain of our comments has already been reached and the final report will reflect suggested changes deemed appropriate by your staff. I will not repeat these agreed upon changes here. Rather, I would like to add the following comments documenting our general response to i sy recommendations in the report and raising issues that have not been resolved in our informal talks. (1) We do not agree with the recommendation that the Regional Commissioner authorize the use of informants. We believe that this authority should be maintained within the District. The District Director should have the authority to authorize the use of informants, as this is the management official responsible for all of the IRS activities within a district. The control of the use of informants by the District Director is consistent with this overall responsibility. Our current procedures will be revised to include this provision and will be issued within a month. We feel that our current procedures with the modification for the authorization to approve the use of informants will be adequate and proper and should be given a chance to work. Therefore, we would welcome GAO to return and review our procedures after they have been in effect for a year or so. (2) The draft GAO report recommended on page 77 that all requests to use informants state "...why the informant is needed, how the informant was developed and determined reliable, 55 APPENDIX I APPENDIX I -2- Mr. Victor L. Lowe and how the information will be used." IRS Manual Supplement 9G-40, Intelligence Division Guidelines on Informants, dated February 3, 1977, provides that the Chief, Intelligence Division, in evaluating the potential use of an informant, will consider such factors as: (1) available background information on the informant, such as the address, criminal record and associates; the reliability of information provided to the Service and to other agencies if known; and the informant's source and means of securing information. In addition, the Chief, Intelligence, has available to him/her as part of his/her review the special agent who contacted or was contacted by the informant as well as the information item prepared by the special agent which would provide the specific information furnished by the informant. The same sources can be made available to the District Director for his evaluation of the use of informants. Also, the procedures in our Manual Supplement 9G-24, Confidential Expenditures, dated December 10, 1975, pertaining to the payment of informants specify that the following material be developed in a written request to the authorizing official: (1) the name of the taxpayer or entity; (2) the case or project number when applicable; (3) the specific amounts being requested; (4) the location of the imprest fund to be utilized; (5) the proposed plan of action; (6) the period of time over which the expenditure(s) u:11 be made; and (7) a complete descritpion of the purpose of the expenditure(s) for laying the groundwork for procurement of information. The description of the purpose should also set forth any tentative arrangements and amounts previously discussed with the informant. We feel that the procedures in Manual Supplements 9G-24 and 9G-40 as described above cover all the points recommended by GAO. (3) We agree with the concept expressed in the GAO recommendation th'at the Internal Audit Division should review our informant activities to provide an independent view of how we are following our prescribed written procedures. However, this would not be a new responsibility for Internal Audit since they have been and will continue to be involved in a review of informant activities (see comment n~. 7 below). We feel that the combination of their reviews and our National Office and regional reviews will ensure that the procedures relating to the use of informpets will be properly followed by our field officials. We also feel that such reviews will compel the involvement of managers with regard to the control and use of informants, as well as to the value of the results obtained from the use of informants, which may have been lacking in the past. (4) The GAO report contends that, "IRS did no. ilave a workable system or procedure for evaluating and assu.ing effective use 56 APPENDIX I APPENDIX I -3- Mr. Victor L. Lowe of the information received from informants." We strongly disagree with this statement. Until March, 1976, it had been a long-standing procedure that when the Intelligence Division received an information item, it would be evaluated by the Chief, Intelligence Division, or his/her delegate for possible criminal potential. If the information item did not have criminal potential, but did have Audit or Collection potential, the Chief, Intelligence, would forward the information item to the appropriate function for evaluation. In March, 1976, the evaluation process was transferred to the Service Centers. We believe that the problems that GAO discovered were related to the handling of information items generated from long-term informants, which were not prepared and evaluated as expeditiously as would have been expected. However, this was not completely a problem with the procedures, but rather in the fact that a few Chiefs of the Intelligence Division did not follow existing procedures. Latest guidelines outlined in our Intelligence Divisions Guidelines on Informants, dated Febrtary 3, 1977, are designed to guard against the possibility that this could happen again. (5) With regard to a further GAO recommendation that IRS distinguish between two types of informants, those who furnish information on their own initiative and those who gather information at IRS' request or encouragement, we again feel that our procedures in Manual Supplement 9G-40, Intelligence Division Guidelines on Informants, already make this distinction. Section 3 of the Supplement provides procedures for the use, control and evaluation of those informants which the Service is encouraging or requesting to gather information. Cases in which informants provide IRS with information on their own Initiative generally involve a one-time, non-paid situation in which the informant provides the information either in person, by telephone or by letter. For this type of informant, we feel that the only procedures necessary relate to the processing of the information received from them, and control procedures for this are found in Section 6 of the Supplement. Under these procedures, the Intelligence Division would either engage in information gathering activities pertaining to the particular taxpayer(s) reported by the informant or input the information into either the Centralized Information Item Processing System or the Information Indexing System. We feel that these procedures are adequate and furthet clarification is, therefore, not necessary. (6) We also propose two language changes which we feel are necessary to accurately reflect the scope of GAO's review. The first change relates to the first paragraph of the cover summary. 57 APPENDIX I APPENDIX I -4- Mr. Victor L. Lowe We suggest the following language (our changes are in parenthesis): "Primarily because of inadequate management attention, IRS had not done all it could to maximize the benefits and minimize risks in dealing with (long-term paid confidential) informants. (Such) informants were (sometimes) used in ill-defined and overly broad intelligence gathering efforts; inadequate procedures existed for evaluating information received from such informants; and the use of these informants was not subjected to systematic management review." Our other language change relates to page 8 See GAO of the GAO report, second paragraph, under the sec'-ion "Review Note 1, Objectives and Scope." We suggest that this para,raph be changed to read as follows: "To do the work, we reviewer the procedures and practices followed by the National Office, two regional offices and five district offices for initiating and carrying out intelligence activities in which confidential informants had major roles. (As stated earlier, the Intelligence Division paid 2,000 confidential informants about $800,000 during the fiscal years 1973 through 1975. For this same period, the two regions we reviewed had 662 confidential paid informants, who received $471,000. This represented 33 1/3% of the total informants and about 59% of the total monies paid the informants. However, our in-depth review was limited to 52 confidential paid informants.") (7) The GAO draft report suggests that the Internal Audit Division can provide an independent third-party review of confidential informants (page V of the Digest and 77 of the Report). Thus, GAO leaves an implication that Internal Audit was not involved in independent third-party reviews prior to the Operation Leprechaun review. This implication is incorrect and distorts the report. Internal Audit historically has provided coverage to the controls on use of confidential informants. The Deputy Commissioner is issuing a memorandum to all Regional Commissioners advising them that our procedures are being amended to require District Director approval on the use of informants, with the provision that this authority may not be redelegated. In this memorandum the Deputy Commissioner will emphasize the need for close adherence to our guidelines in this sensitive area and the necessity for close management monitoring. Also, the Assistant Commissioner (Compliance) has requested the Assistant Commissioner (Inspection) to give high priority to this activity in Internal Audit program reviews and to require an audit of this area in the Internal Audit conducted of every District Intelligence Division. In addition, each Rogional Commissioner will review this area as a part of his management responsibilities of each district office. The National Office Review Program, which evaluates the effectiveness of each region, also will continue to follow-up and review this program area. 58 APPENDIX I APPENDIX I -5- Mr. Victor L. Lowe I hope these comments will be useful to you in developing your final report. We will, of course, be happy to discuss any of these issues with you or your staff if further clarification is necessary. With kind regards, /-.? Sincerely, GAO notes: 1. We have added most of this suggested language to our report. Some of the language was not added because it was either unnecessary or incorrect. As stated on p. 13 we reviewed 52 requests for approval to expend confidential funds on intelligence gathering efforts involving informants. However, our re- view was not limited to 52 informants. Some of the efforts involved more than oale informant. We believe that our scoDe which was augmented by the work Internal Audit did, was sufficient to enable us to draw conclusions on the Intelligence Division's use of confidential informants. 2. Page references in IRS' comments may not correspond to pages in the final report. 59 APPENDIX II APPENDIX II OPERATION LEPRECHAUN In early 1975 a series of articles appeared in Miami, Florida, newspapers which led to considerable interest and controversy regarding an IRS intelligence gathering effort known as Operation Leprechaun. The articles alleged that IRS was paying undercover agents and informants to gather data on the sex lives and drinking habits of Federal judges and other public officials in Miami. The newspaper disclosures were followed by a sequence of events in IRS which included -- the immediate revocation of IRS' guidelines on confidential expenditures and the requirement that all such expenditures be approved at the national office, -- a detailed examination of Operation Leprechaun by the Internal Audit Division, -- investigations of 25 IRS employees by the Internal Security Division, 1/ -- a serious rift among IRS officials and employees over the merits and faults of the intelligence gather- ing effort, -- the departure from IRS of several key personnel associated with the effort, and --major changes in guidelines affecting practically all aspects of Intelligence Division operations. After the newspaper stories, several congressional committees and subcommittees held hearings on Operation Leprechaun and IRS' intelligence gathering efforts in general. (See app. IV.) A Federal grand jury also inquired into Operation Leprechaun and the allegations surrounding it. We reviewed Operation Leprechaun to determine the nature of the effort and why it was undertaken, the extent of manage- 1/As of March 1977, IRS had closed all but one of these inves- tigations and had taken no administrative action against any employee for misconduct arising out of Operation Leprechaun. 60 APPENDIX II APPENDIX II ment's involvement in the effort, and the type of information gathered and its usefulness to IRS in enforcing the tax laws. We also wanted to know whether IRS' management of this effort differed markedly from its management of other efforts we reviewed. WHAT WAS OPERATION LEPRECHAUN AND WHY WAS IT UNDERTAKEN? Operation Leprechaun was not directed toward the sexual and drinking activities of taxpayers. Rather, the effort, which was approved at IRS' Jacksonville district, southeast region, and national office, was directed at organized crime and alleged corruption among public officials in the Miami area. Operation Leprechaun was not an official IRS designation. Until the newspaper allegations in early 1975, the words had little meaning to anyone except one special agent in the Miami office of the Jacksonville district. The special agent coined the term to refer to an intelligence gathering effort assigned to him from May 1972 until September 1973. In September, ad- ditional special agents were assigned to the effort, and the special agent initially in charge stopped using the term. For ease of presentation, we will also refer to this intelligence gathering effort as Operation Leprechaun. Why_Operation Leprechaun was initiated Operation Leprechaun was initiated because certain Fed- eral and local officials strongly believed that political and judicial corruption was deep seated and widespread in Miami and because certain Department of Justice and IRS officials wanted IRS to become more involved in the Strike Force program. IRS national office and Department of Justice personnel were concerned about the decline in IRS' participation in the Strike Force program. IRS intelligence officials in the na- tional office were dissatisfied with IRS Strike Force efforts in Miami and wanted to use undercover agents and informants to develop information on organized crime in that area. De- partment of Justice officials were also concerned about the drop in IRS staff-years devoted to the Strike Force proqram nationwide. For example, at a Strike Force representatives' 61 APPENDIX II APPENDIX II seminar in November 1972 attended by officials from the Department of Justice, the Department of the Treasury, and IRS, the Assistant Attorney General made the following comments. "The basic purpose for setting up a Strike Force program was that we wanted to bring into our concept your [IRS'] knowledge of the system in which you operate. To date I would qrade the IRS 90 plus for its cooperation. "More and more we will be choosing sophisticated type individuals as targets. "What concerns me now is the decrease of IRS manpower to combat organized crime. I strongly suggest your consideration in manpower. There is so much work to be done." At that meeting, the Chief of the Organized Crime and Racketeering Section in the Department of Justice, commented: "In the Jastice Department we consider and the Audit and Intelligence Divisions of IRS recognize as the strong arms in the fight against crime. IRS has accounted for approximately 60% of our criminal prosecutions. "One of the most recent concerning factors Strike Force program is the reduction in in the number of IRS examinations and investigations pending. Man-years expended are also on a downward trend. Most importantly, we need to impress the district offices for the needed manpower to identify new cases." The Director of the Intelliqence Division emphasized the importance of continuing to improve IRS performance in the Strike Force program, and stated: "In Strike Force we need new ideas, new approaches in case development and investigations, and we need to provide leadership in district offices. 62 APPENDIX II APPENDIX II "District offices have a responsibility to buy informa- tion and conduct surveillance. Greater emphasis should be placed on the development and use of informants." In addition, the Department of Justice attorney acting as Chief of the Miami Strike Force at the time of Operation Leprechaun encouraged IRS' Jacksonville district to become more actively involved in the Strike Force. In a June 1972 Strike Force meeting attended by Justice and IRS personnel, the Chief stated that corruption of local public officials would be emphasized and estimated that "* * * at least 50 percent of all judges and commissioners are corrupt." In that meeting the Chief described the role he wanted IRS to play in gathering information on certain specific individuals. Authorization to expend confidential funds Local law enforcement agencies, in cooperation with the Miami Strik? Force, had been probing political and judicial corruption in Miami. The Dade County Department of Public Safety and the Miami Police Department had conducted joint surveillance which produced information indicating that public officials were taking bribes. During the probe the Chief of the Miami Strike Force held several discussions 'with the Chief of Intelligence in the Jacksonville district and the Assistant Regional Commis- sioner for Intelligence in the southeast region regarding corruption among local public officials. In one meeting, the Chief of the Strike Force provided the Chief of Intel- ligence with the names of 15 persons identified during their surveillance, and the Chief of Intelligence agreed to initiate an intelligence gathering effort. On Mar-h 30, 1972, the Chief of Intelligence requested the Assistant Regional Commissioner for Intelligence to approve confidential expenditures to develop intelligence through informants. The request stated that (1) an acute need existed to finance the covert gathering of intelligence because Miami had become "* * * a watering hole for racketeers of national notoriety as well as a spawning ground for local corruption," (2) Strike Force efforts represented a significant portion of the investigative effort of the Intelligence Division in 63 APPENDIX II APPENDIX II Miami, and (3) the Chief of the Strike Force had identified loan sharking, extortion, gambling, labor racketeering, and corruption of law enforcement and judicial officials, at all levels, to be areas of particular concern. The request also stated that: "Intelligence gathering via the usual techniques provides only surface information of little real value. Meaningful intelligence can only be obtained from inside the organizations of interest by the use of paid informants. While this latter technique has been utilized with a great deal of success on a case- by-case basis, we now most urgently need to utilize this technique in the gathering of intelligence con- cerning the activities of organized criminals and others in the areas of concern previously enumerated." The Assistant Regional Commissioner for Intelligence con- curred in the request and estimated $30,800 would be needed to finance the effort through June 30, 1973. Because the request exceeded the $5,000 limit for regional approval, the Assistant Regional Commissioner for Intelligence sent it to the national office and noted that the effort would include the services of three to five undercover agents and five to seven local informants. The Acting Director of In- telligence approved the request on April 25, 1972. In February 1973, after the Assistant Chief of Intelli- gence had evaluated the intelligence gathering effort, the Chief of Intelligence in Jacksonville asked the Assistant Regional Commissioner for Intelligence to approve an additional $17,000 in confidential expenditures to continue the effort through June 30, 1973. The Chief stated that a joint investi- gation by local enforcement agencies and the Miami Strike Force disclosed widespread corruption of public officials, particularly criminal court judges, which made it imperative to continue the effort. This request was forwarded to the Director of Intelligence who approved it on February 28, 1973. In May 1973, the Group Manager recommended to the Chief of Intelligence that confidential funds of $50,000 be author- ized to continue the intelligence gathering effort through 64 APPENDIX II APPENDIX II June 30, 1974. The Chief sent a request for confidential funds of $50,000 through the Assistant Regional Commissioner for Intelligence to the Director of Intelligence. The Director granted an interim authorization of $10,000 and requested the Jacksonville district to submit a plan for confidential expenditures. The Chief submitted a plan in October 1973 to probe loan sharking, extortion, gambling, labor racketeering, and corruption in order to identify "the extent of national participation of organized crime." He estimated confidential expenditures of $51,500 would be required and stated that undercover agents and ten informants or operatives, selectively placed, would be used. He also stated: "Space rentals and utilities will be necessary in order to conduct observations, to meet confidential informants or undercover agents periodically and in one situation, to set up a business front.. Rental autos will be used to meet informants and for other intelligence activities." In late 1973 and early 1974, national and regional in- telligence officials reviewed confidential expenditures and information gathered in Operation Leprechaun. In February 1974, the Director authorized an increase in confidential funds from $10,000 to $40,000 for use through June 30, 1974. INADEQUATE STAFFING AND EVALUATION OF OPERATION LEPRECHAUN Not long after Operation Leprechaun was initiated, problems began to arise and the overall effort began to deteriorate because: -- The Intelligence Division had not adequately staffed the effort, but instead had left almost everything to the discretion of one special agent. -- Management oversight and control was inadequate during most of the special agent's 16-month in- telligence gathering effort even though the scope of the effort was greatly expanded and a number of problems surfaced during the 16 months. 65 APPENDIX II APPENDIX II -- The information accumulated by the special agent during the effort was not systematically evaluated to determine its tax-relatedness and usefulness. Placing one special agent in charge of a broad intelli- gence gathering effort and not exercising reasonable manage- ment control over his activities or the information he gath- ered created unnecessary risks and hampered IRS' chances of maximizing the effort's contribution to the Intelligence Division's enforcement mission--a problem that was not unique to Operation Leprechaun. Inadequate staffing After confidential expenditures were first authorized in April 1972, the Intelligence Division assigned a special agent to establish a network of informants and left him in charge of setting up the operation, identifying targets, and determining sources and techniques to be used. He was given the broad mission of gathering intelligence on alleged political and judicial corruption in Miami but was given little guidance or assistance in setting up and carrying out the effort. The then Chief of Intelligence told us that he selected the special agent for the assignment because the agent had been successful in two previous assignments and was in contact with informants in Miami. The agent was furnished the names of 12 to 15 persons identified by the Chief of the Miami Strike Force and was instructed to develop intelligence files on them. As the intelligence gathering effort continued, however, the agent began to accumulate information involving hundreds of individuals. For example, he prepared 594 records of con- tact with informants, containing an average of about 10 names each. It is not clear how many of the persons named in those documents were targets of intelligence gathering because no formal process existed for identifying and approving targets. Instead, the special agent identified individuals as targets based on (1) his knowledge of suspected tax violators and his experience over the years, (2) information provided by other law enforcement agencies, (3) the intelligence gathering effort itself, and (4) newspaper stories about organized crime figures. The special agent relied primarily on confidential infor- mants to gather information on targets. In September 1972, 66 APPENDIX II APPENDIX II about 4 months after Operation Leprechaun's inception, the special agent sent a memorandum through his group manager to the Chief of Intelligence evaluating his network of informants. The Assistant Chief of Intelligence had requested the evalu- ation because during a September 6, 1972, Strike Force meeting attended by IRS regional and district officials, the special agent had asked for help in handling informants. In his September 1972 evaluation, the special agent said the primary purpose of Operation Leprechaun was "to infiltrate and subvert organized crime and corruption to the end that information, leads, and evidence are obtained peculiar to criminal tax investigations." The evaluation outlined the procedures the agent followed in establishing an informant network which included: 1. Identifying potential informants in the areas of interest. 2. Recruiting and training informants. 3. Obtaining personal background information on each informant. 4. Directing informants and receiving information from them. 5. Verifying informants' allegations and information. 6. Disseminating informat on to other agents and to the information gathering and retrieval system. 7. Periodically reviewing informants' files and planning future strategy. 8. Identifying employment opportunities for informants in the areas of interest. 9. Paying informants. 10. Coordinating with the Strike Force and other agencies. The agent began recruiting informants by asking other IRS special agents and representatives of other law enforcement agencies and State agencies about individuals who might assist 67 APPENDIX II APPENDIX IT in information gathering. Some informants used in Operation Leprechaun were recruited in his previous work, some were re- cruited through his personal contacts, and some were recruited by other informants at the direction of the special agent. According to the special agent, the confidential informants came from all ethnic, social, professional, and occupational ba-kgrounds and some had criminal records. In his memorandum the special agent described the scope of his intelligence gathering, the risks involved, and some of the problems he was having as follows: "There are presently 34 informants within OPERATION LEPRECHAUN, but only a handful are paid. There are several informants which I nave never met. They were developed by other informants. In other words, o help offset limited time and manpower, I use informants to instruct and even to pay other informants; to conduct surveillances on suspects arid other informants; and to conducL background ar.nd public records checks on suspects. Admittedly, this it,not orthodox and is a risky proposition." * * * ** "In my opinion the assignment to set up and operate a network of paid informants is next to impossible for one man to manage on a continuing basis in an effective and efficient manner. However, I strongly believe in tho need and propriety of such an undertaking. Along with it is a need for office facilities separate from the general Internal Relenue Offices, a full group of Special Agents with Supervisor and Clerical Staff, various investigative equipment (including electronic devices), and financing. There is no doubt in my mind that_a group such as this would be hJ_il __successful if properly fiaancta, staffed, and_equpped." ineffective man:agement control IiS officials outside the intelligence .nction -ere not kept informed of. developments and therefore did not exercise any control over Operation Leprechaun. IRS in'elligence of- ficials at all levels, however, were maade await from time to 68 APPENDIX II APPENDIX II time of the nature, scope, problems, and risks of Operation Leprechaun but did nothing to change the situation. Management oversight The Chief of Intelligence in the Jacksonville district, the Assistant Regional Commissioner for Intelligence in the southeast region, the Director of Intelligence in the national office, and others at each management level were generally aware of (1) what the nature and purpose of Operation Lepre- chaun was, (2) who the targets were, (3) how the information was being gathered, and (4) what type information was being gathered. IRS intelligence officials in the southeast region and the Jacksonville district were informed of the special agert's activities through requests for authority to expend confi- dential funds; memoranda, progress reports, and project eval- uations prepared by the special agent conversations with the special agent's group manager; and regional and district evaluations. In his September 1972 memorandum, for example, the special agent specifically described the problems he wasWe having with informants and emphasized the risks involved. found no evidence that, as a result of the September 1972 evaluation, IRS intelligence officials in the southeast region or Jacksonville district took steps to change the way in which the special agent was carrying out the operation and, indeed, the agent continued to operate in the same way over the next year. In a January 1974 memorandum to the Director of Intelligenc? the Assistant Regional Commissioner for Intelligence summarized the type of information being gathered by the special agent. This review was requested by the Director to help him decid whether to continue expending confidential funds in the .ntel- ligence gathering effort. The report tu the Director included information similar to that reported oy the special agent in his Septemiber 1972 evaluation, such as: -- Informants were being used to instruct other informants. -- Informants were being used to conduct surveillances and to perform background and public record checks. 69 APPENDIX II APPENDIX II -- Informants had been placed on the payrolls of suspects and in other positions close to either the suspects or their confidants and aides. Included in the report were the results of a detailed examination of information obtained from three informants who received a substantial part of the funds expended during Operation Leprechaun. The report contained references to the type of information obtained and stated that the infor- mants -- recruited other informants and performed surveillances on and took pictures of subjects, -- identified a subject's girlfriend, -- infiltrated a major pornography dealer's operation and recruited his manager, -- were socially close to the targets, -- determined the subject owned three houses/apartments and kept girlfriends in two of them while he rented out the third, and -- determined a target took a trip to Santa Domingo in 1971 to join a girlfriend. After receiving the report in late January 1974, the Director expressed serious concern over the way the special agent had used informants. He advised other Intelligenzce Division officials that either instructions were unclear or district intelligence personnel did not accurately understand the purpose of paying informants. He also strongly empha- sized that information gathered must be tax related and that payments should not be made for information relating pri- marily to violations under the jurisdiction of other law enforcement agencies. He said that guidelines had to be clarified and controls made more effective, and he requested a review of the matter. In response, Intelligence Diviscn personnel stated that a revised policy statement on confidential expenditures was being reviewed and that new guidelines would be prepared. The policy and guidelines were not changed, however, until 1975, after the Operation Leprechaun disclosures. In addi- 70 APPENDIX II APPENDIX II tion, the Director continued to authorize confidential ex- penditures for the Miami office including a $30,000 authori- zation on February 4, 1974. Supervision The special agent was assigned to IRS' Miami office under a group manager who was in charge of special agents gathering intelligence on and conducting investigations of taxpayers as part of IRS' efforts under the Strike Force program. Tie degree to which the group manager supervised the special agent in charge of Operation Leprechaun has been a matter of debate and disagreement, and questions have been raised about whether the special agent was actually under the direction and control of the Chief of Strike Force (a Department of Justice attorney) and the extent to which the agent reported to him. The special agent in charge of Operation Leprechaun received little direct supervison from the group manager. The special agent selected individuals as targets of intel- ligence gathering, chose informants, decided on intelligence gathering procedures and techniques and, to a large extent, controlled the payments made to informants and the information received. The group manager's role was limited and passive. His principal duties were to act as custodian of the funds and to process paperwork generated by the special agent, such as fund requests, expenditure documents, and memoranda con- taining information received fror informants. He did not direct or participate in the special agent's day-to-day operations such as informant contact, surveillances, and other "street work." In describing the group manager's role, the Chief of Intelligence stated that the group mana- ger approved the information gathered by the special agernt and dispersed funds to "ne agent to pay for the information. Arrangements made by the Chief of Intelligence may have restricted the group manager's supervision of the special agent's activities, but the facts are not clear as to the scope and duration of those arrangements. In a March 1975 affidavit furnished to IRS during the joint Internal Audit/ 71 APPENDIX II APPENDIX II Internal Security investigation of intelligence gathering activities in Jacksonville, the group manager stated: "It was [the Chief's] further direction that 'L., special agent] would consult directly and closely with [the Chief of Strike Force] about the corrup- tion in Dade County. [The Chief] advised me that I would only learn of [the special agent's] activi- ties on a need-to-know basis. [The Chief] asked me if T could work with [the special agent] under such an arrangement and I told him that I saw no problem. My responsibility would be to fund [the special agent's] activities concerning the payments to infor- mants; and to conduct the initial review of the efficacy of any such payments. "[The special agent] met with [the Chief of Strike Force] at the Miami Strike Force headauarters almost every day concerning 'Operation i .rechaun.' I have no knowledge of the specific nature of their meetings; however, I assumed it was to keep each other abreast of any current developments involving the individuals [the special agent] was developing information on." Similarly, in a July 1975 interview conducted as part of that same Internal Audit/Internal Security investi- gation, the group manager stated: "I was [the special agent's] supervisor with respect to 90 percent of the normal supervisory relationships, I would judge. * * * With respect to particular information gained from his contacts at the Strike Force, * * * and his contacts with our Division Chief, I did not enter into those, in the normal supervisory relationship ' * *." In February 1976, the group manager said that the Chief of Intelligence had told him in May 1972 that the special agent would be receiving highly sensitive infor- mation from the Department of Justice which was to be revealed to the group manager only ca a need-to-know basis. The group manager also said that a few days later the Chief of Intelligence told him that he was to know 72 APPENDIX II APPENDIX II everything about the special agent's work. In May 1976, the special agent told us that the Chief's original instructions were in effect only a short time. According to the special agent in charge of Operation Leprechaun, he visited the Strike Force offices in Miami to discuss his intelligence gathering activities, but only infrequently. The Chief of Strike Force said he did not have daily, or even weekly, contacts with the agent or supervise his intelligence gathering efforts. Because the meetings between the special agent and Strike Force attorneys were not documented, we could not determine the frequency of the meetings or whether the Strike Force attorneys directed or supervised the soecial agent's efforts. Information was not systematicall y evaluated and generallywas not related topossible tax violations The special agent in charge ot Operation Leprechaun was given the broad mission of gathering intelligence on political corruption and organized crime in the Miami area. He was not responsible for conducting investigations of specific individuals; this responsibility was assigned to other special agents who initiated investigations using informnation from Operation Leprechaun and ocher sources. To the special agent, qat.l- ing infurmltation tended to be an end in itself, btcause ie was not responsible fo' using it to initiate investigations or for other pur- poses. 73 APPENDIX II APPENDIX II At the time of Operation Leprechaun, IRS policy was to evaluate information, whenever practicable, before pay- ing foL it. As with other intelligence gathering efforts we reviewed, however, the special agenit did not always obtain and evaluate information before paying his infor- mants. The special agent used informants to gather infor- mation and many times made regular, periodic payments to them. In addition, some informants obtained information from and paid other informants for the special agent. These practices did not always permit the special agent to determine the value of information before paying for it. After he received information the special agent had to decide what to do with it. No procedure or system had been developed to control, process, and evaluate the infor- mation. In March 1974, at the group manager's request, the special agent developed local guidelines for administrative control of informants' information. The special agent sug- gested establishing a separate set of informant control files for each agent and the group manager. The agent stated: "* * * Because of the several different * * * posts of dut-y and the agent's need to continually evaluate his confidential sources, it becomes necessary for him to maintain a control file on each C.I. [confi- dential informant] * * * *." In describing the controls to be implemented by the group manager, the agent wrote: "Any intelligence organization faces a necessary .ind continuing flood of paper; an ever increasing need of clerical and technical personnel; irternal organizational problems; and building and maintain- ing a network of confidential sources. The added need for strong control becomes evident when (gents are dispersed throughout the state and each is workinn with paid C.I.'s. In my opinion, this need for st -g and effective control prevails over the nominal duplicity that is required in certain in- stances." 74 APPENDIX II APPENDIX II In March 1974, the group manager transmitted the special agent's suggestions to other agents in the group and directed that the controls be implemented immediately. Intelligence officials reviewed and evaluated informa- tion accumulated by the special agent only .s the need arose to answer specific questions. In December 1972, the group manager advised the Chief of Intelligence that funds were almost depleted and suggested a decision be made on whether to continue the project. As a result, the Assistant Chief of Intelligence visited the Miami office and discussed the special agent's intelligence gathering efforts. Similarly, in January 1974, national end regional intelligence offic- ials completed a review of the special agent's files. The Director of the Intelligence Division asked for this review after he received a request for additional funds. We reviewed 594 memoranda and other documents containing information obtained from informants by the special agent. The information in those memoranda and documents can be cate- gorized generally as information on: --Individuals' personal lives, such as their acquaint- ances, girlfriends, boyfriends, mistresses, drinking habits, and personal problems. -- Individuals' daily activities and personal contacts resulting from informants' surveillances. -- Individuals' lifestyles, such as their spending habits, purchases of personal property, and mistresses being financially maintained. -- Financial transactions that might involve kickbacks, payoffs, loan sharking, prostitution, and pornography. -- Taxpayers taken from police reports and newspaper articles. Most of the information was general and did not relate ,o possible tax violations. Some of it, however, was speci- fically tax related and useful to IRS. For example, there were records relating to a skimming operation at a movie theater which the special agent reported involved unreported income exceeding $125,000 a year. The information was obtained 75 APPENDIX II APPENDIX II from informants in connection with a joint audit/intelliqence investigation which led to the proposed assessment of taxes and penalties totaling $344,300 and conviction of the tax- payer. Extent to which information was gathered on the sex lives and drinking habits of private citizens As a result of newspaper articles alleginq that IRS gathered personal information on prominent citizens, Internal Audit expanded an investigation of the Jacksonville district' s intelligence gathering activities to include Operation Lepre- chaun. A June 1975 report on that investigation stated that of 594 documents containing information that the special aqei.t obtained from informants, 135, or 23 percent, included refer- ences to sexual and/or drinking activities. According to the report, only 70 of the 135 documents contained tax-related information. National office intelligence officials, on the other hand, reviewed the 65 supposedly non-tax-related docu- ments and told us that most were tax related. We found that the documents accumulated by the special agent in Operation Leprechaun contained sporadic references to taxpayers' sexual and drinking activities. This type of information was a minor part of the overall content of the documents, however, and generally consisted of the casual or incidental mention of individuals in terms connoting or des- cribing sexual cr drinking activities. In all cases the ref- erences were intermingled either with other information of a general background nature or with information which had spec- ific tax implications. Some references to sexual activities were includeu in information which was not tax related; they did relate, however, to the objectives of the intelligence gathering effort--organized crime and political corruption. Several documents, for example, contained rather explicit references to c judge's sexual relations. Although the information had no appa-ent relevance to possible tax violations, the docu- ments indicate the information Hid relate to political cor- ruption--a blackmail scheme in v iich the judqe was allegedly coerced into judicial actions favorable to the perpetrators of the scheme. 76 APPENDIX II APPENDIX II The documents having references to sexual and/or drink- ing activities were generally quite lengthy and detailed. Many were submitted to the special agent by informants as periodic reports of their activities. In preparing the documents, the special agent and informants spoke freely of taxpayers' activities and often referred to their social, as well as business and financial, affairs. The references to sexual activities often were made in the context of the taxpayers' business operations such as pornography, prosti- tution, and the operation of massage parlors. In identifying references to sexual and/or drinking activities in the 594 documents, Internal Audit used a broad definition of the terms and made no distinction as tc the type or extent of such information in each document. In addition, their June 1975 report did not explain the nature or context of the references to sex and drinking. Among the 135 documents identified by Internal Audit as containing such references were 39 documents in which refer- ence was made to individuals in terms such as: "girlfriend," "a lover," "a mistress," and "a pornographic writer." In- cluded also were 50 documents in which mention was made of individuals or their activities by reference to their sexual or drinking proclivities: "a known homosexual," "an alcoho- lic," "a lesbian," "a prostitute," and "prcstitution." A third category (46 documents) included more explicit refer- ences to individuals' sexual or drinking activities such as: " * * * an alcoholic and usually is 'smashed' * * * ," "* * * an orgy of women and narcotics * * *," "* * * have had social and sexual relations with * **," "* * * became intoxicated one night and *," and, "* * * a lesbian homosexually married to * * *." An Internal Audit official in the southeast region told us the auditors classified as sex related any document that implied or contained any statement specifically involving the sex activities, habits, or relationships of the individual. He said they took the most conservative approach because the news media had been accusing IRS of a "cover up," and he had understood that the results of their work on the type of infor- mation collected would be qualified. In the memorandum trans- mitting the results of the investigation to the national office, the Regional Inspector responsible for Internal Audit and Internal Security activities in the southeast region wrote: 77 APPENDIX II APPENDIX II "The attached reports also comment on the extent that 'sex related' material or information was located in the Intelligence files. There is some danger that these results will be misunderstood. "In brief, there is sex data in the file. It should be noted as well that for the purpose of these reports 'sex data' has been broadly applied. eg. 'Mr. X has a mistress', was interpreted by us to be 'sex data.' Caution therefore, should be exercised in drawing con- clusions that IRS was involved in 'sex spying'. CONTRIBUTION OF OPERATION LEPRECHAUN TO THE INTELLIGENCE MISSION We reviewed the result oration Leprechaun as re- ported by the Intelligence b and the Internal Audit Division. Although both diva.ons reported on the results of the operation, they differed as to the extent of addi- tional taxes expected to be collected. In July 1975, the Intelligence Division summarized the results of the special agent's efforts. It attributed 15 tdx cases to the special agent's work, involving actual or estima- ted taxes and penalties of $6.9 million. The Division said the information was still being processed and considerably more revenue was expected. In August 1975, Internal Audit evaluated the Intelligence Division's assessment of tax cases and revenue resulting from the special agent's efforts and determined that the special agent's information was valuable in 13 cases and that the estimated taxes and penalties in those cases totaled $4.7 million. Of that amount, $2.6 million had been assessed, and the other $2.1 million represented cases awaiting review and approval within IRS. The $2.2 million difference between the estimates by Intelligence and InteLnal Audit resulted primarily because the two disagreed on the value of the special agent's infor- mation in initiating or developing tax cases. Internal Audit determined that information obtained by the special agent was 78 APPENDIX II APPENDIX II Intelli- of value in initiating two cases not included in the gence Division's estimates. Also, Internal Audit determined that information obtained by the special agent was of no value esti- to four cases included in the Intelligence Division's mates. The net effect was to cause Internal Audit's esti-of mate to be $1.7 million lower. The remaining difference recent and $500,000 resulted because Internal Audit used more more formal estimates of tax assessments. As of January 1977, IRS had assessed taxes and penalties $49,000. in the 13 cases totaling $2.7 million and had collected of Included in the $2.7 million, however, was an assessment against one taxpayer who subsequently died. IRS $2.6 million wrote off $1.6 million as uncollectible, leaving an assessment of $1 million outstanding. IRS also took criminal prosecution action against five Oper- taxpayers based in part on information gathered during ation Leprechaun. As of January 1977, the status of these five cases was as follows: --One case resulted in conviction and sentencing. -- One case resulted in a fine after the taxpayer pleaded guilty. -- One case was scheduled by the Department of Justice for prosecution. -- Two cases were recommended to the Department of Justice for prosecution but were declined. 79 APPENDIX III APPENDIX III 0MGQESOnU&UlaO Department of the Intmnal RosnUe I 93G-152, Amend. I l(1l;)-91, l supplem ent CORRECTED COPY* Treasury Service Amend. 41c-lS, Amend. 1 1l-11R, Amend. I 5(12)c-25. Amend. 1 42c-328, Amend. 1 61C-3, Amend. I 45C-231, Amend. 1 71G.-Q, Amend. 1 March 16, 1976 Information Gatherin Guidelines q4r-57, Amend. I Section 1. Purpose The purpose of this Supplement is to amend the basic Supplement in conformance with a memorandum to all Regional Commissioners and'All District Directors dated December 2, 1975. Section 2. Memorandum of December 2, 1975 .01 The effect of the December 2, 1975 memorandum was to revise Section 5.06 of the basic Supplement to read as follows: ".06 Employees assigned to projects or individual information gathering may obtain information from sources outsidea the Service for purposes of verifying the filing of required returns, payment of tax, exempt status, proper reporting of income, deductions or credits, or otherwise determining compliance with the tax laws. These guidelines are for information gathering only and do not affect information item evaluation as set forth in IRM 9311.204). The information obtained musr be directly tax related and necessary to the administration of the tax laws. (See Sections 4.01 and 4.02) The '...sources outside the Service...' which may be contacted and the activities which may be performed by special agents in connection with authorized information gathering aessinments are as follows. "1 Inquiries at federal, state and local governmental agencies, including, but not limited to: "a law enforcement b,dies "b crime commiasaions "c regulatory and licensing brarcicas "d motor vehicle registration "e real estate records "2 Inquiries at state ane local tax authorities. "3 Contact with the orig.,nal informant and other informants who are believed to possess pertinent information. "4 Contact with foreigr governments (both tax treaty and other nations). "5 Surveillance (approved by the Chief). "6 In making these inquiries the special agent is allowed to disclose the name of the taxpayer for identification purposes in an effort to secure information that is directly tax related and necessary to the administration of the tax laws." Section 3. Effect on Other Documents 5G-118l, 5(12)G-25, 61C-3, 71-9, 94.-57. This "effect" should be annotated by pen and ink on the basic Supplement with a reference to this Amendment. * Dispose of all previous copies of :anual Supplement, also dated M.arch 16, 1975, dt ich 7 inadvertently showed CR "l(15)G-9h,eni Aodistant Commsi r 1" instead of "1(15)G-91, Ame nd. 1." (Compliance) Distribution: 80 APPENDIX III APPENDIX III 91G-118 maoxodf~IgG~ 5(12)G-25 Supplem .t 42G-105 420-328 61G-3 71G-9 fOOPWU oM4006 i t 91G-33 94G-57 June 23, 1975 ,Inf,&ionth n Guidelines Section 1. Purpose .01 This Supplement implement, Policy Stateent P-1-1 (Approved 6-23-75). attached. and provides guidelines for the gahering of information that may be solicited, obtained and retained for use by Service personnel as background material prior to the aseignment of a caer for collection. examination or investigation. .02 Them guidelinos are not intended to alter in any y the pgathering. solicitation C and documentation of tel related fact aend evidoece nacesury In developin cases that have - been aseigued for collection of taxes. examintion or investigation of a t&x liability. Section 2. Backaround .01 Compliance with tlhe tax 1v which the Service is authorized and directed to enforce cannot be determined solely by rexirence to the information on returns and documents filed Withl the the Service. Therefore, the Service must obtain information fro outside sources for effective administration of the tax lasN. .02 Information gathring activities which were suspended by telegrae to All Regional Manul *a Manual Supplement 1 Comissionere on January 22, 1915 (reissued in the Internal Revenue 91RDD-7, CR 41IDD-18 and 51RD0-20 and 71IDD-1) and by telegran to All Regional Coioonerse District Directors and Service Center Directors ne February 7, 1975 (reissued in the Internal udy be Revenue Mnual as Manual Supplent 93G-148, CR 420-323, 45C-223, 5(12)G-22, and 71C-3) resuned in accordance with the gutidelines and definitions met otit in this Mnu l Suppleant Section 3. Record Reteatio end Destructionu .01 No information documnt of any type preently on hand or hereaft-r acquired in the Service concerning Intelligence Information Gatbhering Joint Compliance ProgrSmCoordinated Compliance Projects and Returns Compliance Progre will be destroyed util the Senate Select Comittee and all other official reviewing bodies. completo their investigations of intelligencC activities carried out by or on behalf of the Federal Government. The supenion of destrue- cloa procedures doe not preclude use of such information for civil or cr]ni-ul tax adminia- tration purpouen, provided such use does not include destruction. Instruction concerning records diaposition will be issued as soon as the investigation are cope..ted. .02 District Directors will ensure that documents nd information relating to or arising from information gathering activities (ir,.uding projects and prograM), whether eolicited or not indicate iusolicited, which ate not necesary to ta aedminietration of the tax lanv and do a violation of a Federal 1aw enforced by another agency will be segregated and placed in a eaparate storage are with access limited to Division Chiefs, To the xtunot praeticable, the C data should be filed according to taxpayer nae. An index of all documents from the dis- C continued Information Gathering and Retrieval System should be retained. These records may be transmitted to the Federal Records Center, or destroyed in accordance with IN 1(15)59, when the Congressional investigations specified in Section 3.01 are concluded. .03 Directly tax related documents (defined in Section 4) remaining after the review specified in Section 3.02 *hall be maintained in accordance with the provisions of there gtideline. 1t(15)59, 4100, 4200, 4500, 5100, 5(12)00, 6100, 7100, 9100, 9300, 9400 81
Internal Revenue Service's Controls Over the Use of Confidential Informants: Recent Improvements Not Adequate
Published by the Government Accountability Office on 1977-09-01.
Below is a raw (and likely hideous) rendition of the original report. (PDF)