The Federal Bureau of Investigation is the most powerful and secretive agency in the United States today. Despite the bureau's influence over a broad terrain of American life, independent examinations of the "B," as special agents refer to it, have been rare. (In the sixties The Nation published a special issue on the bureau by Fred Cook, which became The FBI Nobody Knows.) No Congressional committee has mustered the political will and staff to carry out a comprehensive review. With the bureau's political power greater than at any time in its eighty-nine-year history, an examination of the questions of how well it is performing its crime-fighting mission and safeguarding civil liberties is overdue. Beyond the recent series of F.B.I. problems—the tragedy at Waco, the shootout at Ruby Ridge, the Atlanta Olympic Park bombing, the flawed investigations by its scientific lab—are there broader kinds of defects that demand attention? We asked David Burnham, a reporter who has specialized in law-enforcement issues for thirty years, to review the current state of the F.B.I. With research support from The Nation Institute's Investigative Fund, Burnham availed himself of an entirely new kind of information—internal administrative data maintained by the Justice Department, which tracks every instance an investigative agency like the F.B.I. refers a matter for prosecution and the ultimate disposition of these referrals. Data tapes containing this detailed information, from the mid-seventies to fiscal year 1996, were obtained under the Freedom of Information Act by the Transactional Records Access Clearinghouse (TRAC), a research organization formed in 1989 by Burnham and Susan Long, a professor at Syracuse University. They then verified and analyzed the data. The F.B.I. did not respond to our request for comment on the findings of this article. The Editors The F.B.I.: A Special Report by David Burnham
The Nation. August 11, 1997I. A Bureau Full of Troubles
Despite the ugly blots of Waco and Ruby Ridge and the mess at the crime lab, and despite attacks on it by libertarians on the left and right, the Federal Bureau of Investigation is widely regarded as the world's pre-eminent crime-fighting agency. Because of the impressive investigative skills and dogged persistence of some F.B.I. agents, elements of this relatively shiny reputation are deserved. Nevertheless, our analysis of data that have been sitting unused in federal data banks—supported by interviews and evidence drawn from operating manuals, court decisions and government reports—proves conclusively that the F.B.I. today is a sloppy, unresponsive, badly managed, uncooperative and out-of-touch agency that is aggressively trying to expand its control over the American people.
Skewed Priorities
Judging from the data, and borne out in interviews with present and former law-enforcement people, the F.B.I. prefers busting drug dealers and playing cops and bank robbers to attacking the endemic problem of corporate and white-collar crime. In 1996, Justice Department records show, F.B.I. investigations resulted in 11,855 convictions in federal court—45 for every million persons in the United States. The largest identifiable chunk of the 1996 cases—about a quarter of the total—involved 2,919 people found guilty on a variety of drug charges. The next largest grouping, involving 1,898 persons, concerned fraud and embezzlement situations where banks, savings and loan associations and credit unions were the victims, often through the misuse of credit cards. Old-fashioned bank robberies—with 1,341 convictions—made up the third-largest category of cases. Overall, these numbers tell us that more than half of all F.B.I. convictions involved drug dealers, credit card scams against financial institutions and bank robbers.
By contrast, in 1996 F.B.I. investigations resulted in two convictions for the criminal violation of the nation's antitrust laws, one under the statute concerning health and safety crimes against employees, seventeen under the law traditionally used for the prosecution of brutal cops and prison guards and thirty-five for environmental crimes.
Apparently sensing a growing wave of public concern, the Justice Department and the F.B.I. stepped up investigations of medical fraud. With 148 convictions in 1996, the F.B.I. appears to have delivered more than empty rhetoric. But the problem is now estimated to cost U.S. consumers more than $100 billion a year. And it's not just a pocketbook issue, it's a violent blood sport. Untold numbers of old, ailing and bewildered people needlessly die or suffer serious bodily harm.
What is true for the health care industry also holds for many other kinds of corporate malfeasance: manufacturing companies that allow their workers to suck cancer-causing substances into their lungs, industries that poison the nation's drinking water, pharmaceutical companies that illegally market dangerous or improperly tested new products. Carefully documented studies by authorities like Edwin Sutherland, Marshall Clinard, and Philip Landrigan and Dean Baker show that the financial loss, physical pain and death that result from the work of the nation's army of white-collar criminals considerably exceed that of all the murderers, robbers, burglars and drug dealers regularly featured on the 10 o'clock news.
Yet the F.B.I. continues to give heavy emphasis to drug violations and the protection of the banks—areas that could be mostly left to state and local authorities—and its priorities have seldom been subject to outside audit. Instead, official and unofficial judgments about the F.B.I. by the President, Congress, the press and academia have long been based on the self-serving, undocumented claims of the bureau. Since the glory days of J. Edgar Hoover, F.B.I. directors have limited themselves to shouting about grave threats to the United States— from Communists, from politically active blacks and students, from drug lords, from pornographers and, most recently, from terrorists. As a result, far-reaching policy and budget decisions about the F.B.I. are almost entirely based on information (including the now discredited Uniform Crime Reports) that has been artfully edited by one of the government's most high-powered public relations machines.
Striking Out
Another surprising conclusion from the data is that for many years Justice Department prosecutors all over the country have found much of the bureau's investigative work inadequate. From 1992 to 1996, for example, federal prosecutors disposed, in one way or another, of 226,186 F.B.I. referrals. Only 57,871 of them—one-quarter—resulted in convictions; 67,710 of these referrals —about one-third of all they processed—assistant U.S. attorneys ruled legally insufficient. According to the department's data tapes, the prosecutors said they had rejected 25,599 of the matters because they were supported by "weak or insufficient evidence." Prosecutors declined an additional 28,680 on the grounds of lack of evidence that a federal offense had been committed or that criminal intent was involved. For 10,243 more, there was "minimal federal interest." Finally, 3,188 F.B.I. referrals were disposed of because there was "no known suspect."
Let it be acknowledged that prosecutors have broad discretion in choosing the cases they prosecute or ignore and that sometimes they make political or otherwise dubious decisions. Let it also be said that assistant U. S. attorneys sometimes reject matters for defensible reasons. During the five-year period in question, for example, 19,419 were declined on the grounds that they had been referred to "other authorities" for prosecution; 13,294 because of a lack of prosecutorial or investigative resources; and 3,063 at the request of the Justice Department.
Nevertheless, during the past five years federal prosecutors won convictions in 26 percent of all the F.B.I. investigations they processed. (In 1996 the record was 28 percent.) This scorecard looks even worse when compared with that racked up by other agencies. In that same year, for example, 59 percent of Drug Enforcement Administration matters ended with a conviction, 52 percent of those from the Bureau of Alcohol, Tobacco and Firearms and 41 percent of those from the Internal Revenue Service. This suggests that the F.B.I. is forwarding a substantially higher proportion of questionable cases than the other investigative units. For every conviction from 1992 to 1996, the F.B.I. had slightly more than one matter declined on the basis of insufficient evidence, no proof of criminal intent, etc. This compared with about two convictions for every "weak evidence declination" for the I.R.S., three for the B.A.T.F., six for the D.E.A. and seventeen for the Immigration and Naturalization Service.
Part of the reason the bureau does so poorly in comparison with other agencies could be that it sends federal prosecutors more difficult or politically unpopular cases than the other agencies. While drug dealers are "bad people" doing bad things, many white-collar criminals are "good people" doing bad things. Most of the D.E.A.'s clients have a considerably lower social position than those of the F.B.I.'s.
Even when the F.B.I. and the D.E.A. are going after the same targets, however, the bureau's record looks somewhat worse than the D.E.A.'s. Between 1992 and 1996, in cases where the principal charges were based on the government's two most frequently used drug laws, the D.E.A. got the green light from the prosecutors 81 percent of the time, compared with 75 percent for the F.B.I. The D.E.A.'s investigations led to prison sentences 54 percent of the time, the F.B.I.'s 49 percent. By another measure—how long it took attorneys to prosecute a matter—the D.E.A. also did better. Its median was 198 days, the F.B.I.'s 264. Everything else being equal, a high-quality investigation should require less time to prosecute than a flawed one. In only one significant area did the D.E.A. not look measurably better than the F.B.I.: The median prison sentence of offenders investigated by these agencies was the same—sixty months.
Patchwork Performance
An essential element in the fair and effective enforcement of the law is consistency. When Congress passes a law mandating comprehensive national programs to discourage larcenous bankers, environmental outlaws or drug dealers, carefully coordinated national programs of enforcement are crucial. Equally important to achieving success is public support, a goal that requires investigators and prosecutors to live up to the constitutional ideal that the government always tries to treat similarly situated citizens in similar ways.
There is considerable evidence that the F.B.I. is delivering a widely varying level of service in different parts of the country. Because 91 percent of all convictions result from plea agreements between prosecutors and defense attorneys, and because federal judges follow national guidelines 71 percent of the time when imposing prison sentences, the wildly erratic pattern of sentences would seem to indicate that F.B.I. agents in some districts are targeting more serious crimes drawing longer sentences than agents in other districts.
Some district variation is of course appropriate. After all, the crime problems in Baton Rouge or Louisville may well be quite different from those in Los Angeles or Miami. But limit your consideration to the 1996 sentences imposed on F.B.I. defendants in a selected group of districts that include the nation's biggest cities. In 1996 the median prison sentence imposed on all 11,855 people convicted on the basis of F.B.I. referrals was twenty-four months. But in the Maryland (Baltimore) district the median sentence was forty-eight months. This compared with thirty-three months for F.B.I. defendants in California Central (Los Angeles), twenty-seven months in Pennsylvania East (Philadelphia), twenty-four months in Florida South (Miami), fourteen months in Massachusetts (Boston) and thirteen months in New York South (Manhattan and several counties to the north) and six months in California North (San Francisco).
How can it be that the median sentence for people convicted on the basis of F.B.I. investigations in Maryland is eight times that in California North? Part of the explanation, of course, lies in the fact that F.B.I. agents in various parts of the country pursue very different kinds of criminals. In 1996, the largest single group of F.B.I. convictions in Maryland—28 percent—were for bank robbery. In California North, bank robbery convictions made up only 19 percent of the F.B.I. total. Drugs was another area of considerable contrast: 22 percent in Maryland and only 1 percent in California North. On the other hand, with regard to F.B.I. convictions in which theft or embezzlement by a bank officer was the lead charge, California North was more active than Maryland. While such convictions made up 29 percent of the total in the San Francisco area, they represented only 5 percent in Maryland.
But differences in the kinds of crimes that various F.B.I. offices choose to investigate offer only a partial explanation for the wide disparities in prison sentences. Another factor seems to be the competence and diligence of F.B.I. agents and prosecutors working in an individual district. The Justice Department data and interviews with current and former law-enforcement officials indicate that investigators and prosecutors in some areas are considerably less proficient in developing solid evidence against important criminal suspects than those in other areas. The evidence for this emerges from an analysis of the very different median sentences that result from F.B.I. convictions for the same crimes that have been brought in roughly similar districts. It is assumed that in any big-city F.B.I. jurisdiction, the bureau has a surfeit of criminal suspects—drug dealers, bank robbers, fraudulent bankers—whom it can investigate. It is further assumed that a well-run enforcement agency would try to focus its investigations on the most serious criminals—those with the most previous convictions, those who appear to be dealing in the largest quantities of drugs or the largest swindles, those who seem to be engaged in numerous different criminal enterprises. Finally, it is assumed that longer sentences do not reflect tougher judges. The mandatory sentencing guidelines require federal judges to impose longer sentences on those with criminal histories who deal in larger amounts of drugs or embezzle larger amounts of money. Given these assumptions, the considerable variations among districts in median sentences is surprising.
Again consider Maryland and California North, in this case the median sentences imposed from 1992 to 1996 on all convicted individuals where the lead charge against them involved the ten statutes that were most frequently cited by the F.B.I. in the two districts. The data show that when the sentences are compared by lead charge, F.B.I. investigators and assistant U.S. attorneys in Maryland obtained more prison time than their colleagues in California North for nine out of ten of the categories. The comparatively poor record of the investigators and prosecutors in California North in locking up more serious criminals is striking and not easy to explain.
Are the agents in some F.B.I. districts so caught up in bureau pressures to meet their quotas that they focus on a less important class of criminals who end up with shorter sentences? Are the Special Agents in Charge (SACs) in some F.B.I. districts failing to lead their agents in pursuit of serious crimes? Have SACs in other districts established sophisticated mentoring programs, in which experienced agents are assigned to work with rookies, that result in higher-quality investigations? How can the F.B.I. provide effective enforcement of laws dealing with national programs like health care fraud if some of its offices are not fully committed? In the face of all these regional variations, is the constitutional goal of equal treatment under the law a high-sounding fiction?
Testimony: Some Expert Witnesses
In interviews, current and former F.B.I. agents, federal prosecutors, former Justice Department officials, criminal lawyers and experienced investigators from such agencies as the I.R.S. and D.E.A. confirm the statistical evidence showing that the bureau is performing poorly.
Take the bureau's apparent lack of zeal in pursuing corporate and white-collar crime. Investigators and prosecutors all over the country contend that agents frequently are reluctant to work on these kinds of cases. "There's a whole lot of agents—and federal prosecutors too—who will not do serious paper cases because it is a lot of work," said David McGee, a highly regarded federal prosecutor who recently retired after twenty years as an assistant U.S. attorney in northern Florida. "This resistance to paper cases is a big deal, because if you don't do paper you leave out a broad spectrum of criminal matters, often the most important crimes. All that too many agents want to do these days is kick down doors." A federal prosecutor in New England complained, "It really is a tough job in my district to get an F.B.I. agent to read corporate documents."
A retired F.B.I. executive, considered by many to be among the bureau's most effective leaders in the past three decades, explained some of the pressures that distort bureau priorities. "When I was a SAC the trucking industry and the banks drove me bats. Every time they lost a package off a truck, they wanted to make it a federal case. The bankers have done a real job on the F.B.I. The banking associations lean on Washington. And at the local level, bank robberies are pretty nice for the agent. Sometimes they're exciting, and rarely do they require serious work. And don't forget what may be the most important factor: A lot of agents want to please the bankers because one of their favorite retirement jobs is being the chief of security for a bank."
Henry Ruth, a former Justice Department lawyer and the last special prosecutor in the Watergate investigation, was less critical. Although Ruth was the co-author of a scathing Treasury Department report on the F.B.I.'s actions in Waco, he praised the bureau for its nonviolent resolution of the Montana Freemen standoff in 1996. Ruth said that one reason the Justice Department statistics showed comparatively few complex white-collar crime cases is that these require much more investigative time than a bank robbery or a carjacking. As a lawyer who has defended white-collar criminals, he was generally impressed by the quality of F.B.I. investigations. "Given a choice, many agents probably would have been much happier chasing drug dealers than most of my clients. But you have to remember, a lot of white-collar crime is really boring."
And some more complaints: A veteran F.B.I. agent who is widely admired for his handling of complex government corruption and organized crime cases says, "Many younger agents have become obsessed with breaking down doors, with going after the bad guys like they do on television." "The biggest sin of today's F.B.I. is that many agents are just plain lazy," said a prosecutor from the South. "In the broadest sense, the F.B.I. is the most political of all the investigative agencies," said another recently retired federal investigator who spent his last few years in government as a senior manager. "The F.B.I. is the most arrogant and least cooperative agency in the United States government today," said the senior Washington attorney responsible for coordinating the work of a specialized enforcement agency.
A former Republican assistant attorney general, who has also served as a U.S. attorney, summed up: "Unfortunately, the ghost of Hoover still stalks the hallways. As in the past, the challenge of trying to protect the F.B.I. against all kinds of criticism remains the most important single task of successful F.B.I. officials. As in the past, the F.B.I. often does not cooperate with other enforcement agencies. As in the past, the F.B.I. mostly sets the Justice Department's criminal agenda."
II. The Expanding EmpirePoorly managed, poorly trained, poorly motivated, highly political and unresponsive, the F.B.I. has nevertheless been amazingly successful in maintaining and expanding its powers. At what cost to the taxpayer? To the liberties of the people?
The Hungry Ear
The single most important source of power for the F.B.I. is its proficiency in collecting various kinds of information about criminal suspects, about all the people who ever come into contact with them and, quite recently, about every household in the United States. Some of the F.B.I.'s information-collection efforts are regulated under laws that require public accounting. Others are governed by laws that have no public accounting mechanism. The third, and most sweeping, F.B.I. information collection method is entirely unregulated. Here are the knowable dimensions of this massive and sophisticated information drift net:
• For many years, the F.B.I. has operated the Surreptitious Entry Team, a squad of agents who specialize in breaking into houses, embassies, offices and warehouses to secretly install tiny hidden cameras and other recording devices. In 1993, according to a secret document sent by the F.B.I. to the White House's Office of Management and Budget, the team conducted 300 break-ins, six times as many as it did in 1985.
• In 1968 law-enforcement agencies were authorized by Congress to use evidence obtained with court-authorized surveillance devices to convict people charged with a fairly narrow range of serious crimes. In the twenty-nine years since then, Congress has steadily expanded the kinds of situations in which the government may eavesdrop. In 1996, according to a report by the Administrative Office of U.S. Courts, federal investigative agencies obtained warrants to install 581 taps and bugs in connection with criminal cases—more than twice the number they planted only ten years before. During a recent year federal agents listened to some 1.3 million conversations, only 15 percent of which were said to include incriminating material. The court report does not provide a breakdown on how many of the 581 warrants obtained by federal agencies were requested by the F.B.I., but because of the bureau's comparatively large size, it is assumed to operate a substantial majority of all wiretaps and bugs.
• Since 1979 a secret federal court has approved the surveillance of suspects in national security investigations. Under this category of cases, the government reportedly obtained 839 warrants for taps, bugs and break-ins in 1996, 20 percent more than in 1995 and more than three times the number reported in the first year of the program. Because each national security warrant can authorize multiple surveillance actions, the reported number of warrants substantially underestimates this category of taps, bugs and break-ins. Since the F.B.I. has primary responsibility for tracking spies and terrorists, it presumably conducts almost all of these highly classified surveillance activities.
• A third kind of F.B.I. surveillance uses equipment designed to record all numbers dialed from a particular telephone and the numbers of the telephones that are calling a person under surveillance. Between 1987 and 1993, the government reports that warrants for these "pen registers" and "trap and trace" devices more than doubled, growing to 3,423 in 1993. Because a single warrant often involves a request to place the devices on many lines, the actual usage is much higher.
• Under current law, federal investigative agencies are allowed to subpoena long-distance telephone records, credit card records, utility records and other information about an individual held by third parties. Because the agencies are not required to report publicly the frequency of such subpoenas, the F.B.I. count is not available. However, a survey of several telephone companies by former Representative Don Edwards of California indicated that federal investigators scan millions of records of long-distance calls every year. Such large-scale scanning is possible because the companies usually provide the information in computerized format.
• More than twenty years ago, several companies, including Donnelly Marketing Inc. and R.L. Polk, began to collect all the information they could about every U.S. household. Using publicly available records maintained by telephone companies, motor vehicle departments, voter registrars, county tax assessors and the Postal Service, the companies developed amazingly complete data portraits of just under 100 million households. The original purpose of these computerized lists was to allow marketing companies and politicians to target people who might be interested in their wares. Because some of the conclusions about households are based on inference, the lists are not rigorously accurate. Despite this drawback, and industry rules that its data should never be used for law-enforcement purposes, the F.B.I., the D.E.A. and possibly the I.R.S. have set up elaborate databases with such information that are now instantly available to their agents.
The F.B.I. thus controls a vast array of constantly updated personal information about you, even if you have never been a criminal or talked to a criminal. How much money do you make? What car do you drive? What are the names, ages and sex of your children? Are you a registered Democrat or Republican? How much did your house cost? What is the name and breed of your dog? The drift net covers the whole nation.
The latest strike in the F.B.I.'s constant drive to gain access to more and more information involves two programs. Under one, the nation's telephone companies are required to install a new generation of F.B.I.-approved equipment that will make it much easier for the bureau to tap telephones and track individual suspects as they move around the country. With comparatively little public discussion, this far-reaching program was mandated by a law signed by President Clinton in October 1994. Although the F.B.I. contends that the number of its taps will not dramatically increase when the new equipment is in place, it is hard to believe that this major boost in the capacity to tap will not result in a substantial increase.
The F.B.I. campaign to force telephone companies to install new "eavesdropping friendly" equipment is not limited to the United States. In 1995, after a series of meetings between the F.B.I. and representatives of a number of countries, the European Union agreed to launch a global surveillance system to combat "serious crime" and protect "national security." The proposed system is based on standards set by the F.B.I.
A second technical development that almost certainly is whetting the F.B.I.'s current appetite for wiretapping is the expected decline in the cost of transcribing recorded conversations. Wiretaps today cost the government a great deal of money because each conversation must be reviewed by an F.B.I. employee for incriminating evidence. This expense has served as an indirect brake on electronic surveillance. With rapid advances in computerized voice recognition systems able to search for key words, sentences and even broad subject matters, however, some of this reviewing effort can probably be handled by machines, so the cost of wiretapping will plummet.
The F.B.I.'s continuous effort to expand its access to information involves yet another complex campaign, this one to limit the spread of powerful, low- or no-cost computer programs that allow users to protect their electronic messages from snooping. The F.B.I., actively supported by the National Security Agency and the White House, has been a leader in a broad government effort to limit the sale of advanced encryption programs overseas and in the development of specially coded programs to allow the bureau to read secret messages. In Congressional testimony and speeches, F.B.I. Director Louis Freeh has said he is considering asking Congress to pass a law prohibiting the sale or use of encryption programs that do not include a "trap door" for the government.Stings—Anti-Corruption or Anti-Democratic?
Another development that has worked to expand the F.B.I.'s power and influence goes back to the period immediately after Watergate, when the Carter Administration approved a revolutionary change in F.B.I. investigative tactics. Until then, F.B.I. regulations had prohibited agents—in their investigation of crime—from involving themselves in criminal activities. But with the early and enthusiastic backing of Attorney General Griffin Bell and Assistant Attorney General Philip Heymann, sting operations became part of the F.B.I.'s investigative repertoire. In the view of William French Smith, President Reagan's first attorney general, it was essential for the federal government to "interject its agents into the midst of corrupt transactions. It must feign the role of corrupt participant. It must go underground."
Annual reports to Congress, available for a short time, disclosed the number of long-term F.B.I. stings. In 1991, for example, there were 171. But because that report did not include the operations that were completed in less than three days or involved national security, the official count substantially underestimated the actual use of the tactic. If these additional categories are included, the F.B.I. is now running several thousand stings a year. The tactic has in many ways been highly successful. Operation Abscam, first publicly revealed in 1980, resulted in the conviction of one senator, six representatives and several state officials on charges of taking bribes from a con man who pretended to represent Arab sheiks. But a number of F.B.I. undercover operations have caused serious harm to innocent citizens who were the accidental victims of the make-believe criminal organizations set up by the bureau. An even bigger risk, associated with undercover operations aimed at public corruption, is the destruction of the public's confidence in government institutions. This concern was the central focus of a 1984 report by the House Judiciary Subcommittee on Civil and Constitutional Rights after a four-year investigation of stings. "While investigations of public corruption may be intended to restore the public's faith in the integrity of the affected institutions," the subcommittee's report said, "ill-conceived and poorly managed undercover operations are likely to have precisely the opposite effect."
The subcommittee documented two egregiously perverse F.B.I. stings. In one 1983 operation, with the full approval of senior Justice Department lawyers, the F.B.I. created a political problem in North Carolina that did not exist, initiated a petition ostensibly designed to resolve the problem, obtained a referendum on that issue and then influenced the outcome of the referendum. The F.B.I.'s Special Agent in Charge told the Charlotte Observer that the bureau's extraordinarily manipulative investigation was aimed at rooting out corruption that had become "a way of life" in the state. In the vote-buying trial that followed, however, the evidence gathered by the F.B.I. failed to demonstrate that any voters had been paid or that anything other than traditional political organizing had occurred.
The second misbegotten sting, started in 1977 and completed in 1982, involved a heavily publicized investigation of the Cleveland Municipal Court. It began flamboyantly when a team of fifty F.B.I. agents and local police descended on the court at 9 A.M. on a busy weekday to seize three years' worth of court records. To facilitate the seizure, agents and officers were stationed behind the judges while their courts were in session. Shortly thereafter, the F.B.I.'s SAC and the acting U.S. attorney ominously announced at a news conference that "more than one judge" was implicated. In the end, however, the costly and time-consuming investigation, appropriately dubbed Operation Corkscrew, did not result in the indictment of a single judge. The apparent reason was the F.B.I.'s misplaced trust in one low-level court employee who turned out to be a liar and a thief.
Presumably to head off Congressional critics of this powerful investigative technique, the F.B.I. and Justice Department have adopted more elaborate internal review procedures. But sting operations can still go very wrong.
This February, for example, a federal district judge in South Carolina threw out the convictions of five former legislators on vote-selling charges. Judge Falcon Hawkins wrote that he had taken the unusual step of dismissing the indictments because of the "outrageous" and "egregious" misconduct of both the F.B.I. and the federal prosecutors. To obtain the convictions, the government had lied to the court, withheld evidence that might have favored the defendants, allowed false testimony from an F.B.I. agent and ignored its own regulations. In addition, the judge said, there was evidence that the investigation and prosecution were political hits that focused on Democrats while ignoring a close ally of the Republican governor.
The failure of the South Carolina operation is hardly unique. During the Reagan years, the F.B.I. worked with the three Republican U.S. attorneys in Alabama in the systematic investigation and prosecution of black Democratic activists and political leaders in the state. In the end, almost all these efforts— including an elaborate undercover investigation of Birmingham Mayor Richard Arrington—were abandoned, resulted in acquittal or were thrown out on appeal.
Stanley Brand, a Democratic Party activist, Washington lawyer and former counsel to the House of Representatives, contends that a great deal is at stake in the F.B.I.'s growing appetite for political investigations, especially against state legislatures. Brand does not argue that corruption is unimportant. But in a letter several years ago to the National Conference of State Legislatures, he warned that the F.B.I.'s methodical probes of state governments in Pennsylvania, California, Arizona and South Carolina could culminate "in the federal government's attempt to put state legislatures into criminal conservatorship."
Many of the F.B.I.'s targets were hardly models of representative democracy. But giving the F.B.I. and the frequently political Justice Department the authority to launch sweeping investigations of an entire municipal court, a state legislature or a large group of House and Senate members may have altered the basic relationship among the three branches of government.III. Civil Rights: Looking Away
In late 1992 a 15-year-old girl being held in a juvenile jail in Jackson, Mississippi, told the police she had been raped by one of her guards. The girl added that three other young female prisoners had also been raped. Partly because of an ongoing dispute between the city's police chief and the county's district attorney, the local authorities did nothing.
A job for the F.B.I., one might think. Under several Civil War-era statutes, the F.B.I. has long been the lead agency in the investigation of charges that government officials have used their powers to abuse the rights of any person, even someone in prison. But for at least nine months after the young woman's allegations came to the attention of Joseph Jackson, Special Agent in Charge of F.B.I. enforcement activities in the state, he took no action. When finally ordered by Washington to allow a bureau investigation to go forward, Jackson retired.
The F.B.I.'s nine-month delay in investigating the Mississippi prison guards was not an isolated event. In fact, the reluctance of the F.B.I. to investigate local law-enforcement agencies has been a matter of concern throughout its history. "There are SACs all over the country who resist the investigation of police abuses in their communities," said one senior Justice Department official with extensive experience as an administrator in Washington and as a prosecutor in a major northern district. "This problem is known to everyone. The mindset of J. Edgar Hoover—that the F.B.I. should wherever possible avoid offending local and state law enforcement—is a disturbing reality more than twenty years after his death."
A little-noticed report published every year by the Justice Department, along with internal administrative data obtained from the Executive Office of United States Attorneys under the Freedom of Information Act and interviews with federal investigators and prosecutors, confirms that the government has for many years treated civil rights violations entirely differently from the way it treats almost all other law-enforcement problems.
In 1996, according to a report by the Justice Department's civil rights division, 11,721 people came to the federal government with complaints that their civil rights had been violated. A substantial majority of the complaints involved allegations of physical abuse by law-enforcement officers. After preliminary screening, the F.B.I. investigated 2,619 of the matters, most of which were then referred to prosecutors. Criminal charges were brought in seventy-nine cases. (According to the department data tapes, the two largest categories of civil rights convictions included seventeen under the law reserved for brutal law-enforcement officers and an additional forty-three under a broad conspiracy law.) Justice Department records show that the 1996 pattern—in which the prosecutors filed charges in less than 1 percent of all complaints and less than 4 percent of all referrals sent them by the F.B.I.—has pretty much repeated itself for the past twenty years.
As described by the civil rights division report, more than three-quarters of civil rights complaints are screened by special agents and then rejected for formal investigation. Given the real limits of the law, the challenge of rounding up supporting evidence and other factors, many of the decisions against a full inquiry are valid. Nevertheless, a small but important number of the complaints that should be subjected to a full investigation are purposely ignored. Neither the Justice Department nor Congress has ever conducted a systematic audit to determine the integrity of the F.B.I.'s initial screening process.
The exercise of discretion by individual F.B.I. agents often produces widely disparate outcomes, as shown by a comparison of civil rights matters referred by agents from 1992 to 1996 in Alabama's three judicial districts. In Montgomery, civil rights referrals made up 23 percent of all bureau referrals, while in Mobile and Birmingham they made up 7 and 9 percent. When looked at in relation to population, the pattern is a bit different. Birmingham, with the largest population, had 88 referrals per million people while Mobile had 137 and Montgomery 168. A second factor in the Justice Department's broad failure to deal with official misconduct is the poor quality of agents frequently assigned to investigate the cases that survive the initial review. "Unless it is a really high-profile matter like the 1964 murder of the three civil rights workers in Mississippi, the F.B.I. hands these cases to its youngest and least-experienced agents and the tired old dogs," said a recently retired assistant U.S. attorney from the South. "And the cases they manage to put together often are insufficient." This view was echoed by a senior supervisor with experience in several different enforcement agencies. "A few years ago I became aware of a really bad situation along the border with Mexico," he said. "A lot of I.N.S. people in the field were beating the crap out of the illegals. The I.N.S. was out of hand. The F.B.I. was the agency responsible for investigating the situation. But the kids it sent out to conduct the investigations had two very serious problems: They didn't know diddly about asking the necessary questions, and even worse, they were unable to ask them because most of them didn't speak Spanish. It was a bad scene."
The last barrier to prosecution is the federal attorney's decision on whether or not to act on the referral. At this point many factors come into play: What are the chances of winning? Have the investigators marshaled sufficient evidence? Does the case have jury appeal? Will the prosecution be in the public interest? Will it advance the prosecutor's career?
Justice Department data show that prosecutors handle different kinds of matters in very different ways. In 1996, for example, assistant U.S. attorneys went forward on about 78 percent of the immigration matters referred to them, 40 percent of the official corruption matters and 4 percent of the civil rights matters. Federal prosecutors have sometimes argued that their decision to prosecute only a tiny fraction of civil rights matters is guided by the philosophical principle that the federal government should serve only as the court of last resort, that it is important to encourage local governments to deal with their own messes. They have no good answer, however, when asked why these high-sounding principles of local responsibility are followed so rigorously when it comes to prosecuting brutal cops but largely ignored, comparatively speaking, when the target is corrupt local and state officials.IV. National Insecurity
Fears that America's national security interests are threatened by spies, traitors or subversives have led a long line of Presidents and Congress to gradually enlarge the bureau's powers. In the past few years the perceived menace of international and domestic terrorism has provided the rallying flag. As a result of the Oklahoma City bombing, for example, Congress granted the F.B.I. $133 million in additional funding to hire new agents and modernize its facilities. F.B.I. Director Freeh told the Senate Appropriations Committee in June that this extra funding would allow the F.B.I. to double its "shoe leather" for the investigations of domestic and international terrorist groups. And in his requests for additional legal authority to conduct more electronic surveillances, Freeh always cites the growing menace of terrorism.
Because of the secrecy that surrounds efforts to defeat these threats, the Justice Department's enforcement data provide only a tiny window onto the F.B.I.'s national security activities. Nevertheless, the facts are disturbing: In an investigative and posecutive area that must be the F.B.I.'s absolute top concern, the bureau's record has been mediocre. Only 22 percent of F.B.I. internal security and terrorism cases disposed of by federal prosecutors from 1992 to 1996—forty of a total of 182—resulted in a conviction. In other words, more than three-quarters were declined by the prosecutors, dismissed by the judges or ended with a not-guilty verdict. Among the vast majority of cases that the Justice Department declined to prosecute, 65 percent were tossed out because the evidence was weak or they were otherwise legally flawed.
The principal charges brought against the forty convicted people involved a wide range of crimes. Eight of those convicted had been charged with gathering defense information or delivering it to a foreign government. In seven cases the lead violation was the sale or receipt of stolen goods, in four it was counterfeiting or forgery and in two it was money laundering. There was one conviction under a law prohibiting the harboring or concealing of a person wanted in a national security case, one against the importation of obscene materials, one for the use of a fictitious name and one for the use of a counterfeit access device. Curiously, the lead charge in two of the forty convictions was malicious mischief. How is it possible that in the highest-priority F.B.I. investigations, a substantial majority result in a declination, dismissal or finding of not guilty?
How is it possible that two-thirds are turned down because of weak evidence? Remember, some of the F.B.I.'s worst abuses have occurred under the national security banner. If federal prosecutors find so many flaws in the national security cases that the F.B.I. hands over for prosecution—thus becoming subject to outside review—how much confidence can be placed in the care and attention that the F.B.I. devotes to all those national security investigations that never become public?V. Who's in Charge Here?
Because the F.B.I. operates within the Justice Department, most people assume that the attorney general supervises it. This is incorrect. From his appointment in 1924 to his death in 1972, J. Edgar Hoover was his own boss. With the passage of a law in 1976 giving the director a ten-year term and establishing that he can be removed only for cause, the F.B.I. director remains very much his own master. The F.B.I.'s independence has been especially obvious during the Clinton Administration, because Attorney General Janet Reno —beginning with Waco—has given the bureau pretty much anything it wanted.
The extraordinary freedom of the F.B.I. from meaningful supervision is not just an inside-the-Beltway reality. As we have seen, F.B.I. district offices exercise considerable discretion in deciding whether to investigate civil rights complaints and many other matters as well. Frank Tuerkheimer, a former U.S. attorney who is now a professor at the University of Wisconsin Law School, says, "A big part of the F.B.I.'s power is passive. They usually don't fight you, they just wait you out. They know you will be gone in a few years, but they will not." In addition to this indirect exercise of power, Tuerkheimer said, the F.B.I. sometimes simply refuses to work on certain kinds of cases. A prosecutor in the New York area confirms the practice: "The SAC just would not involve himself in a serious official corruption matter in the district, and my boss— the U.S. attorney—was not prepared to make an issue out of it. To get around the problem, I had to shop around for help, finally recruiting some I.R.S. agents."
SACs frequently are the de facto bosses of all federal enforcement in their fiefdoms. "The theory is that the U.S. attorney in New York or Chicago or wherever is the top dog in the area when it comes to federal enforcement," said one senior investigator. "But in many districts the SACs have more experience and political connections in Washington than the U.S. attorney."
This dynamic, combined with the inability of a long line of attorneys general to set and implement their enforcement priorities, means that across the country a small, invisible and generally conservative cadre of senior F.B.I. agents often have a stronger voice in determining what crimes are prosecuted than the political appointees who theoretically run the Justice Department or the members of Congress who go on passing new laws.VI. A Matter of Accountability
The F.B.I. is not performing in a manner compatible with the principles or practices of representative democracy. It is an improperly political, badly managed and ineffective law-enforcement agency that is constantly seeking to grab more power. Stating the problem, however, is much easier than outlining a feasible remedy. On the one hand, to avoid the dangers of improper political influence, the F.B.I. must be isolated from the direct control of elected officials. On the other hand, to avoid the threat of a freewheeling national police force that is allowed to set its own agenda, the F.B.I. must be responsive to appropriate political review and direction.
One beginning point might be to compel the attorney general actually to manage the Justice Department and its investigative agencies in accordance with laws and executive orders going back to shortly after the Civil War. But other laws and traditions have given the U.S. attorneys considerable independence, making the Justice Department a scattershot "adhocracy" rather than a tightly run, top-down bureaucracy. This absence of centralized management extends to the F.B.I. One reason is that the bureau for many years has had an independent power base on Capitol Hill. A second is that many attorneys general have been reluctant to challenge the F.B.I. because they are fearful of the dark secrets it collects and has used to embarrass or blackmail their administrations. Another reason is that senior F.B.I. officials are often a lot more experienced than their putative bosses in Justice, who seldom hold office for more than two or three years. Finally, attorneys general have failed to manage because they lack a sufficiently large and expert staff to advise them on the bureau's performance.
Unfortunately, shoring up the attorney general's power might create a whole new set of problems. With a truly coordinated team of federal investigators and prosecutors, an ambitious and unprincipled A.G. like Reagan's Edwin Meese 3d or Nixon's John Mitchell or Harding's Harry Daugherty or Wilson's A. Mitchell Palmer could threaten the very essence of democratic government by throwing the full weight of the Justice Department against an administration's political enemies.
Given the potential hazards of sweeping reforms, some more focused changes seem desirable and doable. They are all designed to pierce the deep secrecy in which the F.B.I. has long operated and to provide for greater public accountability.
• Congress should pass legislation requiring the Justice Department's inspector general to publish an annual audit answering the following questions: How many matters did the F.B.I. refer to prosecutors for prosecution? How many involved criminal violations of drug laws, bank laws, environmental laws? How many were prosecuted? How many resulted in a conviction? How many agents are working in each district? How have the F.B.I. enforcement activities changed over time? The annual report should document instances when the F.B.I. refuses to cooperate with other agencies in matters of mutual concern.
• The law requiring the inspector general to publish an annual audit should also contain a provision eliminating Attorney General Reno's unpublished November 1994 order prohibiting the inspector general from investigating misconduct in the F.B.I. and the D.E.A.
• Congress should require all U.S. attorneys to publish annual reports describing federal enforcement efforts—including those by the F.B.I.—in their districts. These would provide the press, public interest groups, the academic community and lawyes with precise, easy-to-understand information about the current priorities and practices of the investigative agencies and the prosecutors. U.S. attorneys who fail to provide such information should be disciplined.
• All F.B.I. surveillance activities should be tightly controlled. Almost forty years ago Congress approved a law regulating wiretapping that prohibited all such surveillance except by government enforcement agents who had obtained a judicial warrant. In addition, the law required the Administrative Office of the Courts to publish an annual report giving limited details about every tap and bug aimed at criminal suspects. Congress should establish similar mechanisms to govern the bureau's essentially unregulated investigative use of new technologies, long-distance telephone records and marketing data. It should mandate the publication of reports that describe how, when and where these new-age tools are being used.
• Congress should establish a judicial-warrant and public-reporting requirement for F.B.I. undercover operations, particularly those aimed at members of Congress, state legislators and judges. Undercover operations are clearly much more intrusive than electronic surveillance. If agents must obtain a court order to tap a phone, it is hard to understand why a judicial warrant should not be required when they deploy an undercover operator in an office. While few requests for wiretaps are refused, it seems likely that the warrant requirement and the publication of a summary report by the courts have worked to minimize abuses.
A major reason for the F.B.I.'s overall poor performance and its abuses of power is that it has evaded the harsh light of public scrutiny to which some other agencies of the federal government (like the I.R.S.) have been subjected. Buoyed in a sea of uninformed apathy, it plies an erratic course. Congress, the press, academics, lawyers and ordinary citizens must be informed and awakened. Greater public awareness of the bureau's shortcomings is an essential first step in building a political consensus for real reforms. Curiously enough, in his testimony before the House Judiciary Subcommittee on Crime on June 5, Louis Freeh summed up why rigorous oversight is so important. "We are potentially the most dangerous agency in the country," said the Director of the F.B.I.
Note: The links in this article lead to data published by the Transactional Records Access Clearinghouse. Because of space limitations this article could present only the highlights of the data amassed by TRAC. Those interested in more information should check out TRAC's Web site on the F.B.I. at: http://www.trac.syr.edu/tracfbi/. The site contains some 23,000 pages of materials, including maps, graphs and text. There are detailed tables on a variety of enforcement indicators, along with staffing patterns, salaries, etc. for each F.B.I. district. . . .
David Burnham, Transactional Records Access Clearinghouse
