Tracking the Fed with Trac - The Transactional Records Access Clearing House
by David Burnham
In 1789 the thirteen states agreed to govern themselves under a new Constitution that was driven by two somewhat contradictory forces. The founding fathers recognized the need for a powerful national government. From the very beginning, however, they also saw the need to establish sufficient checks and balances to prevent this government from abusing its powers. Unlike the powerless creature erected under the short lived Articles of Confederation, the new national government was given a legal authority essential to the existence of any state: the right to force its citizens to pay taxes. But as a counter weight to this and other grants of power, the new government was divided into three competing authorities; the executive branch, Congress and the courts.
Almost immediately, however, a majority of the American people decided that the checks and balances provided by the constitutionally mandated division of power were not sufficient. The Bill of Rights, the first ten amendments of the Constitution, were swiftly adopted by the Congress and the necessary number of states. The first amendment, of course, checked the authority of the federal government by giving citizens and the press an almost unfettered right to speak their minds about government and all other subjects.
The various legal powers of the executive branch of the federal government were imposing indeed, but so were the counter balancing powers of Congress and the judiciary and, under the First Amendment, the public and the press.
The legal authority of the people to challenge their government in a variety of forums had been spelled out in a clear and relatively unambiguous way. Working through a number of surrogates including Congress, the press, the defense bar, public interest groups and scholars, the public had been formally empowered to hold the federal government accountable.
For a variety of complex reasons, however, these five separate institutions of oversight have not done a very good job. Although there have been occasional exceptions, the Congress, the press, the defense bar, academia and the public interest groups have generally failed to carry out one of the single most important tasks of representative democracy.
Constructive oversight attempts to answer two inseparable questions. First, are the agencies of government effective, do they work? Second, are the agencies fair? Because Congress has given the agencies a number of crucial tasks - collecting taxes, reducing the distribution of illegal drugs, regulating the use of nuclear power, guaranteeing a working banking system, and protecting the environment, to mention just a few - making sure they achieve their stated goals is of real significance. Just as important, however, is the second question: in the pursuit of their primary missions are the agencies acting in a fair and even handed fashion?
For many years, civil liberties and civil rights activists throughout the United States have sought to guarantee the equitable treatment of citizens by concentrating their concerns on particular programs and individual cases. As a general rule, they have not concerned themselves with the broader effort to maintain general oversight of the government. As stated in the currently popular cliche, they have been mostly reactive, rather than proactive. But in the last decade of the twentieth century, an age when Congress has seen fit to steadily enhance the underlying power of almost all federal enforcement agencies, constant, informed, and probing oversight has become an ever more important element of maintaining representative democracy. It simply is not sufficient to have the formal institutions of oversight in place. What good are Congressional committees, reporters, public interest groups and scholars - empowered by the Constitution and the Bill of Rights to act as the eyes and ears of the public - if they are unable to examine the actual performance of the government?
The situation, however, is not hopeless. Technology, in this case computer technology, can be harnessed in the important cause of oversight. All that is required is to teach reporters and editors, congressional staff members and other representatives of the public how to meet the challenge. In the age of the computer, genuinely effective oversight is possible if the overseers can be coaxed into developing the necessary tools, understanding and skills.
This is the mission of the Transactional Records Access Clearing House (TRAC), a new tax exempt organization associated with Syracuse University. TRAC's purpose, simply stated, is to extract daily operating data from federal enforcement agencies and help people perceive the policy issues revealed by them. TRAC was created two years ago by Susan Long, an associate professor of quantitative methods in Syracuse University's School of Management, and myself, a writer and reporter.
Before presenting some examples of how TRAC functions, lets examine information environment in which it swims.
Number One. The universal adoption of computers during the last two decades means that information about the daily operations of all federal agencies are now stored in a way that is vastly more easy to retrieve than in the age of paper records.
Number Two. The federal courts have in a general way ruled that the Freedom of Information Act applies to government records, regardless of whether the requested information has been stored on paper or reels of computer tape.
Number three. The rapidly increasing ability of computers to analyze complex collections of information at relatively little cost has given the citizens who know how to use computers a powerful new tool for monitoring the actual performance of their government.
The combined effects of these three developments is that the people of the United States are now in a position to vastly improve their ability to monitor what government agencies actually are doing, rather than totally relying on what the heads of these agencies say they are doing.
There are, however, several formidable barriers that must be overcome before we reach the new age of oversight. To my way of thinking, the single most difficult barrier is cultural. All of us are literate. Almost all of us were educated in the pre-computer world where documents -treaties, press releases, contracts, memos, reports -were both the foundation stones of action and the threads from which the fabric of memory and history are woven. But the documents that we have been required to rely upon since the days of ancient Mesopotamia have a flaw, they record what the people who wrote them say has happened, rather than what happened.
As noted above, however, modern bureaucracies have in the last twenty years established computer systems to record the millions of official actions performed by their agents. The recording, of course, is not new. Even Hitler's henchmen kept careful written records of their awful work. What is new is that the transactional records of all federal bureaucracies are now maintained in a format which makes it possible to analyze them without employing a massive army of expensive clerks.
But because of our centuries of cultural conditioning, very few of us have yet grasped the potential power of the government's transactional data. I have been an investigative reporter and writer most of my working life. As a result of this background, I have some strong feelings about the general failure of the press to cover the actual performance of government. (I also have some pretty strong feelings about the passive stance of many academics and Congress, but I will limit my comments to the press.)
With the exception of the Muckraking Era just before World War One and a short period that reached its height during the second presidential term of Richard Nixon, American newspapers have rarely undertaken any serious investigative reporting. If reporters were honest with themselves, they would see they mostly serve as high priced stenographers. When analyzed in the cold light of day, the role of the press during most of the 20th Century has been very simple. Write down what the people in power say is happening. Never, or almost never, make a serious effort to determine what the institutions of society - the police departments, the universities, the hospitals and the federal government - actually are doing. Just look at your newspapers. They are filled with official statements by President Bush. By your governor. By your local district attorney. By the heads of your museums and universities. Examine the so called investigative articles in your paper. It is obvious that most of these sorry excuses for independent thought actually are based on unsourced leaks from one of these people of power who is trying to undermine one of their enemies. A recent example of this tawdry practice involved Attorney General Richard Thornburgh leaking of false negative information about the House Democratic Whip, Representative William H. Grey.
Does that seem too cynical? After 32 years in the news business I don't think it is. Furthermore, I temper my cynical views of America's press with the optimistic belief that a small but revolutionary improvement in the performance of some American reporters and Congressional staff members and public interest groups and scholars is possible. All that is needed is an awareness that these records are available and the skills to utilize them.
This was the thought process that two years ago led Susan Long, a professor at the Syracuse University, and myself to form TRAC. Our goal then, and our goal today, is to break down the cultural and technical barriers that prevent the representatives of the public from acquiring and using the information that is available to them.
So far I have been presenting a general theory. I now will recount the course of events that led Long and me to form TRAC and then describe some of the findings from our initial studies.
I was hired by the New York Times in 1968 to cover the criminal justice system of New York. My hiring came after a three year stint on the staff of the President Johnson's Crime Commission. Because of my experience watching how the commission's experts analyzed criminal justice problems I decided I was going to make a change in the definition of news. Why not spend at least some of my time looking at what the government was doing?
One of my first stories involved robbery, an area of great public concern. I knew that New York, like most American cities to this day, collected statistics such a way that it was impossible to assess how the police, prosecutors and the courts actually dealt with the people who were being arrested for robbery. So I went to the police and asked them to give me a random sample of the reports the patrol officers had filled out the previous year while they were booking people they had arrested for robbery. Then I went to the court houses and found the records that described the outcome of each of my sample cases. The finding was truly shocking. Only one percent of all those arrested for robbery were convicted of robbery. A full thirty percent of the charges had been dismissed. The balance pleaded guilty to minor charges, usually being sentenced to four or five months in prison.
The statistics raised good questions. How come a system that was costing the taxpayers billions of dollars a year was only convicting one percent of those arrested for the serious crime of robbery. Were the police officers over booking? Were they poorly trained? Were there an insufficient number of assistant da's and judges?
Another story I did for the Times involved murder. I had a feeling that the public's fear of crime had a large component of racial prejudice. What were the facts about who was killing whom? Once again the official statistics were of no use. I went to the Police Department and asked for the reports filled out at the scene of a random sample of murders that had been committed the previous year. The reports, of course, included the race, age, sex and some other details about each victim. Then I went to the courts and determined the same information about the person who in New York is known as the "perp," the perpetrator. I asked the business department of the Times to work out some correlations for me on the paper's computer. The sample showed that 48 percent of the murders involved African Americans killing African Americans, 21 percent involved Latinos killing Latinos, 13 percent involved whites killing whites. The balance --less than one out of five of the cases- were cross racial. Whites killing blacks. Blacks killing whites. hispanics killing blacks, etc.
At about the same time I was doing my studies in New York, Long, then a young graduate student, began her examination of the IRS. Using the FOIA, Long began asking the IRS for data. The IRS at first was extremely hostile. Long, since that time, has brought a total of 13 federal suits asking the courts to require the IRS to provide the requested information. To date, she has won 12 of her cases, and one is pending. One of these cases was a crucial court of appeals decision holding that the FOIA covers data tapes.
Much of Professor Long's statistical analysis is sophisticated. But when I began working on a book about the IRS, I approached her about a rather simple table. I wanted the district-by-district audit rates, the district-by-district seizure rates and the district-by-district installment agreement rates.
The results were interesting. The taxpayers in San Francisco are three times more likely to be audited than the taxpayers in Boston. Seriously delinquent taxpayers in Richmond, Va., are seventeen more likely to have their property seized than delinquent taxpayers in Nashville, Tenn. Obviously there were great disparities. But disparities by themselves are not enough. The disparities, however, did allow me to ask senior IRS officials some tough questions and, along with other kinds of evidence, provided support for my contention that the IRS is a poorly managed agency that does not enforce the tax laws in an even handed way.
As a result of working together on the tax matters, Long and I decided to form TRAC and collect data from all federal enforcement agencies. For a variety of reasons relating mostly to our fields of expertise, we have focused our initial efforts on the Justice Department, the IRS and the Nuclear Regulatory Commission and the EPA.
Our first published study concerned the Justice Department. It was based on a gigantic data base we obtained from the Federal Judicial Center. The tapes included 5.5 million cases, 60 sperate files, all criminal cases, all civil cases and all appeals from 1970 to mid-1987.
TRAC's first report focused on eleven "big city" federal districts that Census data showed were demographically similar. We found some interesting district to district disparities. In Brooklyn, for example, 40 percent of the federal criminal cases brought between 1980 and 1987 involved drugs. In Baltimore, 33 percent involved drugs. In Chicago, drugs made up 19 percent of the cases. And in Los Angeles, the big city jurisdiction with the smallest proportion of case, it was 17 percent.
Los Angeles is interesting because the U.S. Attorney for most of the period in question recently was selected to be the head of the Drug Enforcement Agency. When the Senate Judiciary Committee asked him about the numbers uncovered by TRAC, he said the prosecution of a relatively few number of cases in Los Angeles was related to his emphasis on quality. That is a possibly valid explanation. But I have been informed that the U.S. attorney's office in Los Angeles for some years has operated a large scale drug project aimed at the lowest level of drug dealers who, of course, are easy to prosecute. If this information is correct, does it undermine the statement that the U.S. attorney in Los Angeles prosecuted a relatively small number of cases because the office was concentrating on the big fishes?
In addition to these interesting geographic disparities, TRAC's analysis uncovered some possibly significant national trends. In 1980, 18 percent of the criminal cases were completed at a formal trial. By 1987, it had dropped to 9 percent. The numbers indicate a dramatic increase in plea bargaining, probably because the system has become heavily loaded with drug matters. This finding suggests that the basic quality of justice provided by the federal courts -- always best protected by a formal trial -- may be eroding.
Since Long developed the enforcement tables for my IRS book, TRAC has undertaken several far more sophisticated tax administration studies that were based on an informed statistical analysis of gigantic masses of IRS data.
One of our studies concerned long term trends in the proportion of American taxpayers meeting their tax obligations. For the last six or seven years, a number of politicians and IRS officials have sought political advantage by been giving semi-hysterical speeches about the soaring rate of tax cheating in the United States. Michael Dukakis adopted this line when he was running for president. It was a lot easier to talk about supporting his ambitious social programs through money raised by grabbing tax cheats than by raising tax rates on the general public. IRS Commissioner Roscoe Egger, President Reagan's first tax man, and many top officials of the agency, also have given repeated speeches denouncing the rapid growth of tax cheating. This always was a good way to justify ever bigger IRS budgets.
There is, however, a little problem. For more than 20 years, the IRS has undertaken a long series of surveys in which a random sample of 50,000 taxpayers are audited every other year to determine the extent to which they failed to pay the taxes they owed. According to most of the official statements of the IRS and Commissioner Egger, a rapidly growing numbers of Americans were ducking their tax responsibilities. The world was indeed falling apart. Social values had gone to hell. The IRS needed an increased budget.
Looking at all the TCMP data tapes, however, Long found that the IRS' own evidence showed that nothing was as announced by Dukakis and Egger. All the smart people at the IRS, and all the Harvard advisers to the Democratic presidential candidate, had failed to make several necessary adjustments. They had somehow forgotten to adjust their tax cheat numbers for inflation. They also had failed to take into account the steady increase in taxpayers. Finally, they had not looked at the margin of error that is found in any survey, even one as large as the IRS series.
When these extremely obvious and required steps are taken, the IRS's own best research shows that contrary to the chicken little statements of Egger and Dukakis there has been virtually no change in compliance since 1969.
TRAC today is supported by grants from nine foundations and contributions from Syracuse University. We now have a total of eight employees and currently are offering a number of basic services.
Using our substantive knowledge of the agencies and the Freedom of Information Act, we identify, extract and organize documented data bases from the enforcement agencies. We also have obtained various tapes that help us put the enforcement numbers into context. Tapes from the EPA give us the pounds of toxic material that American companies report they have released to the environment. Tapes from the Office of Personnel Management tell us how many assistant US attorneys, how many IRS agents, how many Drug Enforcement Agency people are working in each judicial district. FBI tapes concerning reports of serious crime give a very rough indication of the crime problem in each district. Census data provides us the number of persons living in each district and their demographic characteristics.
For those with access to a main frame and the necessary statistical skills, the enforcement tapes may be purchased so that you can do your own research.
From these data bases we can provide special extracts on tape, diskettes or paper. For example, we currently are processing an NRC data base that will allow a local newspaper or public interest group to purchase diskettes with a complete history of NRC inspection activity just for the reactor or reactors located in their immediate vicinity.
Another special product would be the docket numbers on a valid random sample of all those indicted for a specific crime. As in my robbery and murder studies almost 20 years ago in New York, this random sample, with additional research, can provide an accurate picture of how the federal system of justice as a whole or in one district actually is dealing with a particular problem.
TRAC also will undertake special studies or provide consulting services.
TRAC has begun to be noticed. A leading national publisher of specialized news letters has agreed to pay TRAC $4,000 for a special compilation of enforcement data of interest to one of its publications. The Los Angeles Times, Newsday, the New York Times, the Chicago Tribune and many other papers, published articles on our study on big city US attorneys. The Columbia Journalism Review wrote an article about TRAC. The Senate Judiciary Committee quizzed the new head of the DEA about our data during his confirmation hearing. USA Today and the Miami Herald currently are considering buying several of our documented data sets so they can do their own research on their own main frames. The Boston Globe asked us to undertake a special study for them. The New York Times right now is considering asking TRAC to undertake a major research project for it. Kenneth Laudon, a professor at NYU, has approached us with a research effort in connection with a study he is doing on how the adoption of computers has changed the IRS, FBI and Social Security Administration. Lawyers in Philadelphia, New York and Los Angeles have begun to see that our data could help them anticipate the probable policies of particular U.S. attorneys and possibly even judges.
But if reporters are tired and unimaginative, if scholars limit themselves to sterile academic analyses of hypothetical problems, if Congress is unable or unwilling to perceive and understand the day to day operations of the executive branch, if under-funded public interest groups must attempt to study government without readily available analytic tools, then we may be in trouble. Open government is an essential element of representative government. But unless we fight for open government, the natural instincts of the massive bureaucracies of our age will be to stifle the astonishing creativity of the American people.
As usual, Alexis de Tocqueville said it very well almost 150 years ago. Tocqueville worried that the egalitarian spirit of the United States, which had reduced the independent power of the upper class, might someday lead to the development of a replacement, a despotic centralized government. This new despotism, he said, might not engage in the off-with-their heads style of the last Roman emperors. But its reach would be far more extensive.
Tocqueville predicted the new government's power over its subjects would be detailed, regular and provident. "It covers the surface of society with a network of small complicated rules, minute and uniform, through which the most original minds and the most energetic characters cannot penetrate, to rise above the crowd. The will of man is not shattered, but is softened, bent and guided; men seldom are forced by it to act, but they are constantly restrained from acting. Such power does not destroy, but it prevents existence. It does not tyrannize, but it compresses, enervates, extinguishes a people, till each nation is reduced to nothing better than a flock of timid and industrious animals, of which the government is the shepherd."
(David Burnham is the author of A Law Unto Itself: The IRS and the Abuse of Power. Based in Washington, he also is the co-director of the Transactional Records Access Clearinghouse and an Associate Research Professor at the S.I. Newhouse School of Public Communications. He currently is working on an investigative book about the Justice Department.)
Copyright, 1991, Jim Warren & Computer Professionals for Social Responsibility All rights to copy the materials contained herein are reserved, except as hereafter explicitly licensed and permitted for anyone: Anyone may receive, store and distribute copies of this ASCII-format computer textfile in purely magnetic or electronic form, including on computer networks, computer bulletin board systems, computer conferencing systems, free computer diskettes, and host and personal computers, provided and only provided that:
- this file, including this notice, is not altered in any manner, and
- no profit or payment of any kind is charged for its distribution, other than normal online connect-time fees or the cost of the magnetic media, and
- it is not reproduced nor distributed in printed or paper form, nor on CD ROM, nor in any form other than the electronic forms described above without prior written permission from the copyright holder.
Return to CFP'91 Index page.
Return to the CPSR home page.
Send mail to webmaster.
Created before October 2004