with the General Services Administration's self-service stores. We also requested all checks for purchases and sales as well as all invoices which would show both what items had been sold and from whom those items had been purchased, assuming the company was not the manufacturer.

time

The first series of subpoenas produced a roomful of documents and the attorneys for at least one company. One of the suspected companies was James Hilles Associates of Westminster, Maryland. That company had both goose contracts for hanging folders and other items as well as BPA's to deal with all 30 stores in the Region. At the of this subpoena, we had no solid information that they were involved in any criminal acts. The issuance of the subpoena, however, produced their attorneys in my office with the startling information that corruption existed in at least 20 of the stores within the Region. Hilles's attorneys made a request and a pitch for immunity for their clients in exchange for laying out the entire scheme. Their request for immunity was rebuffed with the information that we would review all of their documents and, when that review was completed, we would talk again. They asked how long we thought that would take and we indicated at that time in late September that we hoped to finish the project of reviewing documents by January of 1978.

As a curious sidelight it should be noted that, when they first appeared, Hilles's attorneys believed that the case had been referred to us by the Internal Revenue Service. Apparently an IRS auditor had routinely been through the Hilles books back in January and February of 1977. It was the attorneys' belief that for some peculiar reason the IRS auditor had chosen to bypass the special agents of the IRS criminal investigations group and come directly to the United States Attorney's Office. Obviously, that was not what occurred, but it should be pointed out that in March of 1978 the IRS auditor finally referred the case to the criminal investigations group, having found inappropriate business deductions. As is standard procedure, the criminal investigations special agent contacted the attorneys for Hilles who at that time simply told him to contact our office. The Internal Revenue Service auditor did not have a concept of the scope of the offense or even the real nature of it. A decision was made by our office along with IRS that they would not become involved in the investigation. That decision was made in either April or May of 1978.

In early October of 1977, a meeting was held between Need and his superiors, Jones and his superiors and me. The initial topic of discussion, to my surprise, was whether G.S.A. would remain in the investigation or whether the entire matter would be handed over the to F.B.I. In my mind there was no question that the G.S.A. agents were

central to any successful investigation of their own agency. Apparently, however, prior practice, particularly in the District of Columbia, had been that G.S.A. could only gather the initial facts and then the entire case was to be turned over to the F.B.I. It was our opinion, including that of the Baltimore Office of the F.B.I., that a thorough understanding of the internal workings of G.S.A. was crucial to our investigation. We also decided that any task force established would include an even number of G.S.A. and F.B.I. agents who would work in teams. The number of agents decided upon was six, with the F.B.I. providing two first office agents trained as accountants. G.S.A. supplied office space on the 7th floor of the courthouse along with a typewriter, many desks, tables and, very crucially, a xerox machine. The office opened on November 2, 1977, and proved to be an important factor in the success of the investigation. What the office did was to establish a neutral ground where the agents were working together on an investigation and not working on either the G.S.A. 's investigation or the F.B.I.'s investigation. All the agents were treated equally with both G.S.A. having a case agent, Robert Jones, and the F.B.I. having one, Vance Need. The agents found it fairly easy to work with each other, and the neutral territory provided a place for them to be free of their supervisors with no one other than the United States Attorney's Office looking over their shoulders.

Certain guidelines were established for this group as a result of the massive size of the investigation as well as the peculiarity of having G.S.A. investigators investigating their own agency along with F.B.I. agents. An initial demand was made and an agreement established that the G.S.A. agents would not report back any details to their superiors. Bill Clinkscales who was the head of Office of Investigations was entitled to learn most of the details, but he was not to report them to any of his superiors or to any other individual within the General Services Administration. This was all done as a result of Rule 6(e) and the 6(c) listing of agency personnel provided to the Court went no higher than Bill Clinkscales. Fortunately, Clinkscales did not care to know the details so that when he was questioned by others within G.S.A. he could claim ignorance.

It was determined that when interviews were done, they would always be done on the basis of F.B.I. advice of rights form and that the F.B.I. agents would have the responsibility of preparing reports of interviews (302's). It was further determined that, as individuals' names surfaced within the investigation, separate files would be opened on each. To keep original documents in some semblance of order, copies of the documents which indicated an individual's complicity would be placed in that person's file. Similarly, every time an individual's name was mentioned in

an interview, a copy of that interview went into the file of the person whose name was mentioned. As a result of this procedure being followed, it never became necessary for the F.B.I. agents to do individual reports on any one case because the file jackets kept on the 7th floor contained all mentions of that particular individual or that individual's store. All of the original documents were kept in the order and in the manner in which they were received from the company which had responded to the subpoena.

It was also decided that no G.S.A. documents would simply be taken from the agency by the G.S.A. agents. All G.S.A. documents were subpoenaed by issuing subpoenas to the head of G.S.A. Region 3. We felt this to be was essential, although we understood that some within G.S.A. might be offended, because we would not then face a challenge a year later that the G.S.A. did not have proper authority to seize its own employees' records. The subpoenas were simply the safest means by which to obtain the records and to withstand a possible challenge later on. The original subpoenas were issued by a routine grand jury sitting in Baltimore and it was not until the original documents were returned and examined that it was determined that a special grand jury would be needed. The court was requested to convene a special G.S.A. grand jury and that was done in January of 1978. All of the previously obtained documents were then transferred to the new grand jury pursuant to Rule 6(e).

The agents settled in for the winter to begin a review of over 200,000 documents which had been returned in response to the subpoenas. It should be noted that their staying in the room on the 7th floor to review the documents was not a chore gleefully embraced. In fact, subsequent to the Hilles company lawyers indicating that in fact corruption existed at other stores, the agents determined that it might be easier and faster to begin interviewing the other store managers and see if they would confess. In order to test their theory, we permitted the interview of one store manager who appeared to be involved based on a quick review of the Hilles records. When that interview produced a stone wall, the agents realized that all of the documents would have to be sifted through before further interviews could be conducted.

The review process took substantially longer than originally expected. It began in part in October of 1977 and then fully in November of 1977 when their office in the courthouse was ready. That review, which had to be substantially completed before interviews could be conducted, took until April 10, 1978. What the review first produced was the necessity of the issuance of dozens of other subpoenas and in the end well over 250 subpoenas were issued for this branch of the investigation. All of the documents of the

companies were reviewed to determine sales to the government. Those documents were then compared with the G.S.A. documents which showed purchases by the various stores from the companies. The purchases made by the outside companies were then reviewed to determine whether they had in fact purchased the goods allegedly sold to the government.

The review of several of the companies' documents showed that they had not purchased anywhere near the number of goods that they had allegedly sold to the federal government. For example, James Hilles Associates records and G.S.Á. records showed that the federal government had been billed for over 4.4 million hanging folders. Hilles's sole source of supply for hanging folders was the Warshaw Co. in Brooklyn, New York. Hilles's own records of purchases of hanging folders showed that they had only purchased about 1 million hanging folders. There existed a discrepancy of some 3.4 million hanging folders at a total difference of $630,000.00. The proof that these records properly reflected that Hilles had sold more goods than they had purchased came through the issuance of several other sets of subpoenas. First, all of the Warshaw records were subpoenaed directly from Warshaw and correlated with the Hilles records. Second, all of the bank records of Hilles were subpoenaed from the various tanks to establish exactly what Hilles had paid for. Hilles had supplied their own checks but we chose to verify all of their records by subpoenaing the bank records. This method cf cross checking was done with each company as to every purchaser from whom the companies had bought the items which they then subsequently claimed to have resold to the federal government. This became the key to the government's ability to prove its case, because Hilles was obviously a suspect company and its own records open to challenge. The agents could, however, testify at trial that their records were bsolutely accurate because they had checked them not once lut twice. They had checked the checks of the company to (stablish purchases and they had checked the records of the cutside, disinterested manufacturers of the goods who had sold to Hilles, a wholesaler. It should be noted that while Filles is used as the primary example, because they were the primary offender, each of these procedures were followed as to all of the companies whose records were subpoenaed. During the course of this record review, the F.B.I. supplied a full-time secretary for the compilation of the individuals' files and two clerks whom we chose to call document specialists and who, along with the agents, matched many of the documents.

Some peculiar problems arose in doing the documents review and the agents came forward with some ingenious solutions. As to many of the items shipped by Hilles, the method of shipment was directly from the manufacturer to an individual store. It, therefore, became easy to determine what had been shipped to that store and then it could be

determined whether the manager or assistant managers had signed for goods which could not have been delivered because they were in fact never shipped. However, certain types of goods, specifically Pritt Glue Sticks, were shipped to Hilles Associates in large bulk and then were divided and shipped by United Parcel Service (UPS) or parcel post to the stores. The UPS shipping records did not indicate how many went to each store and the only way it could be determined was based on the number of boxes that had been sent by UPS. The agents did an analysis of cost and weight per box and then, by comparing the number of glue sticks allegedly delivered within Region 3 to those outside of Region 3, were able to determine a discrepancy. For all shipments outside of Region 3 there were 48 Pritt Glue Sticks per shipped box. Within Region 3 the number shipped per box varied from 48 upon occasion up to 1,000 on other occasions, according to the billing to the government. It was through that method that the agents were able to determine how much actually went to which store and which stores had accepted false invoices.

At first there were so many names and so many of these names were not G.S.A. employees that it was thought we would need the Treasury Department's computers to determine who these people were. Rather than to go to that expense, however, a decision was made to wait until the G.S.A. interviews were concluded and determine who if any individuals remained unidentified.

With Hilles as with all of these companies, an analysis was done by the agents of those goods that they had purchased which should not have been purchased by companies in their line of business. For example, Hilles had an extraordinary number of purchases of televisions and other appliances from several companies. Those companies' records were subpoenaed and eventually the employees of those companics were interviewed to determine if they remembered who had come into their stores to purchase and pick up items. Given to government employees and others were carpeting, trips, televisions, guns, microwave ovens, a car, a furnace, a tractor, cash monies, and, literally, any other item which could be found in a Sears catalog. A total dollar figure was run on the Hilles purchases of goods which could not have been properly resold to the government. That dollar figure for Hilles alone came to $1.3 million dollars in purchases that should not have been conveyed to the government. They were the largest offender we uncovered, but there were other companies that had amounts of $200,000.00 and $180,000.00 and several that were substantially less.

By April the documents review was sufficiently complete to permit interviewing. It was decided that the interviews would be done by teams of one G.S.A. agent and