The Best Policy

The Department of Forgetting

How an obscure FBI rule is ensuring the destruction of irreplaceable historical records.

What’s lost when the FBI clears out its archives?

I got bad news from the FBI a few months ago. A file I’d requested under the Freedom of Information Act wasn’t going to be available. Ever.

And not for one of the reasons I already knew to expect—that the material was classified, that the file concerned a living person, or that no file existed to begin with. Judging by the FBI’s final response letter, there might have been a file on my subject, a long-deceased Mississippi lawyer name John R. Poole. But if there was, it got shredded.

“Records which may be responsive to your … request were destroyed on July 01, 1995,” the letter said. “The FBI Records Retention Plan and Disposition Schedules have been approved by the United States District Court for the District of Columbia and are monitored by knowledgeable representatives of the NARA.”

NARA is the National Archives and Records Administration, the agency that keeps track of everything from the Declaration of Independence to Lee Harvey Oswald’s rifle. The letter conjured up images of my file getting scrutinized by furrow-browed NARA scholars who decided that, alas, John R. Poole was not of sufficient historical interest to keep around.

At the time, I was new to the weird science of FOIA requesting, so I didn’t know the FBI was allowed to destroy files routinely. Dismayed, I looked into how the Records Retention Plan works, with help from several generous FOIA experts. What they described sounded more like a Records Destruction Plan, since it allows the FBI to discard roughly 80 percent of its files at any given time. The FBI would have you believe the plan is a best-of-all-possible-worlds compromise that preserves the essential and discards only the unworthy. Don’t buy it. Though the NARA experts who helped create the plan tried to come up with a fair, workable system, the bottom line is that the FBI gets to trash mountains of historical source material without adequateoversight. And there is nothing the public—which owns the records, after all—can do to stop it.

Like many people who make FOIA requests, I’m probably hypersensitive to the potential loss of any one file among millions, but that’s how it is when you’re researching something: The people on your punch list become all-important. I’m writing a book about the 1951 execution of Willie McGee, an African-American man from Laurel, Miss., who got the death penalty in 1945 for allegedly raping a white housewife from the same town. For the past two years, I’ve been working to find out anything I can about McGee, Poole, and dozens of other people involved in the case.

McGee’s legal saga was little-noticed at first, but it became so famous that, toward the end, President Harry Truman was getting harangued by people from all over the world who wanted him to grant McGee a pardon—some because they thought he was innocent, some because they thought the sentence was too harsh. Given the time, place, and nature of the offense, the outcome of McGee’s first trial in December 1945 was almost pre-ordained. It lasted a day, and an all-white jury found him guilty after deliberating for 2½ minutes.

During appeals and two retrials over the next five years, the Civil Rights Congress, a Communist-linked activist group based in New York, threw tremendous energy into making the case a cause. McGee’s appeals lawyer was a young, energetic Bella Abzug. Among the prominent figures who spoke out were Jessica Mitford, Paul Robeson, Albert Einstein, and Josephine Baker.

The FBI took an interest because the case involved lefties. I already have its thick file labeled “Willie McGee,” and Iwon’t have much problem gathering info about the famous people—Einstein’s file is so popular that the FBI has put it online. But the smaller fries like Poole are at risk, and to me they’re just as important. Poole was a white lawyer from Mississippi who represented McGee during his third circuit-court trial, and though he wasn’t a Communist, he got Red-baited by fellow lawyers on the other side of the case. This led to his getting disbarred, at the end of a murky process, the basic facts of which are hard to pin down. I’d hoped an FBI file on him might contain useful information.

Which brings us back to the Records Retention Plan. The reason the Poole file might not exist anymore dates back to the early days of FOIA, which was enacted by Congress in 1966. For the first several years, the FBI was largely successful in fending off FOIA requests, which is exactly how J. Edgar Hoover wanted it. “One precept of Hoover’s was that these are our files, they’re nobody else’s, and nobody else can have them,” says Scott Hodes, a Washington, D.C.-based lawyer who ran the FBI’s FOIA litigation unit from 1998 to 2002.

But Congress changed the playing field in the wake of Watergate, making it easier for researchers to get their hands on FBI material. According to Ivan Greenberg, an independent researcher who is writing a book about the FBI and civil liberties called Trouble Times,the new rules led the FBI to conduct a massive internal purging of files, among them more than 330,000 pages from the FBI’s file on “Sex Deviates,” which tracked homosexuals in government.

That phase of house-cleaning came to a halt in 1980, thanks to a case brought by the American Friends Service Committee and a host of plaintiffs, including Daniel Ellsberg and Jessica Mitford. The suit was prompted in part by unchecked file dumping at FBI field offices in cities like New York, Atlanta, Los Angeles, and elsewhere. The plaintiffs won, and Judge Harold H. Greene ordered the FBI and NARA to conduct an inventory and come up with a plan about what would be kept and destroyed.

A 17-member body dove into that task in 1981. According to a summary of the appraisal process written by archivist James Gregory Bradsher, they were confronted with files that took up 500,000 cubic feet of shelf space in 70 locations. (To put that in perspective, the volume of the Washington Monument is just over 1 million cubic feet, so the files would have filled it about halfway.) Obviously, the archivists didn’t sit down and study every word. The process relied on a method developed in the late ‘70s during a review of 35,000 cubic feet of records from the Massachusetts Superior Court. It involved a systematic sampling designed to answer a macro-question: What percentage of the FBI’s total holdings seemed to have genuine historic value?

The team reviewed roughly 20,000 files from FBI headquarters and seven major field offices—with FBI personnel on hand to pull files individually and hand them over—scribbling information about the contents on a data-collection sheet that asked some 75 questions. How thick was the file? (“Fat file theory,” a rather obvious guidepost used in the Massachusetts review, holds that if there’s a lot of stuff inside a folder, it might be important.) What were the results of the case? Whom or what was it about? What forms of intelligence-gathering were used?

Using this data, files were rated according to a scale of perceived research potential, with four main values applied: no, low, medium, and high. By May of 1981, a review of 5,832 headquarters files yielded a breakdown that went like so: 71 percent of files were deemed to have no value; 22 percent had low; 5 percent had medium; and less than 1 percent had high. Trudy Huskamp Peterson, who worked on that review and later served as the acting archivist of the United States during the Clinton administration, told me the group looked at samples of records from each of the 214 filing classifications used by the FBI. Then as now, the Bureau arranged its material in large categories that are each assigned a number. For example, 44 is Civil Rights, 76 is Escaped Federal Prisoner, and 100 is Domestic Security, which covers subversive activities on the left and right. New categories have been added since the big review in the ‘80s, covering crimes like Tampering with Consumer Products (251) and Weapons of Mass Destruction (280). Some but not all of these categories have subsequently been sampled and analyzed by NARA.

For each classification, the 1981 assessmentgroup came up with a rule for the FBI, describing in big-picture terms what they had to keep and what they could consider for destruction, a process that led to a final retention rate of about 20 percent. Civil Rights came through with broad protection. For that large category—there was a total of 234,379 cases in the headquarters and field offices combined—everything created prior to 1977 was marked “permanent,” thanks to the assumedhistorical value of this subject. Presumably, John Poole wasn’t a class 44 Civil Rights case, or his file would not be sleeping with the banana peels. My initial guess—the FBI’s response didn’t say—was that he was filed under 100, Domestic Security. There were more cases created in the Domestic Security category than in Civil Rights—1,790,191 as of 1981. The destruction guidelines on this class are looser, however, and if Poole was in there, he didn’t make the cut.

The system’s fundamentals make sense, I guess—very complicated sense—but to me the disturbing part comes at the end of the line. At some point 25 years after a case closes, a file that isn’t marked “permanent” gets pulled and looked at by one or two people inside the FBI. There are no “knowledgeable representatives of the NARA” monitoring this crucial moment. If it’s decided internally that the file isn’t important, it’s gone.

Michael Ravnitzky, an FOIA researcher based in the Washington, D.C., area, is no fan of the Records Retention Plan and likens it to an open-ended manual for strip-mining a priceless public record. “The FBI got a list of exceptional files given to them by historians, and they said, ‘We’ll keep that,’ ” he says. “We’ll keep large files. Smaller files, we’ll keep a sampling. Everything else gets tossed. That’s what the plan is.” Based on documents Ivan Greenberg obtained from the FBI, he estimates that 250 million pages were destroyed between 1986 and 1995.

But isn’t the FBI destroying only junk? I doubt it. Ernie Lazar, an independent researcher in California whose particular interest is in far-right groups, sent me a list of “destroyed” responses he’s received over the years from FBI headquarters and field offices. There are dozens. We’ll never know if they were significant—they don’t exist anymore—but they sure look interesting to me.In 1994, for example, the Baltimore field office destroyed a file called “Arab Participation and Influence of Hate Literature in the United States.” Also destroyed in the ‘90s or later: files on “race riots” from Birmingham, Ala.; a Dallas file on the John Birch Society; and a headquarters file on the political activities of actor Walter Brennan.

When I moaned about this to Peterson, she acknowledged that the system wasn’t perfect, but she said there was no other choice because of the volume. “It is a mammoth quantity of material,” she said, “and to save it all is just impossible.”An FBI spokesman told me thatthere are now 56 linear miles of files scattered in headquarters and field offices. That’s approximately 300,000 cubic feet. I have no idea how much new material has been added since ‘81—the FBI wouldn’t tell me—but there’s obviously been a huge net loss since then.

Maybe the volume istoo much to manage, but I have to wonder if it might be time to put on the brakes and reassess. Evenif the Records Retention Plan team had scrutinized every page, I wouldn’t trust their ability to decide now what might be significant to someone 100 years down the road. Also, the rate of handoff from FBI to NARA seems awfully slow. At this point, NARA has been sent onlyabout 13,100 cubic feet of records for permanent archiving. What do they have? Good luck figuring that out. There’s no general index to the NARA holdings that lists this information using comprehensible subject headings like “John Birch Society” or “Judge Crater.” You sort of have to know they have it before asking for it, which you find out by sending them or the FBI a FOIA request.

My final gripe: The volume of the FBI files isn’t that mind-boggling. NARA has a much bigger load in the attic—29,019,647 cubic feet of material in its main D.C.-area facilities, regional hubs, and Presidential libraries—and NARA manages to preserve it without institutional meltdown. That’s 60 times as much as the FBI files targeted for destruction. The half-million cubic feet of FBI documents from 1981 would have fit into about a dozen McMansions, packed floor to ceiling. The stuff was already cataloged and cross-referenced, so a simpler strategy would have been to keep it all together. The Feds have no shortage of storage space. Time published a report in 2006 on the nationwide glut of empty government buildings that require expensive upkeep, including Chicago’s Old Main Post Office (2.5 million square feet of floor space) and the General Services Administration (376 “vacant and underused” buildings, courthouses, labs, and warehouses). To protect this priceless collection of FBI material, all it would have taken was shelves, guards, and about 20,000 smoke alarms.

Meanwhile, there’s a new glimmer of hope on the John Poole front. I was able to get an interview with an FBI official who said the destroyed Poole file was listed under the classification that covers interstate transport of stolen vehicles. That’s probably not myJohn Poole, but attempts to confirm it either way have met with foot-dragging. Separately, I sent a Poole request to the Jackson, Miss., field office and was told that material “which may pertain to your subject” has been released to the big National Archives storage facility in College Park, Md.

Butwhen I contacted College Park, I was surprised to find out that this material, fully freed up by the FBI, can be obtained only with a separate FOIA request to NARA. I’ll do that next, and I’ll let you know whether I get the information or just decide to lie down and go nuts—whichever comes first.