This week’s news that the CIA’s Office of Inspector General destroyed two copies of the Senate Select Committee on Intelligence Report (SSCI Report) on the CIA’s Detention and Interrogation Program comes on the heels of last week’s DC Circuit decision in ACLU v. CIA that the SSCI Report is not an “agency record” subject to the Freedom of Information Act (FOIA). The ACLU v. CIA decision is another setback for transparency over torture and means that a review of the complete 6,700-page Report for public release will not come through FOIA anytime soon.
The DC Circuit’s decision does not mean, however, that agencies can destroy their copies of the SSCI Report or return them to the SSCI, as Sen. Richard Burr (R-NC) demanded earlier (and has apparently repeated since the DC Circuit’s decision). Agencies should still have to preserve their copies of the SSCI Report as agency “records” under federal record-keeping laws, which should free them to finally open the Report, read it, and learn the full history of our use of torture in order to never repeat it — precisely what the SSCI intended in the first place. Unfortunately, the official empowered to make a determination binding on all agencies that the Report is a “record” — the Archivist of the United States — has thus far been reluctant to enter the fray despite demands from Sens. Dianne Feinstein (D-Calif.) and Patrick Leahy (D-Vt.), as well as a group of NGOs. Perhaps the realization that the lack of a mandate from the Archivist could have already contributed to the CIA’s “inadvertent” destruction of two copies of the Report — and the DC Circuit’s FOIA decision — will finally persuade the Archivist that he must act to protect agency copies of the Report.
There Are Records and Then There Are Records
The difference between “agency records” under FOIA and agency “records” under the federal record-keeping laws is far more than a technicality, especially in the case of congressional material. The difference results from the fact that FOIA is about disclosure while the record-keeping laws are about preserving documentation.
The best explanation of the distinction comes directly from the seminal 1978 FOIA case on congressional records — cited throughout last week’s ACLU v. CIA decision — Goland v. CIA. At issue in Goland was a CIA copy of a transcript of a congressional hearing relating to the history of the CIA that Congress had released to the CIA “for limited purposes as a reference document only” and subject to congressional secrecy restrictions.
The Goland court stressed that Congress directly excluded itself from FOIA’s disclosure requirements and therefore if a record containing congressionally restricted information were an “agency record” under FOIA its “release could be required, regardless of Congress’s wishes, unless the CIA could prove a specific exemption.”
Yet even as the Goland court held that the transcript was thus not an “agency record” under FOIA, it dropped a lengthy footnote expressly contrasting how the CIA’s copy of the transcript “may well” constitute a CIA “record” under the federal recordkeeping laws, for which Congress had a much broader purpose. The court stressed the broader meaning of “record,” noting “we doubt Congress would agree that an ‘agency record’ under [the federal record-keeping laws] is an ‘agency record’ under [FOIA].”
Indeed, under FOIA standards developed since Goland, even documents created by agencies in response to Congress can be held to not be “agency records” for purposes of FOIA. See, e.g., Judicial Watch v. Clinton. Yet to treat such agency-created records as not “records” under federal record-keeping laws would directly undermine the historical record of an agency’s interactions with Congress.
Similarly, FOIA jurisprudence on whether agencies “control” records for purposes of determining if they are withheld “agency records” — see, e.g., Kissinger v. Reporters Committee for Freedom of the Press or US Department of Justice v. Tax Analysts — should not be conflated with the record-keeping definition of “record.” As the case of Hillary Clinton’s emails highlights, records completely outside the possession or control of an agency can still constitute agency “records” under the record-keeping laws — they are simply improperly alienated records that need to be recovered.
Applying the Federal Records Law to Copies of the SSCI Report
Such distinctions are precisely why the DC Circuit’s decision in ACLU v. CIA would neither preclude, nor be inconsistent with, a determination that agency copies nevertheless constitute “records” the agencies must preserve. Here briefly is what that analysis would look like.
The federal record-keeping definition of “record” is expansive and includes “all recorded information” that is (1) “made or received” by agencies “under Federal law or in connection with the transaction of public business” and (2) “preserved or appropriate for preservation” by those agencies as evidence of the “operations” or “activities” of the US government. Let’s quickly take those elements in order.
“Made or Received”
The statute indicates no special meaning for “received” beyond the normal understanding of that word. Under National Archives regulations, “received” also simply means “the acceptance or collection of documentary materials by or on behalf of an agency … in the course of their official duties” regardless of who sends it and regardless of how it is sent.
Similarly, courts have applied “received” very broadly even in situations where there are serious questions about who should control access to the content. In Rohrbough v. Harris, for example, the Tenth Circuit held that deposition transcripts in civil cases related to the 1999 shootings at Columbine High School were court “records” under the federal records laws. This was despite the fact that the transcripts were never filed in any case, never used in conjunction with any motion, never cited, nor ever considered by the judges. Yet the mere fact that they were stored in a room at the courthouse under the supervision of a special master was enough for the Tenth Circuit to conclude that the district court had “received” them and therefore they were federal records. This was separate from the analysis over restrictions on access to these records.
Thus, even if the SSCI were to assert some “control” over access to the content, as the DC Circuit believes it does, the SSCI Report has certainly been “received” by the various agencies under the normal meaning of that term. Even the DC Circuit does not deny this, even acknowledging that the Report’s distribution in December 2014 “undoubtedly gives the Executive Branch some discretion to use the Full Report for internal purposes.”
“Preserved or Appropriate for Preservation”
This part of the analysis also ought to be straightforward. The content of the publicly released Executive Summary provides overwhelming evidence of the historical value of the SSCI Report making it “appropriate for preservation” as a permanent record by all of the agencies who received it.
There is a small historical rub here, however. During the battle over the fate of Henry Kissinger’s telephone transcripts in the late 1970s, the DOJ Office of Legal Counsel issued an opinion in 1981 concluding that the Archivist of the United States lacked the legal authority to make determinations about whether material is a federal record that is binding on other federal agencies. Reflecting this, National Archives and Records Administration regulations since the early 1990s have stated that “appropriate for preservation” means documentary materials that “in the judgment of the agency” should be preserved.
The effect of ceding this discretion to agencies, which “left to themselves, have a built-in incentive to dispose of records relating to ‘mistakes’” in American Friends Service Committee v. Webster, has been disastrous. The most high-profile example was the CIA’s destruction of the interrogation tapes, which was based, in part, on the fact that the CIA determined that they were “not federal records as defined” by the federal record-keeping laws. (For a more lengthy analysis of that situation, see my article here.)
However, with the passage of the Presidential and Federal Records Act Amendments of 2014, Congress finally and expressly restored this authority to the Archivist, allowing him to make determinations about whether material is a federal record that “shall be binding on all Federal agencies.”
ACLU v. DOD & The June 2009 Letter Agreement
To be clear, the application of the definition of “record” is further complicated by the DC Circuit’s conclusion in ACLU v. DOD that the June 2009 letter agreement between the SSCI and the CIA encompasses even the December 2014 final version of the SSCI Report. In particular, the 2009 agreement recites that the SSCI’s documents and reports will remain “congressional records” in their entirety and as such “these records are not CIA records under the Freedom of Information Act or any other law” (emphasis added).
While this language may give the Archivist some pause, there are at least two reasons why this does not alter the conclusion that agency copies of the SSCI Report are records.
First, the agreement does not determine the law. Two agencies agreeing about the record status of a document is not dispositive of whether it is a “record” under the law. If it were, it could invite mischief and collusion by agencies seeking to avoid the accountability mechanisms the federal records laws provide. The DC Circuit itself did not end its FOIA analysis at the text of agreement, but instead used the agreement as an indication of intent relevant to its independent FOIA analysis. Similarly, the vague reference to “any other law” does not preclude a determination that, despite the agreement, copies of the SSCI Report have become federal records under the federal records laws.
Second, and more importantly, while both the DC Circuit and the district court analyzed “the” SSCI Report as one “thing” for purposes of assessing its disclosure under FOIA, the federal record-keeping laws subject each copy of the SSCI Report to a different analysis. Under National Archives regulations, “Multiple copies of the same document … may each have records status depending on how they are used in conducting agency business.” This means, for example, that even if the copy of the Report the CIA Office of Inspector General “received” from the SSCI were not a CIA “record,” the separate electronic copy that same Office “made” of the Report when it uploaded it to its computer system could have been a CIA record, whose destruction may well have violated the federal record-keeping laws. This is also, I strongly suspect, why the CIA may be reluctant to comply with Sen. Feinstein’s letter requesting that the CIA provide a new copy to the CIA Office of Inspector General to replace the ones it destroyed. That new copy could be a new “record.”
Moreover, even accepting for the sake of argument, that the 2009 SSCI-CIA letter agreement covered the copies of the Final Report sent to the CIA, that does not mean that the same analysis applies to the copies of the Report that the SSCI separately sent to — and were “received” by — other agencies not a party to that agreement, including the Attorney General, the FBI, the Director of National Intelligence, the DOD, and the Secretary of State. (The copy sent to the President would be subject to a separate analysis under the Presidential Records Act.) Those copies ought to be “records” of those respective agencies.
Your Move, National Archives
The remaining question is whether and when the Archivist of the United States will get involved. In his response to Sens. Feinstein and Leahy, the Archivist gave the impression that it was “routine” policy for the National Archives to refrain from “interceding” in a matter that is the “subject of ongoing litigation.” This is inaccurate. Last year, for example in another high profile situation, the Archives made determinations about whether a group of Hillary Clinton’s emails constituted federal records despite the existence of so many ongoing lawsuits it would be difficult to list them all. Taken together, these situations leave the unfortunate impression that the determining factor for whether the Archivist exercises statutory authority (or refrains from doing so) is whether the DOJ believes it will advance another agency’s litigation position.
To be fair, given an apparent DOJ recommendation that the Archivist stay his hand pending litigation and given assurances made to the court that the status quo would be preserved, it was a reasonable decision not to enter the fray prematurely and risk exhausting the limited administrative capital of the National Archives. But as the revelations of the destruction by the CIA’s Office of Inspector General illustrate, the failure of the Archivist to mandate the preservation of these materials may already have had negative effects. And the more concerning prospect is that the Archivist is taking an overly narrow view of the robust authority Congress has granted his office. The Archivist cannot allow this authority to rot on the vine and must act protect to the SSCI Report.