Federal courts issue hundreds of decisions in FOIA cases every year. Most do not break new legal ground or attract media attention. In honor of Sunshine Week, and before 2025 gets too far away (we recognize we’re later than usual), the legal eagles of FOIA Advisor — Allan Blutstein (AB) and Ryan Mulvey (RM) — look back at 2024 and discuss the decisions that stood out to them (in no particular order).
1. Citizens for Responsibility & Ethics in Wash. v. DOJ (D.D.C. Mar. 31, 2024) — on remand from the D.C. Circuit, finding that the names of federal contractors who supplied pentobarbital to the Federal Bureau of Prisons (BOP) qualify as commercial information under Exemption 4; further finding that BOP demonstrated that the disputed contract terms were confidential and that foreseeable harm would result from disclosure of the contractors’ names and contract terms.
AB: To quote the Honeymooner’s Ralph Kramden, “how sweet it is.” This was precisely the decision I was hoping to see on remand, at least with respect to whether the identities of these contractors met the very low threshold of “commercial” information. As the district court nimbly reasoned, “the fact of the contractors’ names is intrinsically linked to the fact of their commercial activity in a particular market. Where a contractor’s identity cannot be separated from its commercial activity, its identity is commercial information. Put another way, the names of the pentobarbital contractors are necessarily commercial not because they are names but because they are names that appear in government contracts.” Notably, plaintiff did not appeal this decision. Do you think they should have, Ryan? Ideally, Congress should pass a shield law to protect lethal drug manufacturers and others, as several states have done. If it doesn’t, nothing stops future requesters from re-litigating this issue. Alternatively, perhaps the federal government should use firing squads for capital punishment, a method of execution Justice Sonia Sotomayor has endorsed.
RM: I don’t have strong views on whether the contract terms CREW sought were properly treated as “confidential.” The district court’s conclusion on that point in this remand turned in large part on how the parties strategically structured their arguments and exemption claims. I’m more frustrated by the court’s decision with respect to the pentobarbital contractors’ identities. Judge Friedrich was wrong the first time around on that issue, and she’s wrong again. As I explained last year, when we discussed the D.C. Circuit’s 2023 reversal of the district court’s original opinion, the beauty of the majority’s test for “commercial” was its objectivity: commercial information must be intrinsically or inherently commercial, that is, “commercial in and of itself.” This decision on remand distorts that standard. Judge Friedrich instead adopted the kind of functional inquiry endorsed by Judge Sentelle in his mediocre concurrence—one that conditions the “commercial” quality of the contractors’ names on the context of the records at issue. To wit, “the names of the pentobarbital contractors are necessarily commercial not because they are names but because they are names that appear in government contracts.” Yet those names sometimes appear on their own, outside the context of any contract! So how are the names commercial “in and of themselves”? It boggles the mind. I do think CREW should have appealed, but I understand they settled for attorney’s fees this past September. So, in a sense, they still won.
AB: There are all sorts of reasons—some unrelated to the merits— why CREW may have abandoned ship. I’d like to think that the prospect of a loss before the Supreme Court, which nullified the D.C. Circuit’s longstanding “substantial competitive harm test” in 2019, was at least in the back of their mind. The government has a strong case that “commercial” information includes any information in which a person has a commercial interest, i.e., it has a value to the person that disclosure threatens to diminish. I am not sure why your mind is boggled by the district court’s reasoning, which essentially is a mosaic theory. Thus, the name of a business may alone appear insignificant, but when combined with other information (such as a government contract) it can reveal important commercial activities. Courts have accepted mosaic theories of harm in the context of Exemptions 6 and 7(C); let’s hear it for Exemption 4.
RM: The district court wasn’t supposed to engage in “harm” analysis. In fact, Judge Friedrich got in trouble the last time around by looking at “commercial or financial repercussions” when deciding the names were “commercial information.” Circuit Judge Pillard’s opinion for the CREW court was clear: Exemption 4, in relevant part, only applies to information that is “commercial in and of itself.” “In and of itself” is commonly understood to mean intrinsically or inherently, and without consideration of contextual factors. Yet Judge Friedrich started her analysis by observing that, “insofar as the contractors’ contracts . . . concern the[ir] commercial operation[s],” then anything within those contracts, including names, is ipso facto “commercial information.” Let’s set aside the fact that the D.C. Circuit explicitly stated names can’t be “commercial” because they “reveal the existence of a contract.” On its own Judge Friedrich’s reasoning fails for incorrectly focusing on the nature of the underlying records (the contracts), rather than the nature of the information in dispute (the contractors’ names"). Again, she didn’t grapple with the “commercial nature” of the names qua names—regardless of their function in the records, or their connection to “an ongoing ‘business purpose,’” or their revelation “that the government is a customer.” And have I mentioned yet how she still thought (again, incorrectly) that the confidential quality of the names was relevant to treating them as “commercial”?
2. Am. Oversight v. HHS (D.C. Cir. May 17, 2024) — reversing the district court’s decision granting judgment to the government and ruling, in most relevant part, that communications between agencies and Congress (or their staffs) did not fall within the Exemption 5’s consultant corollary doctrine, as interpreted by the U.S. Supreme Court in Klamath, because “each side had an independent stake in the potential healthcare reform legislation under discussion”; the dissent contended that FOIA’s “text, purpose, structure, and legislative history” supported withholding under Exemption 5, notwithstanding Klamath, and that “the ramifications of the majority’s contrary interpretation of FOIA are actually quite breathtaking.”
AB: I am not thrilled about the consequences that might ensue from this decision (outlined by the dissent), but I acknowledge that the majority’s holding might be correct. I have no doubt, however, that this case is last year’s most significant one. Ever since Klamath was decided, DOJ has interpreted the Supreme Court’s Exemption 5 threshold test as going “no further than excluding ‘an interested party [that is] seeking a Government benefit at the expense of other applicants.’” DOJ/OIP, Supreme Court Rules in Exemption 5 Case, FOIA Post, Apr. 4, 2001. In DOJ’s view, Klamath did not apply “the Ninth Circuit’s more general test that disqualified an outside party due to the existence of a self-interest alone.” Two of three panelists flatly rejected DOJ’s longstanding position, opining that self-advocacy was, in fact, the “fundamental distinction” identified by Klamath and that “no appellate court” had adopted DOJ’s test. Regardless of how you come out on what the disputed records reveal in this case (i.e., self-interested negotiations, consultations, etc.), the majority’s opinion on the legal standard to be applied is momentous.
As for the dissent, I was confused by the argument that Klamath doesn’t govern this case. The government never suggested that, so I’ll let the majority’s reply speak for itself. Better is the dissent’s contention that the consultant corollary test was met on the grounds that DOJ’s interpretation of Klamath was accurate; the evidentiary record established a common interest between Congress and the Executive (both controlled by the same party); and Congress and their staff have similar obligations to exercise good judgment and to support the Constitution. I wonder why DOJ did not seek certiorari on that ground. Did it concede defeat? Waiting for Congress to amend the statute? Will a future case provide better facts? I can only speculate, of course.
RM: I agree this case might be the most important decision of 2024. The circuit not only rejected the consultant-corollary doctrine as applied to many types of interbranch communications, but it called into question well-established circuit precedents like Public Citizen v. DOJ and Ryan v. DOJ. There’s long been divergent lines of cases in the jurisdiction when it comes to records reflecting agency interactions with Congress; I’m glad we finally got some clarity, but we have yet to appreciate long-term impact.
Some might try to argue the dissent overstated any ill consequences. Judge Garcia, defending the majority, suggested there could still be “[c]ases in which members of Congress or their staffs” satisfy the Exemption 5 threshold. I’m not sure you’re likely to find those cases with any regularity. When is anybody in the Congress acting in a totally disinterested way, as if he had stepped into the “shoes” of an agency employee? That’s the standard if we take Klamath seriously…and in my mind, it’s the proper understanding of the corollary, which is entirely atextual and a judicial invention.
One final note: I was struck by the court’s oblique suggestion that, while Exemption 5 might no longer be useful for withholding records of communications between agencies and Congress, “[n]othing about today’s decision changes the fact that Congress itself is not subject to FOIA requests.” At least in this case, HHS didn’t dispute that everything was an “agency record.” But I imagine we’ll see more attempts by Congress—and the President, especially if DOGE is determined not to be an “agency”—to impose conditions or enter agreements with agencies to retain control over correspondences, so that all such records are either “congressional records” or “presidential records.” I first warned about that possible trend eight years ago.
3. Cox v. DOJ (2nd Cir. Aug. 5, 2024) — affirming district court’s decision that a congressional committee’s report concerning the CIA’s post-9/11 detention and interrogation program was a “congressional” record, not an “agency” record subject to FOIA, even though it was disseminated to various federal agencies; in reaching its decision, the Second Circuit found that because the Committee “manifested a clear intent to control the report at the time of its creation, and because the Committee's subsequent acts did not vitiate that intent,” the agencies that possessed the report did not “control” it under the “intent test” adopted by the Circuit in Behar v. DHS (2nd Cir. 2022).
RM: For those familiar with the D.C. Circuit’s caselaw on the “modified control test” and the treatment of alleged “congressional records,” the outcome here wasn’t terribly surprising, especially given the Second Circuit’s explanation of the relevant control analysis in Behar. Moreover, as the Cox court describes at length, the "CIA “torture” report at issue was already the subject of a rather prominent FOIA cases several years ago in another jurisdiction, ACLU v. CIA, 823 F.3d 655 (D.C. Cir. 2016). The requester in ACLU filed an unsuccessful cert petition, arguing that multi-factor tests like Burka and United We Stand /Judicial Watch v. Secret Service conflict with Tax Analysts. That Mr. Cox didn’t seek Supreme Court review is more likely a reflection of his pro se status, rather than a calculated decision that such an argument isn’t worth making again. To be sure, there are some interesting and important aspects of the modified control test that need to be worked out. What matters most for our purposes, though, is that Cox, together with Behar, reflects a definite embrace of the D.C. Circuit’s approach to the control question, at least for alleged presidential or congressional records.
AB: Describing this ruling as not terribly surprising is being too kind to Mr. Cox; it was a near-slam dunk. Mr. Cox deserved the Second Circuit’s wrap on the knuckles for failing to cite Behar in his opening brief. He’s no ordinary pro se FOIA plaintiff, but an associate law professor and experienced litigator. His prayer for divine intervention, i.e., an intervening reversal, was an interesting strategy. I do not blame the professor for not seeking certiorari in light of the unsuccessful bids by ACLU and Behar. I doubt a third attempt would have been the charm.
4. Nat’l Sec. Archive v. CIA (D.C. Cir. June 7, 2024) — affirming district court’s district decision that the CIA properly invoked Exemption 1 to withhold a 1989 report drafted by Leonard Peroots concerning a 1983 nuclear crisis with the Soviet Union; rejecting plaintiff’s argument that the CIA was precluded from withholding the memo because the State Department previously published a version of the memo with the CIA’s blessing.
RM: To start, let me propose that the background of this case and the record at issue is as interesting as the decision. Shout out to my fellow requester, Nate Jones of the Washington Post, who wrote a book on “Able Archer 83.” A lot of his research depended not only on archival efforts, but also FOIA requests, as I understand it. Turning back to the D.C. Circuit’s opinion, I think the application of the three-factor “official acknowledgement” doctrine was probably correct. It is unclear to me how the State Department’s redacted and transcribed version of the Perroots Memo could effect a waiver of the CIA’s exemption claims, especially since the “record” was neither “as specific” nor identical to what was published in the Foreign Relations of the United States series. Alas, the opinion seemed to focus more on why the State Department’s acknowledgment for the CIA’s “contribution” in publishing the transcribe memo was legally irrelevant. There are two other points that I find significant. First, I have mixed feelings about the Circuit’s rejection of the “public domain doctrine” as something distinct from “official acknowledgement.” I’m not sure the doctrines are actually “interchangeable,” as Judge Childs suggests. There is some confusion here about a record being accessible to the “public”—say, because it was leaked—and it being a “permanent public record,” such as a court filing. Second, I thought it odd that the opinion opened with a discussion of standing. I suspect that is a reflection of Judge Rao being on the panel.
AB: Picking up on your last point, I initially assumed the discussion of standing was triggered by NSA’s attempt to subject the disputed memo to a reclassification review, but the panel ultimately declined to consider that issue. Standing is uncontroversial in FOIA cases litigated under 5 U.S.C. § 552(a)(3), which Judge Rao surely knows, and there’s no indication that the government disputed it. Be that as it may, I agree with all of your points. It would have been helpful if the court had distinguished or reconciled the two “but see” D.C. Circuit “public domain” cases citied in the opinion. For example, in Cottone v. Reno (D.C. Cir. 1999), I presume the tapes introduced in open court during a criminal trial were, or would be, deemed “officially acknowledged” by the government. And in Students Against Genocide v. Dept. of State (D.C. Cir. 2001), there was no or little dispute that U.S. Ambassador Madeleine Albright displayed certain photographs to the U.N. Security Council; the issue was whether such display constituted a public release. Perhaps a law student would like to address this more thoroughly in a journal article.
5. Shapiro v. Dep’t of Justice (D.D.C. Aug. 1, 2024) — faced with “the latest in a very long series of disputes between the parties,” ordering the FBI to process and produce the potentially responsive contents of the agency’s “Freedom of Information and Privacy Act Document Processing System (‘FDPS’)”; rejecting the FBI’s arguments that “substantive entries” in its FDPS were alternatively (1) outside agency control, (2) lacked “sufficient permanence” to be considered a “record,” or (3) would require “record creation” in order to disclose; similarly rejecting the FBI’s argument that production would “involve a ‘gargantuan’ effort,” as creating screenshots would not be “any different [than] the burden imposed in myriad FOIA cases.”
RM: Poor Judge Moss. This Shapiro case is interminable. It was filed in 2013, and it’s still going. The FBI is in the midst of producing the screenshots of the FDPS, as ordered by the court. As for the outcome, a few observations: I think the agency misfocused by arguing about “agency control.” There’s no serious basis, in my mind, to conclude that the contents and structure of an FBI-owned and operated database, which is regularly used by FBI employees for officials purposes, is not under agency “control.” It would be more persuasive to focus on the antecedent definitional question of whether the database tabs are even “records.” The FBI tried to do that by highlighting the supposed ephemeral quality of certain types of information in the FDPS, but it’s arguments didn’t land. Again, part of that is likely because of Judge Moss’s impression that there was some conflation going on with the control argument. Also, many types of records are dynamic in agency recordkeeping systems—that point by itself means very little, I think. The strongest argument was really the one about “record creation.” And that’s been a hot topic lately. There were big cases in the Second and Fifth Circuit not too long ago—ACLU Immigrants’ Rights Project v. ICE and Rutila v. Dep’t of Transportation, respectively. Rutila, especially, could have been persuasive authority, as it rejected the notion that an agency had to create screenshots of “tabs” in an agency database. Certainly, the line between record creation and the exporting of databased-stored information is a hard one. But the FBI’s downfall was likely Judge Moss’s math about the burden of complying, which he determined to be in simlar to with “regular” FOIA processing.
AB: I wholeheartedly agree that these database cases are challenging—especially so for me because I am not savvy with technology beyond the 1980s. Atari anyone? The FBI gave it the old college try on the “records creation” argument, but the court viewed this as a non-burdensome matter of extracting existing data/information. So, even if cited by the FBI, I am skeptical the court would be persuaded by the Fifth’s Circuit’s 2023 decision in Rutila, which held that DOT was not required to take screenshots “of information it does not store.”
6. Freedom Coal. of Doctors for Choice v. CDC (N.D. Tex. Jan. 5, 2024) — determining that plaintiff’s request for 7.8 million free-text responses to agency’s COVID-19 vaccine safety monitoring system would not be unreasonably burdensome for agency to process, because: (1) the volume of the responsive texts would yield between as little as 83 thousand pages and at most 650 thousand pages; (2) CDC conceded that 93 percent of the responses would require no redaction at all; and (3) any necessary redactions of personal identifying information pursuant to Exemption 6 would be “simple” and “capable of automated assistance”; and (4) plaintiff established a “substantial public interest” in the records.
AB: The Northern District of Texas has become a popular venue for plaintiffs to bring COVID-related FOIA claims, because judges there have not hesitated to order HHS components to process voluminous records at seemingly breakneck speed. In this instance, the court decided that a “mere” 83 thousand pages or “at most” 650 thousand pages would be reasonable to process in one year with monthly quotas. Plaintiffs are wasting no time citing this decision in support of their own requests seeking production of voluminous records from other agencies. In response, defendant-agencies should certainly point out that the judge’s ruling was expressly guided by a substantial public interest concerning a once-in-a-century event. Notably, more reasonable judges in the DDC have granted stays to HHS based on the enormous processing burdens imposed by N.D. Texas See, e.g., Informed Consent Action Network v. FDA (D.D.C. Nov. 20, 2024).
RM: Yes, Judge Kacsmaryk has gotten a reputation for these sort of scheduling orders that impose huge burdens on agency components, especially at HHS. I don’t know if I have much to add, except to emphasize that the case implicated significant public-interest reasons for expediting production, and that the court took a fairly detailed look at the burden its order would impose. I don’t think things would have turned out so favorably for the requester if the agency hadn’t admitted that nearly all of the records (93%) would require no redaction, or if the remainder weren’t “simple” in terms of the application of Exemption 6.
7. Louise Trauma Ctr. v. Wolf (D.D.C. Sept. 18, 2024) — ruling that: (1) plaintiff was eligible for attorney’s fees because its litigation was the “most natural explanation for the sudden movement” on plaintiff’s FOIA requests; (2) plaintiff was also entitled to attorney’s fees, notwithstanding the government’s seemingly valid argument that plaintiff was “a front for the collection of attorney’s fees”; and (3) no award was warranted, however, because plaintiff’s time records were “vague, inadequately descriptive, or made in error,” the request was “grossly out of line with requests in similar cases, reflecting an extraordinary lack of billing judgment,” and plaintiff has been “admonished repeatedly” for engaging in the “same unreasonable and improper billing practices.”
AB: You have to hand it to plaintiff’s counsel for blowing this one. How difficult is it to keep adequate billing records? Interestingly, in another fee-related decision issued three weeks earlier, the court noted USCIS’s seemingly valid concerns that Louise Trauma Center’s executive director and its counsel shared the same last name, that plaintiff does not operate as a non-profit or disseminate any information beyond posting records on its website, and that plaintiff or counsel may have a personal incentive to use FOIA to obtain fees. Although Louise Trauma Center was awarded fees in that case, the amount was slashed by 61 percent for unreasonable billing. Better than nothing!
RM: As our summary paragraph notes, the court went so far as to highlight the government’s complaint that this particular requester is a “front for the collection of attorney’s fees.” Although that didn’t prevent the judge from concluding that the requester was entitled to fees, I do think it colored the analysis—including on the reasonableness of the fee aware—and perhaps rightfully so. On a related note, Judge McFadden has claimed repeatedly that non-profit groups are ruining FOIA. That particular claim has been debunked, in my mind, despite its persistence among some jurists. (See, e.g., here, here, and here!) But groups like this requester help perpetuate the stereotype.
Honorable Mentions
Reclaim the Records v. U.S Dep’t of State (S.D.N.Y. Aug. 7, 2024); Reclaim the Records v. U.S. Dep't of State (S.D.N.Y. Aug. 26, 2024); and Reclaim the Records v. U.S. Dep’t of State (S.D.N.Y. Sept. 9, 2024) — ruling that State Department performed adequate searches for indices, lists, or extracts of various information that could not be extracted from computer systems with technical limitations, and that compiling the requested records would either entail the creation of new records and/or be unduly burdensome.
McWatters v. ATF (D.D.C. Feb. 16, 2024), aff’d (D.C. Cir. 2025) (per curiam) — on renewed summary judgment, finding that ATF properly relied on Exemption 7(C) to withhold portion of a tape recording made by one of the 100 victims of a Rhode Island nightclub fire in 2003; crediting ATF declaration that faint human voices could be heard on the recording and that surviving family of the deceased had privacy interests even if voices could not be attributed to specific victims; further, rejecting plaintiff’s asserted public interest as nothing more “than having the information for its own sake,” similar to case involving the recording of the last minutes of NASA’s Challenger shuttle.
Insider, Inc. v.. GSA (D.C. Cir.) — affirming district court’s decision that GSA properly invoked Exemption 6 to withhold the names of several low-level members of President Trump’s and Vice President Pence’s outgoing transition teams; reasoning that such individuals were not government employees, disclosure would not shed light on GSA activities, and the mere possibility that such individuals might be aware of government conduct was too speculative to qualify as a public interest.
