FOIA Advisor

FOIA Commentary (2017-2024)

Commentary: 2020 FOIA Metrics

FOIA Commentary (2017-2024)Allan BlutsteinComment

On May 13, 2021, the Department of Justice’s Office of Information Policy released a summary of the annual FOIA reports prepared by federal agencies for fiscal year 2020. The staff of FOIA Advisor—Allan Blutstein (AB), Ryan Mulvey (RM), and Kevin Schmidt (KS)—reacts to OIP’s summary.

AB: I used to wait impatiently for these summaries, but in recent years DOJ has made all of the raw data available months beforehand on the website FOIA.gov. If you have not seen that data, which we highlighted here, you will not be terribly surprised to learn that the total number of requests submitted in fiscal year 2020 decreased and that the government’s backlog increased—each likely due in part to the pandemic. Those looking for good news might point to a decrease in the average processing time for “simple requests” from 39.3 days in FY 2019 to 30.23 days in FY 2020. Indeed, OIP touts that fact in the conclusion of its report. But one could rain on that parade by noting that last year’s processing time is still worse than in fiscal years 2012 through 2018. As a parting (positive) comment, I would include U.S. Customs & Border Patrol as one the agencies that performed well despite the pandemic; it reduced its backlog of requests from 10,466 to 1729. Keep up the good work on all fronts.

KS: What stood out to me was the 10 percent increase in staff devoted to FOIA in FY2020 (5,559 full-time FOIA staff) compared to FY2019 (5,002 full-time FOIA staff) along with the 12 percent increase in cost of FOIA related activities. I’m not sure where these increases came from, but hopefully more progress can be made on requests and appeal backlogs in FY2021 if the staffing stays at this level or increases.

RM: Honestly, I was rather surprised by the drop-off in requests, especially given the contentious nature of the federal government’s COVID response. I would have expected some decrease, but not an effective return to FY 2016 levels. I suppose this data could confirm a sort of “Trump” surge in requests over the last four years; a lot of requesting saeems to have been driven by political interests, above all, rather than the sort of “natural” trend we’ve witnessed over the past decades. Opponents of the previous Administration may have seen FOIA as a useful tool for political ends. (I reserve opining on whether that’s a good thing.) For example, if you dig into the data that’s downloadable from FOIA.gov, you can see the agencies where there is the greatest difference in new requests between FYs 2019 and 2020: EPA, USPS, SSA, DOJ, and NARA (very significant spread, there). It would take a bit more investigation to confirm this hypothesis, but the end of the Trump Administration may be more to blame, as it were, than the pandemic.

I also agree with you, Kevin, that the increase in staffing levels, however slight, is interesting and promising. There’s a lot that could be said for agencies devoting more time and energy to FOIA. It seems that they’ve done so, to some extent, and without anything specific in the way of dedicated line item appropriations from Congress.

AB: Just when I thought I was out, they pull me back in. I am skeptical that the decrease in requests had much to do at all with Trump fatigue or the end of the Administration. The fiscal year ended one day after the first presidential debate, about six weeks before the election, and roughly four months before Trump left office. A once-in-a-lifetime pandemic likely sapped the time, ability, and energy of FOIA requesters more than politics. One more comment : if an 8-percent decrease in requests is so significant as to engender disbelief, Ryan, why is a 10-percent increase in staffing so insignificant as to warrant characterizing it as “slight.”

KS: If we are turning to evidence-optional speculation on the drop off in requests in FY2020, I’m going to cut in here before Ryan can jump in. Let me remind everyone of the absolutely insane (fiscal) year that was 2020. It included a presidential impeachment, COVID, passage of the CARES Act, national protests (and some riots), and a Democratic primary. For the most part, these events that drove the news during the year didn’t have a direct nexus to FOIA and many of them required the time and energy of politically active and followers of politics (journalists). Combine that with Trump fatigue and the once-in-a-lifetime pandemic, and you get an 8-percent decrease in requests.

AB: Here we go again. A decrease in requests because of Trump fatigue? Wrong. None of the national events you cite, Kevin, were unconnected to FOIA; records indicating how the Executive Branch handled or reacted to them unquestionably remained of interest. So there. Stuff your request for evidence in a sack.

RM: Let’s talk evidence then. Consider the case of NARA, which I already mentioned witnessed the largest decrease in new requests in FY 2020.

In FYs 2017, 2018, and 2019, the Archives received on average just north of 60,000 requests. That average itself is noticeably elevated compared to earlier annual totals. Maybe the private email controveries towards the end of the Obama White House are to blame, along with continued (and fierce) interest in problematic records management practices under President Trump? Not to mention other Trump kerfuflles, such as the 2017 Women’s March photo scandal. In any case, in FY 2019, NARA received approximately 67,500 requests. Yet in FY 2020, it only received 25,738 requests. That is a tremendous decrease—62%—that may not be explained by the pandemic alone. Or so I respectfully submit.

If you look at data from FOIA.gov, no other major agency, as far as I can tell, has such a huge shift in number of FOIA requests submitted between 2019 and 2020. The next largest decrease is the Social Security Adminitration (35%), followed by Interior (25%), OPM (23%), EPA (22%), FCC (18%), FTC (15%), and USDA (14%). I didn’t notice a huge uptick anywhere. When you glance at these agencies, they all have one thing in common: they were somehow engaged in an especially controversial aspect of the previous President’s policy agenda.

If the pandemic were the principal cause of the decrease in requesting, shouldn’t we expect to see similar trends on other agencies? Why did people care so much less about NARA? Let’s keep in mind that, the decrease of requests submitted to NARA between FYs 2019 and 2020 makes up roughly 61% of the total difference in FYs 2019 and 2020 across all agencies! Moreover, according to NARA’s Chief FOIA Officer Report, at least, these numbers do not reflect requests for archival records. So we’re not talking about researchers making in-person requests at the Archives, but requests for records under the actual legal control of the agency.

I could be wrong. I don’t think I’m misreading the numbers, unless NARA is omitting something. And I do grant there’s a variety of factors that could have influenced the requesting community, as Kevin suggests. Given how public discourse about FOIA and records management generally reached fever pitch under Trump, I’d still wager his departure has influenced requesting trends.

AB: It’s a good thing you’re not actually a gambler, Ryan. I contacted NARA about our debate and here’s the written explanation from NARA’s General Counsel and Chief FOIA Officer Gary Stern:

The significant drop in the number of FOIA requests that NARA received in FY 20 (25,738 in FY 20, as compared to 67,466 in FY 19) was due to the impact of the COVID-19 pandemic. All of NARA’s facilities were forced to operate at a significantly reduced capacity. In particular, NARA’s National Personnel Records Center (NPRC), which receives the vast majority of NARA’s FOIA requests, was limited in its ability even to log in most of the FOIA requests; instead, the NPRC concentrated principally on responding to emergency requests for records concerning medical and burial claims, which are handled outside of the FOIA process.

The Academy Award-winning film Annie Hall includes a glorious scene in which Woody Allen brings out Marshall McLuhan for a cameo appearance to settle an argument—in Woody Allen’s favor—about a professor’s views of Mr. McLuhan’s work. Gentlemen, I believe I’ve had my “Marshall McLuhan” moment here. If life were only always like this.

FOIA commentary: Top FOIA decisions, July 2020-Dec. 2020

FOIA Commentary (2017-2024)Allan BlutsteinComment

On July 21, 2020, FOIA Advisor lawyers Allan Blutstein (AB) and Ryan Mulvey (RM) discussed six FOIA decisions that stood out to them as the most significant of the first half of the year. Today they discuss eight decisions of note that were issued during the second half of 2020. Relatedly, Ryan will be co-teaching an American Society of Access Professionals webinar on FOIA case law developments with the redoubtable Richard Huff on January 26, 2021. (Registration closes on January 22th!)

Ctr. for Investigative Reporting v. DOJ (9th Cir. Dec. 3, 2020) -- in a 2-1 decision, reversing and remanding district court’s decision and holding that data requested from agency’s Firearms Trace System was not protected by a rider to a Consolidated Appropriations Act because it failed to cite to Exemption 3; remanding case for further proceedings to determine whether agency could reasonably search database for requested records; Everytown for Gun Safety Support Fund v. ATF (2nd Cir. Dec. 23, 2020) -- reversing district court’s decision and holding that data requested from agency’s firearms database was protected from disclosure pursuant to a congressional rider even though that rider did not specifically cite Exemption 3.

AB: A pair of split Circuit court decisions within three weeks was an exciting way to end the year. Barring congressional intervention or an en banc reversal by one of the courts, we may be looking at a U.S. Supreme Court clash in 2021. I do not have a strong opinion on how SCOTUS would or should rule on the disputed provision of the Open FOIA Act of 2009, but I lean towards the Second Circuit’s view.

RM: It would be quite something to get three consecutive years of FOIA at the Supreme Court! Apart from the particular issue of the Firearms Trace System, the broader questions of how appropriations riders function, whether they can qualify as withholding statutes, and how or when subsequent riders impliedly repeal their predecessors are all vitally important for understanding the proper operation of 5 U.S.C. § 552(b)(3)(B). (As an aside, I think these cases are a lesson for the imprudence of Congress trying to prohibit disclosure indirectly through appropriations legislation.) That being said, I agree with the Ninth Circuit; I don’t see how the 2008 and 2010/2012 riders at issue are mutually consistent and given the passage of the 2010 rider only two months after the Open FOIA Act, it is difficult to imagine Congress didn’t know what it was doing. The requirement for new withholding statutes to reference Exemption 3 needs to be taken seriously, and the Second Circuit’s decision renders that provision of FOIA entirely nuggatory. (I also think the Ninth Circuit decision offers a useful reminder to agencies that retrieving and producing data stored in a database does not entail creation of a “new” agency record. This has been established in other jurisdictions, but I still see agencies fight against the rule.)

AB: Contrary to your suggestion, it is now entirely unclear whether the “requirement for new withholding statutes to reference Exemption 3 needs to be taken seriously.” We’ll see, as the Zen master in Charlie Wilson’s War was fond of saying. Congress could moot this issue with new legislation, of course, and I hope it would step in if the government ultimately loses.

Machado Amadis v. DOJ (D.C. Cir. Aug. 21, 2020) -- affirming, in most relevant part, district court’s decision that OIP properly relied on Exemption 5’s deliberative process privilege to withhold portions of staff’s appeal recommendation forms, and finding that agency reasonably explained why statute’s foreseeable harm provision was satisfied—i.e., agency had considered "information at issue’” and concluded that disclosure “‘would’ chill future internal discussions.”

RM: The much awaited “foreseeable harm” decision from the D.C. Circuit!  In many ways, I expected this opinion to be much worse (from the requester’s perspective), considering the facts and arguments presented to the court.  I think the decision leaves many questions unanswered.  The good news?  The Circuit appears to recognize that the foreseeable harm standard is not merely a codification of previous practice but imposes some sort of new burden on agencies.  “Generalized” assertions of expected harm—particularly vis-à-vis “chilling” of agency deliberations—are inadequate.  But we still don’t have clarity on the process of identifying interests protected by certain exemptions, or exactly how we should distinguish the harms necessary to establish use of a privilege (e.g., the “technical” requirements under the second part of Exemption 5 to raise the deliberative-process privilege) from further “foreseeable” harm that would result from disclosure of the specific records at issue in any given case.  As other cases work their way up on appeal, including those cases implicating Exemption 4 post-Argus Leader, I think we will see some more useful precedents established.

AB: What would be useful is if Congress amended the ambiguous foreseeable harm provision instead of leaving it to the courts to figure out. Better yet, Congress should repeal it. Agencies have a challenging enough task in determining whether records fall within an exemption. If that burden is met, nothing more should be required; the disclosure of exempt information is inherently harmful. Asking FOIA personnel to become prognosticators like the Oracle at Delphi would be folly and an unwarranted gift to the plaintiff’s bar. Thankfully, the D.C. Circuit set a standard here that doesn’t appear to be all that difficult for agencies to meet.

Whitaker v. Dep’t of Commerce (2nd Cir. Aug. 15, 2020) -- affirming district court’s decision that : (1) the First Responder Network Authority, an independent entity within the Department of Commerce’s (DOC) National Telecommunications and Information Administration (NTIA), was not subject to FOIA; and (2) DOC and NTIA properly declined to search for requested records because such searches would have been futile, adopting D.C. Circuit’s standard.

RM: I’m a little surprised this case was appealed. It’s difficult to argue that FOIA is not part of “chapter 5 of title 5,” notwithstanding the fact that FOIA isn’t “commonly referred to as” the APA. I also didn’t understand the requesters argument that Exemption 3, as amended by the Open FOIA Act of 2009, should apply to Congress’s exclusion of particular entities from the meaning of an “agency,” for FOIA purposes. The only aspect of the Circuit’s opinion which does give me pause is its analysis of futility. At least when an agency is asked to search for records involving communications with a non-agency, I think it’s important to avoid eliding the question of the existence of “records” (NB: not “agency records”) and agency control. Based on the facts here, it seems the Department of Commerce never had copies of the requested records because it wasn’t involved in the relevant aspects of FirstNet’s operations. Fine. I would not, however, want an agency to be able to argue that a search would be futile because all potentially responsive records in its possession would be categorically outside of its legal control. As far as I’m aware, nearly all the case law on congressional and presidential records, for example, contemplates a search and review before arguing over control. But I digress…

AB: The appellants were grasping at straws on the agency issue (the plain meaning of the statutes at issue were against them, as you pointed out), but I do not fault them for doing everything in their power to try to obtain the requested records. The loss on the futility issue will have a more lasting impact on future FOIA requests, and your hypothetical case is an interesting one. But it’s not this case, so I’ll let it go.

Nightingale. v. USCIS (N.D. Cal. Dec. 22, 2020) -- finding that USCIS, ICE, and DHS “have a pattern of unreasonable delay in responses to A-File FOIA requests,” and permanently enjoining defendants from failing to adhere to statutory deadlines for adjudicating A-File FOIA requests; further ordering defendants to make determinations on all A-File FOIA requests in backlog within 60 days.

RM: An interesting decision in a class-action lawsuit certified at the end of 2019.  It used to be, I think, that failure to abide by statutory deadlines, in-and-of-itself, was a non-starter for a policy-or-practice claim.  There have been a few exceptions in recent years, and this is another step in that direction.  I’m sure immigration lawyers are happy!  I am somewhat troubled, though, by the reference to the importance of A-Files for immigration proceedings—true, no doubt—as if that were probative of an ongoing failure to abide by the terms of the FOIA, or somehow justifies the extraordinary injunctive relief ordered by the court.

AB: A number of agencies with longstanding backlog problems might have good cause to be concerned about this decision. For those who are curious, like I was, about the challenges of processing FOIA requests for A-Files, I recommend reading the compliance assessment report issued by the Office of Government Information Services in February 2018. An A-File is typically about 220 pages long and the largest A-File ever processed was approximately 40,000 pages. No wonder USCIS resisted the relief sought here.

Seife v. FDA (S.D.N.Y. Oct. 6, 2020) -- concluding that FDA properly relied on Exemption 4 to withhold records pertaining to its accelerated approval of a muscular dystrophy drug; noting that statute’s foreseeable harm provision applied to Exemption 4 and was met in this case.

AB: This opinion might contain the most extensive analysis of the foreseeable harm requirement in the context of Exemption 4. The court, unfortunately, sidestepped the government’s threshold argument that the disputed information was protected by the Trade Secrets Act and therefore not subject to the foreseeable harm requirement. I eagerly await a decision that adjudicates that issue—hopefully, in the government’s favor.

RM: Before Seife, I don’t think any of the post-Argus Leader cases dealing with the foreseeable harm standard grappled with the interrelation of Exemption 4 and the Trade Secrets Act. But I’m not sure the question is so easily decided in favor of the government. It really depends on which exemptions are actually “discretionary” and whether an agency’s foreseeable harm analysis is extricable from its exemption analysis. There is some caselaw to cast doubt on the government’s claim that Exemption 4 is obviously “non-discretionary.” Further, under current precedent, the Trade Secrets Act should only come into play, if at all, once information is decidedly protected by Exemption 4. (And I think there is honest diversity of opinion on the proper role of 18 U.S.C. § 1905 under FOIA.) If foreseeable harm is part of the exemption analysis, and not a secondary consideration—even though we often speak of it as an “additional burden” on the agency—then the Trade Secrets Act cannot be the legal prohibition on disclosure contemplated by Section 552(a)(8)(A)(i). Relatedly, I think most people tend to read “disclosure is prohibited by law” in this part of the FOIA as effectively referring to Exemption 3. That interpretation would seem to avoid the possibility of agencies gaming the system and defeating the foreseeable harm standard, say, by trying to eliminate the possibility discretionary disclosure by regulation. I recognize, though, that the call is a close one, and certainly the legislative history is mixed in resolving any interpretive ambiguity. The Ninth Circuit is scheduled to deal with foreseeable harm and Exemption 4 in the coming year, in Evans v. Department of Labor. Seife is on appeal, too, so perhaps we’ll see another Second/Ninth split!

WP Co. v. SBA (D.D.C. Nov. 5, 2020) -- ruling that SBA failed to show that Exemptions 4 and 6 protected the names of loan recipients and amounts borrowed from the Paycheck Protection Program and Economic Disaster Loans program. In reaching its decision, the court noted that the SBA had notified loan applicants that such information would be disclosed upon request.

RM: This case, and the underlying issue of confidentiality of PPP applicant data, attracted a lot of attention in the press last summer. That may have been due to the Administration’s changing position throughout the course of the litigation. I think the court’s rejection of Exemption 4 was well-reasoned. The agency failed to offer convincing evidence of a direct link between PPP loan data and actual business payrolls. Even then—and more determinative, in my mind—the agency expressly notified potential borrows that loan information could be disclosed in response to a FOIA request. The subsequent rebuff of the government’s motion for a stay was pretty embarrassing, too.

AB: This decision did not break new legal ground, but I was also impressed by how thoroughly the court dismantled the government’s exemption claims, especially Exemption 4. It makes me wonder whether anyone outside of SBA —e.g., loan recipients, Treasury, or the White House—asked the agency to withhold the requested information.

Campaign for Accountability v. DOJ (D.D.C. Sept. 11, 2020) -- rejecting plaintiff’s allegation that all legal opinions of Office of Legal Counsel must be affirmatively disclosed under FOIA’s “reading-room” provision, but concluding that OLC opinions “that resolve disputes between agencies” plausibly qualify for disclosure.

AB: In 2019, after this lawsuit was brought, the D.C. Circuit opined in a separate case that FOIA’s reading-room provision did not require affirmative disclosure of all of OLC’s legal opinions. As such, the plaintiff had virtually no chance of prevailing on that claim here. It is notable, however, that the plaintiff survived the government’s motion to dismiss on a subset of the OLC’s opinions. We’ll have to wait until 2021 or later to see which party ultimately prevails on summary judgment.

RM: Agreed. I thought the court’s rejection of OLC’s argument that proactive disclosure only applies to records concerning the regulation of private parties to be important. That argument was so removed from the statutory text and attendant caselaw. It’s also refreshing to see OLC taken to task for trying to argue, again, that it merely provides “opinions,” as if OLC’s advice were incapable of becoming an agency’s working law. I’m very interested to see where this case ends up going on the merits.

Commentary: Top court decisions of 2020 to date

FOIA Commentary (2017-2024)Allan BlutsteinComment

Federal courts have adjudicated numerous FOIA cases since the beginning of 2020. FOIA Advisor lawyers Allan Blutstein (AB) and Ryan Mulvey (RM) select the decisions that have stood out to them (in no particular order). Please note in this context that Ryan will be co-teaching a course with Richard Huff about recent FOIA court rulings at ASAP’s annual training conference on July 29, 2020.

Doyle v. DHS (2nd Cir.) — affirming district court’s decision that visitor logs for White House and President Trump’s Mar-a-Lago home are not “agency records” of the U.S. Secret Service, relying on D.C. Circuit’s 2013 opinion in Judicial Watch v. U.S. Secret Service.

AB: I agreed with the result, though I would have enjoyed a Supreme Court showdown had plaintiff prevailed. Query whether any plaintiff will try to forum shop in the Ninth Circuit or elsewhere; it would be an uphill battle.

RM: In my mind, there was never any serious chance the Second Circuit would hold these visitor logs to be agency records. The D.C. Circuit addressed the control question for presidential visitor logs in a fairly persuasive way, both under its modified control test and “regular” Burka test. There have been a couple of attempts to get Trump visitor logs, and similar records; they strike me more as efforts to score political points in the press rather than serious FOIA challenges. Your comment about forum shopping is interesting. The Ninth Circuit has given us some unique FOIA decisions in recent years—think of Rojas v. Federal Aviation Administration, which is scheduled for rehearing en banc after a panel rejected the consultant corollary for Exemption 5’s threshold requirement. I concur it’d be an uphill battle.

Cause of Action Inst. v. DOJ (D.D.C.) — concluding that: (1) DOJ properly withheld certain questions to and responses of congressional witnesses as “non-responsive” even though all material appeared in a single document, but that agency could not treat sub-questions and responses thereto as distinct records; and (2) plaintiff did not have standing to challenge legality of Office of Information Policy’s guidance on definition of a “record.”

RM: I must disclose that my colleagues and I brought this case, and we’re currently on appeal to the D.C. Circuit.  I think it’s time the Circuit cleaned up the mess it created in 2016 with American Immigration Lawyers Ass’n v. Executive Office for Immigration Review.  By itself, AILA is an important decision, and it correctly ended the pernicious practice of using “non-responsive”/”outside the scope” as an effective tenth exemption.  But the panel left unanswered the antecedent question of what qualifies as a “record”!  It only offered dicta mentioning the Privacy Act and suggesting that agencies could use their understanding of the subject-matter of a request as a basis for segmenting “records.”

OIP latched on to this dicta when it created guidance on defining a “record” in early 2017.  I believe that guidance conflicts with the statutory language, as well as long-standing principles of FOIA law (e.g., records must pre-exist any given request) and canons of statutory interpretation.  In terms of application of the guidance in this case, I’m not sure why Judge Jackson didn’t think that QFRs maintained as a single PDF file, under one title, and with consecutive pagination qualified as a unified record.  The court’s sua sponte dismissal of the policy-and-practice claim for lack of standing was even more bizarre.  At the least, on the definition of a “record,” we’ve been seeing the district courts go different ways, so it’d be nice to get some clarity.

AB: As I have previously stated in this forum, I have little enthusiasm for the position that requesters should receive more information than they have asked for. I understand that the statute concerns “records” not “information.” I also understand that it might be more attractive to make a federal case out of the issue than to simply make a follow-up FOIA request for the non-responsive material. If clarity is needed, I’d prefer Congress to act; the courts are winging it.

Grand Canyon Trust v. Bernhardt (D.C. Cir.) — affirming district court’s decision that requester was ineligible for attorney’s fees because its lawsuit did not cause a voluntary or unilateral change in agency’s position; further ruling that the proper standard of review for causation was “clear error,” not de novo. In a concurring opinion, Judge Randolph stated that FOIA’s 2007 amendment did not restore the “catalyst theory” because the amended provision “requires only correlation not causation.”

AB: What piqued my interest here is the concurring opinion, which in my view makes a reasonable case for interpreting the statute more favorably for plaintiffs. I wonder whether any district courts within the Circuit will adopt it. I candidly would prefer requiring judicial relief on the merits as the sole criterion of fee award eligibility, but I am not holding my breath for Congress to re-amend the statute at plaintiffs’ expense.

RM: Judge Randolph’s exposition of the word “unilateral” makes his theory of “correlation not causation” particularly persuasive.  I have not had an opportunity to explore the legislative history.  Judge Randolph did not appeal to it, and he should have if it provided any clue about Congress trying to reverse the Supreme Court on the “catalyst theory,” as opposed to simply liberalizing the standard for fee eligibility.

Ctr. for Investigative Reporting v. DOL (N.D. Cal.) — ruling that Department of Labor improperly relied on Exemption 4 to withhold certain work-related injury form received by OSHA from employers, because agency failed to show that records were both customarily and actually treated as private by owners and provided to the government under an assurance of privacy.

RM:  This is an important decision because it carefully distinguishes whether a submitter of information merely “considered” information to be confidential or actually “kept and treated” the information as confidential.  As I have expressed in previous commentary here and elsewhere, I have problems with the new FMI standard as it seems to open the door to submitters trying to game the system and force a subjective standard for confidentiality.  Policing the factual basis for confidentiality claims is a good step towards avoiding that result.  I was glad to see the court take the agency’s assurances of confidentiality (or lack thereof), as well as relevant OIP guidance, so seriously.  The Supreme Court did not resolve the question of whether assurances either way were a necessary part of the (b)(4) analysis.

AB: I agree with you. Although the court sidestepped the parties’ arguments as to whether the foreseeable harm was satisfied (or even applied) in this case, that issue inevitably will reach the Circuit courts in the near future.

RM:  Since we’re talking about Exemption 4, and you’ve mentioned foreseeable harm, I want to reference another lawsuit brought by the same requester: Center for Investigative Reporting v. Customs & Border Protection (D.D.C. 2019), which was decided at the very end of last year.  I find two aspects of the decision intriguing.  First, the court treated the new FMI standard—i.e., (1) “customarily and actually treated as private” and, perhaps also (2) provided “under an assurance of privacy”—as effectively interchangeable with the D.C. Circuit’s longstanding test from Critical Mass for voluntarily submitted information—i.e., “of a kind that would customarily not be released to the public.”  I’m not convinced that FMI and Critical Mass articulate the same test, but I understand why a court would draw on Critical Mass and its progeny as useful guidance. 

Second, and more importantly, the court’s decision to extend FOIA’s “foreseeable harm” standard to Exemption 4 is groundbreaking, and it will be interesting to see if other courts join the bandwagon.  I suspect most agencies assume that Exemption 4 is a mandatory exemption and, accordingly, foreseeable harm analysis is unnecessary.  This court’s attempt to reintroduce a National Parks-like “competitive harm” requirement really tends to undo what I think the Supreme Court sought to do in FMI.  Exemption 4 and “foreseeable harm” are percolating up to both the Ninth and D.C. Circuits.  So we need to see how things turn out!

Mabie v. EOUSA (S.D. Ill.) — dismissing plaintiff’s FOIA claims as sanction for using profane, belligerent, and abusive language when speaking to opposing counsel, leveling unfounded accusations and threatening opposing counsel, insulting the judge and Court employees, and impugning the integrity of judicial proceedings.

AB: This opinion left me slack-jawed. I do not recall ever seeing a FOIA case dismissed because of a litigant’s misbehavior, but if anyone ever deserved it, Mr. Mabie did.

RM: Indeed. It seems that Mr. Mabie is a troubled individual with a track record of frivilous litigation. The recent opinion out of the Eastern District of Pennsylvania in Manatt v. Department of Homeland Security, which ordered senior agency officials to testify at a special hearing concerning repeated failure to comply with court deadlines, left me similarly astonished, albeit for different reasons. I think there has been an noticeable uptick of judges willing to hold intransigent agencies to task—consider orders earlier this year for in camera review of the Mueller report and deposition of former Secretary of State Hillary Clinton.

AB: Since you mentioned this week’s decision in Manatt, it bears mentioning that the senior DHS official who was berated by name in this case is the former director of the Office of Government Information Services.

Sai v. TSA (D.D.C.) — on renewed summary judgment, concluding that: (1) agency demonstrated that it properly withheld records pursuant to Exemptions 3 and 5 (deliberative process privilege); (2) agency failed to demonstrate that it searched all reasonable locations or employed reasonable search terms and proper time frame; (3) agency was not required to organize records into discrete PDF files,but that agency failed to show how it would be unduly burdensome to create “irreversible redactions within fully digital, non-rasterized PDFs” outside of FOIAExpress program that agency typically used.

RM: I always enjoy opinions from cases filed by Sai. They often touch on the intersection of FOIA and technology, and this case is no different. I appreciated Judge Moss’s discussion of the E-FOIA amendments and the meanings of “form” and “format.” I was somewhat surprised that he held the agency to such a stringent standard for demonstrating that it was unduly burdensome or technically impossible to process and disclose fully digital non-rasterized PDF files outside of FOIAXpress. This could be a helpful precedent for requesters.

AB: I rarely enjoy his cases, or at least the opinions, because they are lengthy and highly technical. Feel free to write the summaries of his cases going forward. Curious that the court relied upon a Ninth Circuit ruling in evaluating the “unduly burdensome” standard (are there no D.C. Circuit cases?), but it was reasonable for the court to give the agency another opportunity to fill in the perceived gaps.

Commentary: 2019 FOIA metrics

FOIA Commentary (2017-2024)Allan BlutsteinComment

On June 1, 2020, the Department of Justice’s Office of Information Policy released a summary of the annual FOIA reports prepared by federal agencies for fiscal year 2019. The staff of FOIA Advisor—Allan Blutstein (AB), Ryan Mulvey (RM), and Kevin Schmidt (KS)—reacts to OIP’s summary.

AB: Kudos to the government for its 7.9 percent reduction in the overall FOIA backlog, but that progress will likely be wiped out and then some by the COVID-19 pandemic, which OIP implicitly acknowledges in the conclusion of its report. Two other statistics stood out to me. First, the average processing time for “simple” requests climbed from 30.22 days to 39.3 days, a nearly 30 percent increase. Second, the average time to adjudicate requests for expedited processing increased from 10.27 days to 14.82 days, the second-slowest time in the past nine years.

KS: There appears to be some important progress here in backlog reduction and number of requests processed, but COVID-19 is going to wipe out any gains made in recent years through no fault on the part of agencies and employees. One possible upside is that agencies that are behind on technology may be forced to invest in upgrades to catch up and/or prepare for a future with more remote FOIA employees. Other than the items mentioned by Allan, the administrative appeals backlog increasing by 6.72% caught my eye.

RM: I was surprised there wasn’t any increase from FY 2018 to FY 2019 in the number of requests filed, but actually a slight decrease. And I was similarly impressed by the progress made on the backlog—though you’re both right that the pandemic will likely seriously impact these gains. I’d like to see agencies reporting on when they withhold parts of records as discrete “Non-Responsive Records.” As we all know, this is a “hot” topic at the moment. OIP gives us data on “No Records” and “Not Agency Record” determinations; it seems reasonable to start recording the use of “Non-Responsive Record,” now that we’re in a post-AILA world. Finally, as I believe I’ve mentioned in the past, I’d like to see some effort to separate out first-party/Privacy Act requests from the data, whenever possible. I don’t find it especially helpful to know DHS is receiving 47% of all FOIA requests, when the agency defines “FOIA Request” broadly to include Privacy Act requests.

AB: To Ryan’s latter point, the annual agency FOIA reports used to include the first-party requests of all agencies, which would often raise the total number of requests above 20 million—most of which were submitted to the Social Security Administration (where, coincidentally, I first worked after law school and encountered FOIA/PA). In the wake of the OPEN Government Act of 2007, however, DOJ instructed agencies to include Privacy Act requests in their fiscal year 2008 reports only if the FOIA was utilized in some way in responding to the request. History lesson for the day.

Commentary: Musings about FOIA during the pandemic

FOIA Commentary (2017-2024)Allan BlutsteinComment

When the FOIA Advisor staff last convened to chat, the Department of Justice had just released the government’s annual FOIA data for fiscal year 2018 and we were looking forward to Sunshine Week. Concerns about the coronavirus scuttled all but a few Sunshine Week events, however, and now FOIA personnel and requesters alike are largely working from homeFOIA Advisor staffers Allan Blutstein (AB), Kevin Schmidt (KS), and Ryan Mulvey (RM) share their thoughts about these exceptional circumstances.

AB: We have seen reports of significant processing delays at several agencies, notably at the FBI and the State Department, but FOIA operations at most agencies appear to be functioning. Requesters who have decried delays of any sort during this pandemic should not be taken seriously. No agency should endanger the lives of their employees in order to process FOIA requests, regardless of the subject matter of the request. Impatient FOIA requesters retain the right to file lawsuits at the usual intervals, unlike requesters in some states that have tolled or extended agency response deadlines.

KS: I agree with Allan that delays are inevitable. The time and energy worrying about delays is better spent fighting the attempts to classify coronavirus records and to eventually uncover any mistakes so we can figure out how to fix them in the future. Also of note, the Congressional Research Service recently published a survey on how thirteen agencies are processing FOIA requests. Eight agencies changed their method for receiving requests and surprisingly, only four specifically mentioned delays in response times.

AB: I should add that I do disagree with the FBI’s decision to disable its eFOIA portal. Even if the agency cannot process records remotely, it should not make it more difficult for requesters to submit requests. Perhaps one silver lining to social distancing in the FOIA context is that courts appear to be issuing more FOIA decisions recently. It could be just a coincidence, but I suspect it is because judges have more time to write opinions now.

RM: I also agree that delays are inevitable, given the reality of work life during the pandemic. However, I don’t think that sophisticated requesters and their allies in the transparency community are decrying delays as such. Rather they are highlighting the problems with how records may be processed, as Kevin suggests, or they are bemoaning how some agencies—particularly at the state and local level, it seems—have taken advantage of this crisis to justify shutting down FOI mechanisms, avoiding Open Meeting laws, and so forth. Open government shouldn’t suffer because of coronavirus. If anything, it is more important than ever for there to be robust transparency, especially when government is acting quickly and outside of “normal” procedural channels.

AB: What costs must be paid to ensure that open government “shouldn’t suffer during coronavirus”? Because unless employees are compelled to come into the workplace, open government will suffer to some extent. I can live with that because I do not believe in transparency at all costs. If agencies choose to err on the side of caution, I will not reflexively bemoan that choice as an underhanded attempt to take advantage of a crisis. And with all due respect to “sophisticated requesters,” a group to which we presumably belong, do not underestimate the scope of what they/we complain about.

RM: Yes, open government will suffer to the extent there are processing delays. But I’m more concerned with preserving the structures of open government. Closing online FOIA portals or shutting down processing altogether is completely unacceptable. And, if anything, this crisis highlights the need for serious reform at agencies that are incapable of processing records remotely, or which rely on retired agency employees for getting things out the door. To be sure, there are some records that will need to be processed on site, due to their sensitive nature. But those aren’t the sort of records sought by the average FOIA requester. As for setting the right balance, how do you feel about agencies policing whether employees are using approved methods of communication or properly preserving records? Because there is a valid concern that agency employees are at home using instant or ephemeral messaging, or personal email accounts, to conduct official business. Should we cut them slack because of coronavirus?

AB: We agree more than we disagree. As for your record preservation questions, I would not excuse the use of improper communication methods, e.g., personal email or personal instant messaging. But I am not losing sleep about it, either. Employees are aware of the rules. Hillary’s “damn emails” have not been forgotten, thanks or no thanks to Judicial Watch.

Stay safe everyone!

Commentary: Agency FOIA performance in FY 2019

FOIA Commentary (2017-2024)Allan BlutsteinComment

As we posted on Wednesday, the Department of Justice has published the FOIA data reported by agencies for fiscal year 2019. DOJ will summarize all of the data in the near future, but it did note that the federal government had received 858,952 requests in FY 2019, down slightly from FY 2018, when a record 863,729 requests were received. FOIA Advisor staffers Allan Blutstein (AB) and Kevin Schmidt (KS) share their thoughts on the available data.

AB: Last year I predicted that the number of requests in FY 2019 would set a second consecutive record, but perhaps we’ve reached a saturation point. FOIA lawsuits have also leveled off according to The FOIA Project, which reported 859 filings in calendar year 2018 and 849 filings in calendar year 2019. Barring a major catastrophe (e.g., coronavirus outbreak), I suspect the volume of requests in FY 2020 will again approach or surpass FY 2018.

Beyond this, I looked closely at the data of USDA, DHS, NARA, DOJ, and DOD, which all told receive more than 70 percent of all FOIA requests. Interestingly, two of these agencies experienced significant changes in their workloads from FY 2018 to FY 2019: (1) NARA, whose requests increased 27 percent (from 52,850 to 67,466); and (2) USDA, whose requests decreased 27 percent (from 36,547 to 26,458). I was also struck by the fluctuating backlogs of DOJ, NARA, and DHS over the same time period. Specifically, DOJ’s backlog increased 46 percent (from 17,411 to 25,558); NARA’s backlog increased 85 percent (from 1810 to 3359); and DHS’s backlog decreased 71 percent (from 53,971 to 31,454).

KS: Let’s start with the Interior Department, which was under fire in the media for its policy changes to its FOIA program. Surprisingly, Interior saw the number of requests drop from 8,402 in FY 2018 to 7,973 in FY 2019. Despite the slight downturn in the number of requests, Interior’s backlog increased by about 24 percent.

The number of requests received by the State Department appears to be stabilizing after the huge spike in FY 2016 due to the inquires regarding then-Secretary Clinton’s e-mail. In that year, State received 27,961 requests, but in FY 2019, it only received 8,589, a slight increase over FY 2018.

Lastly, I was curious about the Department of Commerce and U.S Trade Representative, the two key agencies at the heart of President Trump’s trade wars. Commerce hit a three-year high for requests received and pending requests at the end of the fiscal year, suggesting sustained interest, although unclear if it’s entirely trade war related considering the size and scope of the agency. Requests to USTR dropped from 186 in FY 2018 to 136 in FY 2019.

AB: Perhaps Interior’s backlog was exacerbated by the burdens of litigation. The data indicate that Interior incurred $2.5 million in litigation costs in FY 2019, whereas its FY 2018 costs were only $1 million. On this score, I should add that DHS’s litigation costs notably decreased over the same time period (from $7.4 million to $3.8 million), as did the litigation costs of DOD (from $5.2 million to $3.2.million). Meanwhile, DOJ, which bears the government’s largest litigation costs, saw a modest 14 percent increase (from $14.5 million to $16.6 million).

FOIA commentary: National Security Archive takes a swing at OIP

FOIA Commentary (2017-2024)Allan BlutsteinComment

When DOJ announced last week that Bobak Talebian had been appointed Director of the Office of Information Policy, succeeding Melanie Pustay after a dozen years at the helm, the National Security Archive took the opportunity to criticize Ms. Pustay’s leadership:

Under Pustay, OIP regularly published misleading FOIA statistics and provided an incomplete view of FOIA processing to both Congress and the public. In a prime example, the office’s annual summary of agencies’ FOIA reports regularly touted FOIA release rates of well over 90 percent across the government, a laughable claim. OIP arrived at this figure by excluding FOIA requests agencies denied by: overcharging fees (pricing requesters out); referrals (passing the request off to another agency while the requester still waits);  issuing a “no records” response (very frequently the result of inadequate searches); and requests deemed “improper for other reasons” (which ostensibly includes the increasingly-common “can neither confirm nor deny” Glomar exemption). A more accurate release rate calculated by the National Security Archive and others hovers between 50 and 60 percent.

See Lauren Harper, Border Agency Gets OK to Hide Previously Public Info from FOIA, and Much More, Feb. 6, 2020.

NSA’s criticism misses the mark. OIP’s annual summaries spell out exactly how agencies have processed their FOIA requests. In a section entitled “Disposition of Requests,” OIP reports the percentage of agency responses that fall within twelve categories: full grant; partial grant; full denial via exemption; no records; records referred; request withdrawn; fee-related closures; not reasonable described; improper request; not an agency record; duplicate request; and others. For fiscal years 2011 through 2018, for example, OIP indicated that agencies had released records, either in full or in part, at rates between 58.7 and 65.8 percent of all requests submitted—a range that is only slightly higher than NSA’s.

It is true that OIP reports a much higher “Release Rate,” which OIP calculates by comparing the number of requests denied in full pursuant to an exemption with the number of requests in which records were released in full or in part. The value of this statistic is debatable, to be sure. But OIP is upfront about what that rate represents; it is as illustrated by a pie chart that takes up nearly an entire page—and there is no fine print.

Commentary: Top 6 FOIA cases of 2019

FOIA Commentary (2017-2024)Allan BlutsteinComment

Federal courts issue hundreds of decisions in FOIA cases every year.  Most do not break new legal ground or attract significant media attention.  As 2019 draws to a close, the legal eagles of FOIA Advisor -- Allan Blutstein (AB) and Ryan Mulvey (RM) -- identify the decisions that stood out to them (in mostly chronological order).

(1) Rogers v. EOUSA (D.D.C. Apr. 9, 2019) -- ruling that plaintiff’s civil forfeiture settlement agreement with IRS barred FOIA plaintiff’s request to Executive Office for U.S. Attorneys for records related to his case; declining to extend D.C. Circuit’s 2017 holding regarding the validity of FOIA waivers in criminal plea agreements.

AB comment: Those who remember our commentary on the D.C. Circuit’s 2017 decision will not be surprised to read that I was pleased as punch by this year’s district court ruling. Sanity prevails! Hopefully another Circuit will split from D.C.and set up a SCOTUS case in my lifetime.

RM comment: One of my concerns coming out of DOJ v. Price was that the D.C. Circuit’s standard, and its analysis of relevant public policy concerns, would be used to foreclose FOIA waivers ever being included in a plea agreement. Call it “sanity,” if you like, but Rogers at least indicates that the near-unanimous approach of other jurisdictions in permitting waivers will continue to be followed, at least to some extent, in the D.D.C.

(2) Food Marketing Institute v. Argus Leader Media (S.Ct. June 24, 2019) -- reversing the Eighth Circuit’s decision and holding that commercial and financial information is “confidential” under Exemption 4 when it is customarily and actually treated as private by its owner.

AB comment: Any FOIA-related SCOTUS decision will qualify as a “top” decision, and this one had a monumental impact on the landscape of Exemption 4. Stay tuned as to whether Congress reverses it in 2020.

RM comment: I don’t have much to add, but I agree that this decision had a “monumental impact.” I also think the Court got it wrong! There is indeed active bi-partisan legislation to restore the National Parks competitive harm standard, and even expand it slightly by eliminating the distinction between compelled and voluntary disclosures. (This legislation also would codify the D.C. Circuit’s decision in AILA v. EOIR, 830 F.3d 667 (D.C. Cir. 2016))

(3) Animal Legal Def. Fund v. USDA (9th Cir. Aug. 29, 2019) -- in 2-1 decision, holding that statute authorized courts to order agencies to comply with “reading room” provisions -- in this instance, the online posting of “frequently requested” records generated by the Animal Plant Health Inspection Service.

AB comment: FOIA plaintiffs had been losing these cases forever until this breakthrough.

RM comment: Agreed. The Ninth Circuit has been good to the FOIA requester community lately!

(4) Rojas v. FAA (9th Cir. Oct. 22, 2019) -- reversing district court’s opinion and finding that:“consultant corollary” principle relied upon by FAA was “contrary to Exemption 5’s text and FOIA’s policy of broad disclosure, and its legal foundation . . . is tenuous at best.”

AB comment: This matter seems ripe for SCOTUS review now that the 9th Circuit has officially split with seven other Circuits.

RM comment: The Sixth Circuit has rejected the “consultant corollary,” too, in Lucaj v. FBI, 852 F.3d 541 (6th Cir. 2017). The two circuits, however, adopted different views of the implications of DOI v. Klamath Water Users Protective Ass’n, 532 U.S. 1 (2001), and whether the Supreme Court’s analysis of Exemption 5’s “threshold” requirement actually forecloses use of the doctrine. Personally, I think the meaning of “inter-agency or intra-agency” is ripe for revisiting. In light of Argus Leader and Milner, a strict text-based reading of the FOIA would seem likely to prevail. At the same time, abandoning the broad interpretation fashioned (mostly) by the D.C. Circuit would have tremendous consequences not just for “consultants,” but also for records exchanged with Congress, the President, and non-FOIA-subject agencies. So perhaps a very limited opinion is likely, if the issue were ever taken up by the Court.

(5)(a) Nat. Res. Def. Council v. EPA (S.D.N.Y. July 25, 2019) -- holding that statute’s “foreseeable harm” standard “does impose an independent and meaningful burden on agencies” in case involving the deliberative process privilege; and

(5)(b) Ctr. for Investigating Reporting v. DOL (N.D. Cal. Dec. 10, 2019 ) -- declaring that statute’s foreseeable harm standard applied to all exemptions, including Exemption 4.

AB comment:: A pair of pro-requester decisions breathe life into an ambiguous statutory provision that should have died before birth.

RM comment: You and I do not see eye-to-eye on the “foreseeable harm” standard, Allan! I’m pleased to see courts start to flesh out the meaning of “foreseeable harm” and the increased burden that the FOIA Improvement Act of 2016 imposes on agencies. NRDC v. EPA builds on a string of pro-requester decisions in the S.D.N.Y. and the D.D.C. this year, including Judicial Watch, Inc. v. DOJ (D.D.C. Sept. 24, 2019) and Judicial Watch, Inc. v. DOC (D.D.C. Mar. 22, 2019). The D.C. Circuit is expected to hear oral argument next month in Machado Amadis v. DOJ—so that is something to look forward to on the “foreseeable harm” front.

With respect to the decision coming out of California, in American Small Business League v. DOD, another judge in the same district rejected the idea that “foreseeable harm” disrupted the Exemption 4 standard adopted in Argus Leader. Maybe something will percolate up to the Ninth Circuit. Some agencies have argued that Exemption 4 is exempt from “foreseeable harm” insofar as it prohibits disclosure “by law.” At the least, we need clarity on what “interests” are actually underlying Exemption 4. Post-Argus Leader, is it confidentiality as such?

(6) Smith v. ICE (D. Colo.) -- permanently enjoining ICE from applying its standard operating procedure for denying FOIA requests of fugitive aliens, rejecting agency’s argument that such withholdings are justified by Exemption 7(A).

RM comment: Notably, this wasn’t a “policy or practice” claim, and the court adopted a reading of the FOIA’s relief provision that is squarely at odds with the D.C. Circuit. There have been some interesting developments this year with the more theoretical aspects of the FOIA, including subject-matter jurisdiction (e.g., Cause of Action Inst. v. IRS). This is one such development. If the Tenth Circuit were to adopt the same reading, then we would see a potential SCOTUS opportunity!

AB comment: The procedural issue here is notable, but it’s a shame that the agency (and, in turn, the court) retreated from an interesting issue raised by plaintiff -- that is, whether the fugitive disenfranchisement doctrine, which the D.C. Circuit adopted in 1981, is still viable after the D.C. Circuit’s 2016 decision in AILA v. EOIR (ruling that once responsive records are identified, the agency can withhold them only if an exemption applies).

Honorable mentions:

Story of Stuff Project v. U.S. Forest Serv. (D.D.C.) -- determining that records pertaining to the company’s “boreholes” constituted “wells” for purposes of Exemption 9

AB comment: Exemption 9 decisions are rare enough to warrant a nod.

Nightingale v. USCIS (N.D. Cal.) -- granting class certification to plaintiffs seeking access to immigration files, specifically Alien Registration files, citing government’s “systematic failure of making timely determinations” on their FOIA requests.

RM comment: I’ve never heard of a class action FOIA lawsuit. If this sort of approach is approved by other courts, there could be a huge shift in how “policy or practice” lawsuits are used.

AB comment: The class action approach has been attempted before, but this might be the first successful attempt?

EPIC v. Nat'l Sec. Comm'n (D.D.C.) -- holding that the National Security Commission on Artificial Intelligence is an “agency” subject to FOIA, and denying government’s motion to dismiss plaintiff’s expedition claim against DOD.

RM comment: It isn’t everyday that we add an agency to the list of government entities subject to the FOIA, especially when the “agency” is designed to close within a year’s time. Here, most of the members on the Commission on Artificial Intelligence are appointed by Congress; only three of the fifteen are appointed by the Executive Branch. This decision provides interesting discussion of the scope of Section 552(f)(1), including legislative history and the two types of “tests” that have developed in the D.C. Circuit for handling the “agency” question.

AB comment: These “agency” questions irritate me because Congress easily could have included a provision about the accessibility of the Commission’s records. And what are the chances that Congress funds the Commission to process all of its “agency” records? In the spirt of the holidays, I’ll answer that question in Yiddish: bubkis!

FOIA Commentary: Fiscal Year 2020

FOIA Commentary (2017-2024)Allan BlutsteinComment

As fiscal year 2020 gets underway, the staff of FOIA Advisor -- Allan Blutstein (AB), Kevin Schmidt (KS), and Ryan Mulvey (RM) -- discuss their hopes and expectations for the rest of the FOIA campaign.

AB: Among my hopes for FY2020 (and beyond) are that federal courts keep in check the ill-conceived “foreseeable harm” standard by applying it only to deliberative process claims. I’d also like to see the demise of Elizabeth Warren’s proposed FOIA fixes, particularly the “public interest” test that we discussed in our last roundtable. One of the bigger question marks going forward is who will replace Melanie Pustay as director of DOJ/OIP. I’ll hold off on speculating about candidates, but I’d expect an announcement before the end of the calendar year. For fun, I’ll predict that FOIA request totals for FY2019 and FY2020 will approach 950k and 1 million, respectively.

RM: I’m with you, Allan. My biggest hope is that we’ll see some more development on the foreseeable harm standard. The D.C. Circuit is considering an appeal now, in Machado Amadis v. Dep’t of Justice, that may provide a bit of clarity. Some amici have asked the Circuit to provide more general direction on how the standard is supposed to work. If the court accepts the invitation we could see some helpful and persuasive dicta. So far, I’ve seen different agencies taking different approaches to applying the standard for each of the discretionary exemptions. The case law has only involved Exemption 5 and the deliberative process privilege, and OIP refuses to issue guidance. We’re a long way from clarity.

There are two other areas where I’m hoping to see some development. First, application of Exemption 4 in the post-FMI era. OIP issued intriguing guidance, which introduced a novel three-step analysis that wasn’t devised, in my view, by SCOTUS. The import of express or implied indications by the government of possible public disclosure could lead to some interesting case law. And that goes for “step two” and the inquiry into express or implied assurances of confidentiality at the time of submission, too. It’s an interesting time for Exemption 4.

Second, I’m eager to see what sort of FOIA reform may end up coming out of Congress. The Open and Responsive Government Act of 2019 would reinstate the National Parks standard, and even expand its application, as well as codify the D.C. Circuit’s decision in American Immigration Lawyers Association v. Executive Office for Immigration Review. I would welcome both of these developments, so I’m rooting for the bill to succeed. And I’d like to see some other FOIA reform coming into the early stages of development.

KS: I try to stay away from hopes regarding FOIA, but if I”m forced to go on the record, I’ll go with: Meaningful enforcement and/or guidance for the foreseeable harm standard, even if it’s only involving Exemption 5. Ideally, the FOIA reform bills under consideration would include further limiting of Exemption 5 as well. I’m also hoping the new director of DOJ/OIP is committed to improving FOIA more broadly, or at a minimum, will do more to encourage agencies to comply with the law. As you might expect, I’m keeping my expectations for these hopes quite low.

AB: Chin up, Kevin. But how much more money would you ask taxpayers to pay to improve FOIA operations, because a cash infusion is what it’s really going to take (not statutory amendments or finger-wagging from DOJ). Requesters like us certainly don’t help the cause much, contributing less than a fraction of one percent of actual costs. So enjoy the bargain. Plenty of government information specialist jobs are available, by the way, if you want to try to improve the system from the inside.

We may agree about wanting clarification from the courts or Congress on various issues, but it’s apparent we disagree about how we’d like to see those issues resolved. In my view, for example, Congress should repeal the foreseeable harm standard, as well as the D.C. Circuit’s ruling in AILA v. EOIR — that is, agencies should be allowed to “scope out” information as “non-responsive” even on less than a page-by-page basis. I can live with a resurrection of National Parks. No one bit on the number of FOIA requests, so do I win by default?

KS: I’m not necessarily opposed to more money for FOIA operations, but there would need to be evidence that they are trying to do more with what they have to start. For example, “scoping out” information on less than a page by page basis is a waste of time without a strong reason to support it. Frankly, a lot of time and money can be saved by releasing more records, rather than line by line redaction. I’ve seen a lot of ridiculous FOIA requests that leave me sympathetic to agencies and information specialists, but I’ve been given the runaround too many times to believe we are getting full value for the money.

My estimate for request totals: FY2019: 900,000 and FY2020: 920,000 (I think election year may slow it down a bit).

AB: Spoken like a true libertarian!

FOIA Commentary: Is a public interest test in the public's interest?

FOIA Commentary (2017-2024)Allan BlutsteinComment

In case you missed it, Columbia University Press published a collection of 14 articles last year that examine how freedom of information policies are working, failing, and might be improved. In one of those articles, “Striking the Right Balance: Weighing the Public Interest in Access to Agency Records Under the Freedom of Information Act,” authors Katie Townsend and Adam Marshall argue that all discretionary exemptions -- that is, exemptions other than Exemptions 1, 3, and 4 -- should include a public interest balancing test. That proposal was extended even further by the Anti-Corruption and Public Integrity Act introduced in the U.S. Senate on August 21, 2018 (and now reintroduced in the current Senate by Sen. Elizabeth Warren) and in the U.S. House on November 16, 2018, which would apply a five-factor public interest test to all of the exemptions.

The staff of FOIA Advisor, Allan Blutstein (AB), Kevin Schmidt (KS), and Ryan Mulvey (RM), weigh in on the proposed public interest test.

AB: I have two main objections to these proposals. First, balancing tests often undermine certainty and predictability of the law by tempting decision makers to inject their subjective opinions. That temptation is unavoidable under the Anti-Corruption and Public Integrity Act, which expressly requires agencies and courts to consider “any other factors that the agency or court determines necessary.”

My second objection is a more practical one. It is difficult enough for agencies to accurately determine whether requested records fall within a FOIA exemption -- a largely unappreciated burden that agencies undertook in response to 581,416 requests in fiscal year 2018. The disclosure analysis may soon become more challenging depending on how courts interpret the ambiguous “foreseeable harm” language enacted in 2016. Adding a complicated public interest test to the mix is likely to overburden agencies and courts alike, with Mr. & Mrs. Taxpayer bearing the brunt of increased litigation costs.

KS: I think the practicality issue is a real challenge, so my counter proposal is simply to remove the deliberative process exemption under Exemption 5. Outside of national security issues, that’s going to be the Exemption used for the majority of the withholding that would be potentially challenged under the balancing test. The rationale is simple: taxpayers pay for the the work and are subject to the policies crafted in federal agencies and therefore have an interest in understanding how those policies were made.

I’m also a non-lawyer, so I’d also get rid of attorney-client privilege exemption, but I don’t see that one happening and it surely rankles my colleagues here and lawyers everywhere.

RM: I appreciate the idea of a public interesting balancing test, but like my colleagues I share their objections for many of the same reasons. As an initial matter, the sort of five-factor test that has been proposed in Congress would build on, and not replace, the FOIA Improvement Act of 2016’s “foreseeable harm” standard. That standard was intended by Congress to avoid the technical application of FOIA exemptions, especially when the government seeks to hide embarrassing or politically damaging information. Complicating the FOIA by adding yet another balancing analysis doesn’t make sense to me. We still haven’t figured out how the “foreseeable harm” standard is supposed to work; the courts are still deciding.

And this suggests a deeper criticism. Balancing tests rely too much on the judiciary to ensure that the law is being properly applied. Marshall and Townsend lament the development of FOIA jurisprudence over the past fifty years. But the courts are responsible for that development. Courts have chosen to treat de novo review under FOIA as something special (and seemingly less than de novo!). Courts have chosen to defer to agency declarants. Courts refuse to grant discovery outside of extraordinary circumstances. I have little faith that forcing courts to review an agency’s application of yet another variable analysis is going to advance the cause of open government.

The better approach, I think, would be for Congress to focus on amending the FOIA to include clear statutory rules. Already, a bipartisan group of Senators has proposed redefining the term “confidential” for the purposes of Exemption 4, and they also aim to codify the D.C. Circuit’s ruling in American Immigration Lawyers Association, with the Open and Responsive Government Act of 2019. These are helpful, simple, textual reforms that leave little room for courts to play with the law.

What are some other possible “fixes”? Here are some thoughts: (1) provide more detailed direction for agencies to apply the “foreseeable harm” standard; (2) shorten the “sunset provision” that limits the use of the deliberative process privilege; (3) strengthen the role of the Office of Government Information Services, so that it can provide a real check on the agency appeals process; and (4) reverse the Supreme Court’s elimination of the absolute-qualified privilege distinction in FOIA by re-working Exemption 5. I suspect that this last change could provide the same sort of public interest “safety valve” that Marshall and Townsend want, but without introducing a radical new multi-factor test. Agencies would instead be forced to invoke privileges as any other party in civil litigation.

AB: Since we agree about nixing a public interest test, I’ll move on to some of your own proposals. Transparency advocates have lobbied for decades to eliminate the deliberative process, so you are in good company, Kevin. I am not prepared to go that far yet, but I could support Ryan’s suggestion for a shorter sunset provision -- say five years (borrowing from the Presidential Records Act). OGIS would need a significant boost in funding to handle something other than voluntary and non-binding dispute resolution on the appeals side; the government received 15,559 in fiscal year 2018. How would you feel about a mandatory filing fee to partially subsidize that effort? The Supreme Court’s recent interpretation of the term “confidential” under Exemption 4 could not have simplified the statute more clearly, but I do not object to resuscitating the Byzantine “substantial competitive harm” standard. I’ll pass on permitting a needs-based test to Exemption 5, however, which would be a backdoor public interest test.

KS: I’m not sure how much guidance would help for applying foreseeable harm, but I’m not opposed to trying to see if it helps. I would support a sunset provision, but don’t think it goes far enough. There is far less utility in finding out what the government was up to five years ago as opposed to the same year or two. The empowered OGIS proposals are always fun to tinker with. Can you imagine how DOJ OIP would react if OGIS was given the authority to overturn agency withholdings? I’m in favor of rolling out the proposal for that reason alone. I’d be fine with a flat fee structure, say $10 to $25 per case, to help support it.

RM: It seems to me that five years as a sunset provision for deliberative process is about right. I would even give the government eight years—or the span of a two-term presidential administration. I recognize, Kevin, that you’re all for eliminating deliberative process all together. I’m more on Allan’s side in recognizing that it serves a purpose. At the same time, I’ll be the first to admit how agencies overuse the privilege. I am also skeptical that disclosure of truly deliberative material in many cases is ever likely to harm the underlying interests, say, of insulating agency decision-making from undue interference or avoiding public confusion. I’d be OK with Allan’s suggestion of an additional filing fee for a strengthened OGIS process. As long as it is cheaper than the standard filing fee in the D.D.C.! Congress would need to send more money OGIS’s way, too, but I think it would be worthwhile.