After a difficult 17-month tenure as President Trump’s third National Security Advisor, John Bolton was ready to talk about his experiences, including presidential conduct he found “deeply disturbing.” But he was not eager to talk with any of the nation’s elected representatives who were participating in impeachment proceedings — proceedings to determine whether the President’s conduct toward Ukraine constituted an impeachable offense warranting removal.  Rather, Bolten was eager to recount his experiences and observations in detail at $32.50 a copy, in a book prosaically entitled The Room Where It Happened: A White House Memoir (2020). 

Alas, the national security prepublication review process stood in the way of the optimal timing for the book’s release.  The process began with submission of the manuscript to Ellen Knight, Senior Director for Records Access and Information Security Management, National Security Council (“NSC”).  After hours of negotiation between Bolton and Knight, spanning several  months, regarding modifications to the book sufficient to satisfy Knight’s concerns, on April 27, Knight orally advised Bolton that she had “completed her review and was of the judgment  that the manuscript draft did not contain classified information.” Amended Complaint ¶¶31-46.

But the prepublication review process was not yet over — far from it.  On May 2, 2020, at the request of the Assistant to the President for National Security Affairs, Michael Ellis, the NSC’s Senior Director for Intelligence, commenced a supplemental review of Bolton’s manuscript. Id, at ¶51.  Bolton however, after receiving Knight’s oral representations, but no written confirmation, approved Simon & Schuster’s release of the book.  (Bolton had followed up with Knight for 10 days, through May 7, during which time she advised him the review had not been completed, but provided no additional details.) The book was published in early June, while Ellis was still conducting his review.[1]  The White House first sought to enjoin publication of the book, and then sought to capture all the proceeds from the book due to Bolton’s breach of his prepublication agreement.[2]

District Judge Royce C. Lamberth denied Bolton’s motion to dismiss the complaint.  U.S. v. Bolton, 2020 WL 5866623 (D.D.C. Oct. 1, 2020).  The opinion serves as a strong reaffirmation of the prepublication review process and the requirement that employees adhere to its requirements.

I. Judge Lamberth’s Decision

A. The Government’s Claims

In its Amended Complaint, the Government alleged that Bolton had violated the standard employment agreements that federal employees having access to classified information must sign: the Form 312, covering classified information, and the Form 4414, covering sensitive compartmented information (“SCI”).[3]  Information can be classified when its unauthorized disclosure could reasonably be expected to cause identifiable or describable damage to the national security in at least one of eight respects.^[4]  SCI, is a subset of classified information, consisting of “[c]lassified information concerning or derived from intelligence sources, methods, or analytical processes, which is required to be handled within formal access control systems established by the Director of National Intelligence.”[5]

In considering Bolton’s motion to dismiss, Judge Lamberth first discussed Bolton’s failure to complete prepublication review. Then he considered whether Bolton had violated those agreements by sharing classified information with unauthorized recipients, namely the editorial staff at his publisher, Simon & Schuster.^[6]

B. Prepublication Review Obligations

1. Form 4414 Obligations

Judge Lamberth began by outlining five critical aspects of the Form 4414 agreement. First, the signatory agrees to “never divulge” any marked or known SCI without written authorization. Second, to “give the United States a reasonable opportunity to determine whether [a contemplated disclosure includes] any SCI,” the signatory agrees to “submit for prepublication” security review any writings that meet certain triggering conditions, Form 4414, ¶¶ 4, 5. Third, the signatory agrees to “not disclose the contents of [a writing submitted for review] with,” or show it to, anyone “not authorized to have access to SCI” before receiving written permission to do so. Id., ¶ 4. Fourth, the agreement specifies that SCI remains property of the United States. Id., ¶ 8. Lastly, the signatory assigns to the Government “all rights, title and interest, and all royalties, remunerations, and emoluments that have resulted, will result, or may result” from any unauthorized disclosure. Id., ¶ 12.

Focusing on the conditions that trigger the prepublication review obligation, the District Court noted that such an obligation arose only when the current or former employee produces a writing that both (a) meets one of two disclosure thresholds, and (b) satisfies one of three specified nexuses to SCI.  As a threshold matter, the current or former employee must plan to disclose to either (1) one or more unauthorized individuals, or (2) the general public. Id, ¶ 4.  If so, the prepublication obligation exists if the contemplated writing (a) actually contains any SCI or related matter, (b) purports to contain such matter, or (c) contains materials its author has “reason to believe” is derived from SCI. ^[7]

As an initial matter the Court found the Government’s allegations sufficient to plead a claim that Bolton’s book met at least one of the threshold and nexus requirements and that Bolton had not satisfied the obligation of awaiting completion of the prepublication review. 

The Court had earlier laid out the relevant principles of contract interpretation.  It explained that federal common law governs the interpretation of contracts between private parties and the United States.[8] But, it observed, the federal common law of contracts “dovetails … with general principles of contract law,”[9] including those set out in the Restatement (Second) of the Law of Contracts.[10]  Under such principles, “[c]ontract interpretation begins with the plain meaning of the text,”[11] and, when the parties’ intent “can be determined from the face of the agreement,” ends there as well.[12]

After concluding that Bolton’s obligation seemed clear under the SCI agreement’s text, the Court canvassed Bolton’s argument that he had no obligation to comply with the prepublication review requirement.  The upshot of Bolton’s First Amendment and contract interpretation arguments was that the prepublication requirements were inapplicable unless he had some reason to believe his book’s disclosures contained SCI or SCI-related information.  I discuss the First Amendment arguments first, even though that was not the Court’s starting point.

a. Bolton’s First Amendment Arguments

Bolton argued that the First Amendment relieved him of the prepublication review obligation if: (1) he “neither believed nor had reason to believe” that his writings contained material related to SCI, and (2) his disclosures were harmless, by which he meant he had no reason to know his disclosures related to SCI.^[13] 

The Court relied heavily on Snepp v. United States, 444 U.S. 507 (1980), in rejecting each argument. With respect to the first, it explained that in Snepp the Supreme Court had rejected a First Amendment challenge to the CIA’s stricter prepublication review requirements. The CIA’s nondisclosure agreement prohibited CIA agents from publishing “any information or material relating to the Agency, its activities or intelligence activities generally, either during or after the term of [his] employment . . . without specific prior approval by the Agency.”  Given that “the government could constitutionally require Snepp to submit for review any writings related to intelligence activities, . . . the government [could constitutionally] require Bolton to submit for review writings related to the narrower category of intelligence sources and methods.”[14]

Nor did the First Amendment entitle a current or former employee to avoid the prepublication review system simply because that employee concluded his or her disclosures were harmless.  That argument, too, contravened Snepp.  The Court had upheld a blanket prepublication review requirement for any writings related to a signatory’s CIA employment, even if it contained no classified information whatsoever.  In any event, Judge Lamberth explained, the Government has “a compelling interest in reviewing information related to SCI prior to its disclosure.”[15] He observed: “What may seem trivial to the uninformed, may appear of great moment to one who has a broad view of the scene and may put the questioned item of information in its proper context.”[16]

b. Bolton’s Contract Interpretation Arguments

Bolton asserted that, properly interpreted, the SCI-contract itself required scienter, namely that the current or former employee knows or should know that he is disclosing SCI.  In support of his argument he relied on both Form 4414’s text and its structure.

His textual argument was based on inclusion in the agreements of language referring to “contemplate[d]” disclosures of SCI. Bolton argued that such word choice implied that submitting a disclosure to prepublication review was necessary only when the current or former employee knows or should know that he is disclosing SCI.”^[17]  The Court rejected the argument — finding Bolton’s reading flawed and contradicted by other aspects of the Agreement’s text.  In particular, the court noted that the third potential nexus expressly covers material that a recipient “ha[s] reason to believe” is derived from SCI, but that the other two nexuses do not.  Rather, the other two nexuses, by their terms, define strict liability breaches. Moreover, Judge Lamberth explained, liability for breach of contract rests on principles of strict liability.[18]

For his structural argument, Bolton invoked the syntax canons regarding avoidance of absurdity and avoidance of surplusage when construing legal texts.  Bolton argued that construing the agreements to require prepublication review when a current or former employee has no reason to know that his writing contains SCI, would produce absurd results. Such a construction would render three provisions of the contract superfluous.  First, paragraph 3’s requirement that current or former employees consult with the relevant agency to ascertain the agency’s view would become meaningless.  Second, the provision precluding sigantories from disclosing SCI-related material would serve no purpose.  And third, the “reason to believe” nexus would become irrelevant.[19]

Addressing the last argument first, the Court explained that, taken together, the SCI nexus requirements were broader than a simple presence-of-SCI trigger. For example, the three-pronged nexus requirement expressly covers instances in which the current or former employee purports to be disclosing SCI. Bolton also failed to recognize that the nondisclosure and consultation requirements provide supplemental protections for SCI.  And in the national security context, the Government may understandably have drafted broadly out of a sense of caution, making surplusage arguments less persuasive.[20]

Bolton also invoked the canon that contracts, particularly contracts of adhesion, should be interpreted against the drafter, in this case the Government.  The Court acknowledged that the Form 4414 was a contract of adhesion, but did not find sufficient ambiguity in the contract to make the contra proferentem doctrine applicable.[21]

In short, The SCI agreements required Bolton to complete prepublication review.[22]

2. Form 312 Obligations

Turning to the Form 312 classification agreement, the Court noted a basic difference between the SCI agreement and the classification agreement.  Unlike the SCI agreement, the classification agreement did not require Bolton to submit his manuscript to prepublication review.  However, it did require that Bolton “ensure that anyone he shares that information with is authorized to receive it.”  A signatory who is uncertain whether information is classified must “verify its status prior to disclosure.”  The Judge observed that “[c]ertainty, after all, is a high bar to clear—an iota of doubt would render Bolton uncertain.”  The Government’s allegation that Bolton submitted his manuscript for review and that the manuscript actually contained classified information, was sufficient to survive a motion to dismiss.  Similarly, the Government’s allegation that Bolton disclosed the manuscript to his publisher both during and after Director Knight’s initial prepublication review was sufficient to withstand a motion to dismiss. ^[23]

C. Non-Disclosure Obligations

The Court then discussed whether Bolton had violated the Form 4414 and Form 312 non-disclosure provisions.  It concluded that the Government had not adequately pled a violation of the SCI agreement’s non-disclosure provision.  The complaint failed to allege that Bolton knowingly disclosed SCI or that he disclosed material marked as SCI.  By alleging merely that Bolton had “disclosed classified information without prior written authorization,” the Government’s pleading was deficient.^[24]

The Government’s claim that Bolton had violated the Form 312 non-disclosure provisions fared better.  The allegation that Bolton disclosed classified information without prior written authorization was sufficient to state a prima facie claim for breach of the classified information agreement’s nondisclosure requirement.[25]

The Court refused to read a scienter requirement into the agreement’s non-disclosure clause, just as it had refused to read such a requirement into the SCI agreement’s prepublication review provisions.  Indeed, the Court noted, unlike the SCI agreement, which mentioned the discloser’s state of mind in one of the three nexus criteria, the classified information agreement nowhere mentioned any state of mind requirement at all.  As before, the court presumed that violation of contractual obligations subjected the violator to strict liability for breach.^[26]

Bolton argued that requiring him to consult with his agency in cases of uncertainty would be pointless were he strictly liable for disclosing classified information.  The Court rejected the argument, finding the arguably redundant combination of pre-disclosure consultation and post-disclosure liability sensible in the context of protecting classified information. “If money damages—or even a constructive trust—would adequately remedy unauthorized disclosure, the government would be indifferent between compliance and damages.”  But money damages do not provide an adequate remedy for such security breaches — given that “[o]nce information is released, it cannot be resecured.” Damages were a distinctly second-best solution.^[27]

In a critical paragraph, the Judge Lamberth compared the non-disclosure and predisclosure provisions of the SCI agreement with those of the confidential information agreement to reveal the agreements’ two disparate frameworks for protecting classified information: one relying more heavily on ex ante measures and the other relying more heavily on ex post measures.

The SCI agreements impose a stricter predisclosure regime: a full security review. But they also impose a less stringent non-disclosure obligation: a person bound by the SCI agreements may not share marked or known SCI. In short, they heavily burden a recipient’s ability to disclose information prior to prepublication review but give the government less recourse to sue a recipient for disclosing SCI after review. In contrast, the classified information agreement imposes a laxer predisclosure regime: mere predisclosure authorization. But it also prohibits any disclosure of classified information. In short, it imposes minimal upfront burdens but strictly prohibits disclosure of classified information and provides the government broad ability to sue for unauthorized disclosures.^[28]

D. Constitutionality of Use of Constructive Trust As A Remedy

Finally, Bolton had argued that imposing a constructive trust on his royalties would violate the First Amendment’s prohibition on punishing a speaker for the content of his speech.  Acknowledging that the imposition of a public trust might feel like punishment to Bolton, the Court explained that such relief was merely an equitable remedy that served to prevent unjust enrichment.  In particular, a constructive trust is imposed when the trust res never properly belonged to the defendant, to prevent the defendant’s unjust enrichment at the plaintiff’s expense. And preventing such unjust enrichment was just what the Government sought — in the Government’s view the royalties Bolton will earn from the memoir are rightly the government’s property, being the proceeds of misappropriated classified information.^[29]

II. Observations

Given the norm-defying, hyper-partisan Trump Administration and the presidential election cycle, it seems quite likely that neither the prepublication review of Ambassador Bolton’s book nor the decision to take legal action against him were entirely apolitical and dispassionate national security decisions.  It would hardly be surprising to learn that some of the aspects of the book that came under the greatest scrutiny were the ones most unflattering to the President.  Yet one can understand Judge Lamberth’s impatience with Ambassador Bolton’s reading of the Free Speech Clause and the prepublication agreements in a way that would enable Bolton to escape the review process simply because Bolton believed his disclosures harmless.  In any event, Judge Lamberth seems to have resolved the issues presented by the case in a reasonable manner.  Thus, my brief observations will range more broadly.

At base, the prepublication review requirements establish a licensing system for certain speech in which current and former government employees wish to engage; it is thus the classic form of prior restraints disfavored under Free Speech Clause doctrine.  Indeed, William Blackstone focused on the prohibition on prior restraint as the primary, indeed exclusive, protection of freedom of the press and freedom of speech.[30]

Prior restraint systems have five characteristics that make them particularly problematic.  First, a pre-publication review system makes it easy to restrict speech—the censor need merely refuse consent.  Second, censors have an inclination to refuse consent, particularly if embarrassing speech is at issue.  Third, censors have an incentive to suppress at least some speech to justify their own existence.  Fourth, censors often apply standards arbitrarily and inconsistently.  Fifth, pre-publication review regimes involve delay, and put pressure on prospective speakers to modify their speech to satisfy censors in an effort to obtain timely approval.  See generally, Freedman v. Maryland, 380 U.S. 51 (1965); John Calvin Jeffries, Rethinking Prior Restraint, 92 YALE L.J. 409, 421-22 (1983); Michael L. Charlson, The Constitutionality of Expanding Prepublication Review of Government Employee Speech, 72 CAL. L. REV. 962, 981-82 (1984).

The national security prepublication review system established by Form 4414 agreements appears to suffer from several of these maladies.  See, Christopher E. Bailey, Reform of the Intelligence Community Prepublication Review Process: Balancing First Amendment Rights and National Security Interests, 5 NAT’L SECURITY L.J. 203, 205, 207-08, 209-210 (2017) (discussing the inconsistent management practices, delays, and allegedly politically-inspired censorship of the prepublication review process); accord, Jack Goldsmith & Oona A. Hathaway, The Government Prepublication Review Process is Broken, Wash. Post (Dec. 25, 2015);  H.R. REP. NO. 114-573 7 (May 18, 2016)(“The Committee is . . .aware of the perception that the pre-publication review process can be unfair, untimely, and unduly onerous.”)  Critics of the prepublication review process highlight one of the concerns Bolton raised in his defense, the vagueness of the standards for having to submit to the review process.  Goldsmith & Hathaway, supra; H.R. REP. NO. 114-573, supra, at 7-8.

Freedman v. Maryland established criteria for censorship regimes to address the five pathologies enumerated above.  For such an administrative prepublication review scheme to pass constitutional muster, it must meet three procedural requirements: (a) the burden of initiating civil proceedings and proving the unprotected nature of the speech must rest on the censor, (b) any prior restraint must be for a brief time and only insofar as necessary to preserve the status quo, and (c) prompt final judicial determination must be assured.  380 U.S. at 59. These requirements were largely reaffirmed in FW/PBS v. City of Dallas, 493 U.S. 215 (1990).  Freedman and FW/PBS v. City of Dallas both involve the regulation of sexually‑oriented expression by non-governmental speakers.[31] 

The Government has greater power, consistent with the Free Speech Clause, to limit current and former government official’s disclosure of confidential government information to which they had access by virtue of their government employment.  Nevertheless, it is not clear why the Freedman v. Maryland’s procedural standards, standards designed to combat the maladies that typically plague prepublication administrative review regimes, should be inapplicable to the Form 4414 pre-publication review regime.[32] 

So, how does the regime fare under the Freedman v. Maryland criteria?

With regard to the first Freedman v. Maryland requirement, it is not entirely clear whether the Government has the burden of initiating proceedings to prohibit publication of a work it believes to contain SCI or SCI-related material.  But more importantly, the Government certainly does not have to shoulder the burden of showing the current or former employee’s speech is unprotected. 

As to the former, the prepublication agreement states only that the Government may seek to enjoin disclosures made in violation of the agreement.[33]  The Department of Justice, in regulations applicable within the Department, is more specific.  It suggests that if the current or former employee objects to a final determination concluding the preclearance process, the Government will proceed to seek an injunction within 30 working days.[34]  But it is not clear that the Government’s failure to do so constitutes a waiver of its rights.  And indeed, current and former employees sometimes initiate litigation by seeking judicial review of the preclearance decision pursuant to the Administrative Procedure Act (“APA”), presumably because they do not wish to wait around for the Government to file an action.[35]

In any event, the Government is unlikely to have the burden of persuasion regardless of the form the litigation takes.  Under the APA, challengers generally have the burden of persuasion, and some form of deference would ordinarily apply.  But even if the Government initiated the action, the Government’s classification decisions and the application of those classification decisions to the work in question is likely to be reviewed under a quite deferential standard.  Indeed, in FOIA litigation, where the statute explicitly requires de novo review of classification decisions, 5 U.S.C. §§552(a)(4)(A)(vii), §552(b)(1),[36] courts nevertheless accord substantial deference to the Executive Branch’s classification decisions.  See, e.g., ACLU v. U.S. Department of Defense, 901 F.3d 125, 135 (2d Cir. 2018) (deference to the executive’s national security and military judgments is appropriate when those judgments are “logical and plausible”).

As to the second requirement, the prepublication review process is characterized by significant delays, delays which make one commentator pine for the relative expedition of the FOIA process (a prospect few who have had experience with the delays endemic to FOIA would have imagined). 

There appear to be some time deadlines governing the process.  In the agreement itself, the signatory must give the agency 30 working days (approximately 38-40 calendar days) to render a decision.  Then the current or former employee must generally exhaust his or her remedies by appealing an unfavorable decision, which may consume another 15 working days.  If still dissatisfied, the current or former employee must give the Government another 30 working days to file suit to enjoin the publication.  But perhaps just as importantly, the agency may suffer no consequence for missing any of these deadlines.[37]  It is unclear that failure to adhere to the deadlines waives the Government’s rights to object to the publication or bring suit to enjoin it; it is not even clear that such a default waives the Government’s right to insist that the employee pursue the preclearance process to completion.

As to the third element, it is not clear that a prompt judicial determination of the agency’s prepublication review decision is ensured.  See Bailey, supra, at 225.[38]  If the Government seeks to enjoin publication, Federal Rule of Civil Procedure 65(b) suggests that a temporary restraining order could remain in place for 14 or 28 days.[39]  If the current or former employee must bring suit under the APA, the deadline, if any, for ruling the Government’s action improper is unclear.

Conclusion

While courts will surely continue to uphold the basic principle that the Government must have a reasonable opportunity to review potential current and former employee speech to protect classified information relating to intelligence sources and methods, it seems clear that the prepublication review process could use some reform.  See H.R. REP. NO. 114-573, supra, at 7-8 (directing the National Director of Intelligence to consider reforms regarding several aspects of the prepublication review process).[40]  Indeed, if the right plaintiff comes along, a court may impose some Freedman v. Maryland-type procedural requirements as a constitutional matter.

---

[1] This kind of disagreement between government officials is not unprecedented.  In Shaffer v. Defendse Intelligence Agency, 102 F. Supp. 3d 1, 3 (D.D.C. 2015), the former employee received prepublication approval from the agency at which he had last worked, the Army Reserve.  However, despite its obligation to do so, the Army Reserve had not conducted the necessary interagency consultation.  The Defense Intelligence Agency subsequently learned of the planned publication and objected, leading to the Army Reserve to revoke its approval.

[2] Many of the filings in the case can be found at https://knightcolumbia.org/cases/us-v-bolton

[3] Although certain national security agencies had used prepublication review requirements previously, the Reagan Administration routinized and expanded the prepublication review requirements.   See, National Security Decision Directive 84; Michael L. Charlson. The Constitutionality of Expanding Prepublication Review of Government Employee Speech, 72 CAL. L. REV. 962, 965-69 (1984).

^[4] Exec. Order No. 13,526, 75 Fed. Reg. 707, §1.4 (Dec. 29, 2009).  However, classification cannot be used “to obscure illegal activities, prevent embarrassment, or delay the release of otherwise unprotected information.” Id. at § 1.7(a).

[5] Joint Task Force, National Institute of Standards and Technology, U.S. Department of Commerce, Security and Privacy Controls for Information Systems and Organizations 418 (NIST Special Publication 800-53).  See U.S. v. Bolton, 2020 WL 5866623 *1 (D.D.C. Oct. 1, 2020).

^[6] U.S. v. Bolton, supra at *1-*3.

^[7] Bolton at *7-*9 (emphasis added ). I have used the term SCI-related matter to include “[any] description of activities that produce or relate to SCI,” as the phrase is used in paragraph 4 of the SCI agreement

[8] Id. at *4 (citing NRM Corp. v. Hercules, Inc., 758 F.2d 676, 681 (D.C. Cir. 1985)).  The Form 4414 agreement also contains a choice of law provision stating that the Agreement “shall be interpreted under and in conformance with the law of the United States,” Form 4414, ¶ 15 (choice of law provision), as Judge Lamberth noted. U.S. v. Bolton, supra, at *4.

[9] Id. (citing NRM Corp., supra, 758 F.2d at 681).

[10] Id. (citing Bowden v. United States, 106 F.3d 433, 439 (D.C. Cir. 1997)).

[11] Id. (citing Bode & Grenier, LLP v. Knight, 808 F.3d 852, 862 (D.C. Cir. 2015)).

[12] Id. (citing CITGO Asphalt Refin. Co. v. Frescati Shipping Co., — U.S. —, 140 S. Ct. 1081, 1088 (2020), which in turn quoted 11 WILLISTON ON CONTRACTS § 30:6 (4th ed. 2012)).

^[13] Id. at *7-*9.

[14] Id. at *7.

[15] Id. at *8.

[16] Id.  This seems a curious observation when directed at a recently departed National Security Advisor to the President of the United States.

^[17] Id. at *6. (Emphasis added.)

[18] Id., at *6-*7; see RESTATEMENT (SECOND) OF THE LAW OF CONTRACTS (1981) (Introductory Note to Chapter 11) (“Contract liability is strict liability …. The obligor is therefore liable in damages for breach of contract even if he is without fault ….”).

[19] Id. at *9.

[20] Id. (citing Ethan J. Leib & James J. Brudney, The Belt-and-Suspenders Canon, 105 IOWA L. REV. 735, 750 (2020)).

[21] Id. at *8.

[22] Id. at *10.

^[23] Id. at *10-*11 (emphasis added).

^[24] Id. at *11.

[25] Id. at 11.

^[26]  Id.

^[27] Id. at *12.

^[28] Id.

^[29] Id.

[30] He observed: “The liberty of the press is indeed essential to the nature of a free state: but this consists in laying no previous restraints upon publications. . . To subject the press to the restrictive power of a licenser, as was formerly done . . . is to subject all freedom of sentiment to the prejudices of one man, and make him the arbitrary and infallible judge of all controverted points in learning, religion, and government.” 4 WILLIAM BLACKSTONE, COMMENTARIES ON THE LAW OF ENGLAND 89 (Lonang Group elec. ed. 2003.)(downloadable at https://lonang.com/ ).

[31] In the former, the Court considered a motion picture censorship statute.  In the latter, The Court considered municipal ordinance establishing licensing regime for “sexually oriented businesses.”

[32] FW/PBS v. City of Dallas muddies the waters a bit by suggesting that not all prepublication administrative review regimes must meet the first requirement.  However, in this respect Justice O’Connor’s opinion was only a plurality decision, joined by only two other Justices, and Justices Brennan, Marshall, and Blackmun, the three other members of the majority, vigorously disputed Justice O’Connor’s argument regarding the inapplicability of the first element.

[33] “I understand that the United States Government may seek any remedy available to it to enforce this Agreement including, but not limited to, application for a court order prohibiting disclosure of information in breach of this Agreement.” ¶7 (emphasis added).

[34] 28 C.F.R. 17.18(i).  The provision says: “Persons who are dissatisfied with the final administrative decision may obtain judicial review either by filing an action for declaratory relief, or by giving the Department notice of their intention to proceed with publication despite the Department’s request for deletions of classified information and giving the Department 30 working days to file a civil action seeking a court order prohibiting disclosure.

[35] Baily, supra, at 235.

[36] Exemption 1 exempts matters that are: “(A) specifically authorized under criteria established by an Executive order to be kept secret in the interest of national defense or foreign policy and (B) are in fact properly classified pursuant to such Executive order.” (Emphasis added.)

[37] Under FOIA, for instance, if the Government fails to meet the statutory deadlines, it waives its right to interpose an exhaustion of administrative remedies defense should the FOIA requester initiate litigation. Citizens for Responsibility & Ethics in Washington v. Federal Election Comm’n, 711 F.3d 180, 184-85 (D.C. Cir. 2013) (Kavanaugh, J.)).

[38] Charlson likewise questioned whether the prepublication review system satisfied the Fredman v. Maryland requirements. Charlson, supra, 72 CAL. L. REV. at 986-89, 1015-1016.

[39] In the obscenity context, the Freedman v. Maryland Court noted with approval a New York statute that required commencement of a hearing within one day after the issue was joined and a decision within two days after the hearing concluded.  380 U.S. at 60.  Granted, the Supreme Court is likely to be significantly more flexible in the national security area, particularly with regard to ruling on claims involving review of lengthy books.

[40] At least as on November 2018, the Office of the Director of National Intelligence had not completed the work mandated by the statute. Letter from Sally A. Nicholson, Chief, FOIA Branch, Information Management Division, Office of the Director of National Intelligence to Steven Aftergood, Federation of American Scientists, dated November 20, 2018.