Between October and December 2003 U.S. military personnel engaged in “sadistic, blatant, and wanton criminal abuse” of Iraqi detainees at Abu Ghraib prison. See, Seymour M. Hersh, Torture at Abu Ghraid, THE NEW YORKER (May 10, 2004). Photographs and videos recorded the abuse; a few were published when the abuse was first reported. See, id. Thus began a long saga regarding release of the remaining photographic images. Last week, the Second Circuit ruled that the Department of Defense could continue to withhold many of the photographs, American Civil Liberties Union v. United States Department of Defense, — F.3d —, 2018 WL 3977021 (Aug. 22, 2018). But the litigation as a whole illustrates the dangers of “hard look” and “proceduralized” judicial review, harkening back to the heyday of that debate almost 50 years ago.
The Protected National Security Documents Act of 2009 (“the PNSDA”)
In 2008, the Second Circuit affirmed an order requiring the Government to provide previously-unreleased Abu Ghraib photos to FOIA requesters. American Civil Liberties Union v. Department of Defense, 543 F.3d 59 (2d Cir. 2008), vacated sub nom., 558 U.S. 1042 (2009). In response, Congress enacted the “Protected National Security Documents Act of 2009” as a part of the Department of Homeland Security Appropriations Act, 2010, Pub. L. No. 111-83, §565, 123 Stat. 2142 (2009). The PNSDA specified that certain documents, termed “protected documents,” “shall not be disclosed” pursuant to FOIA. Id. at §565(b). Protected documents were defined as photographs taken from September 11, 2001, through January 22, 2009 “relate[d] to the treatment of individuals engaged, captured, or detained after September 11, 2001” by U.S. forces operating abroad. Id. at §565(c)(1)(B). The Act protected such photographs only if the Secretary of Defense had certified that their disclosure “would endanger citizens of the United States, members of the United States Armed Forces, or employees of the United States Government deployed outside the United States.” Id. at §565(c)(1)(A) (emphasis added). Certifications had to be renewed every three years, id. at §565(d)(2), with notification to Congress, id. at §565(d)(4). Unlike FOIA, the PNSDA does not explicitly empower the District Judge to review the withheld materials in camera.
The 2009 and 2012 Certifications
All Abu Ghraib photos were certified in November 2009. The District Court upheld the certification. ACLU v. Department of Defense, 2018 WL 3977021 *2 (2018). But when the Secretary of Defense issued an almost identical certification in 2012, the District Court demurred. Id. at *3. The Court explained that the passage of time justified more rigorous review, id. at 3 & n.5, and noted the conclusion of the U.S. combat mission in Iraq. Id. Because invocation of FOIA exemptions is subject to de novo review, the District Court held that the Secretary of Defense’s conclusion regarding the potential danger to U.S. citizens, military officials, and civilian personnel was subject to de novo review. Id.
In response to the Court’s ruling, the government submitted an affidavit describing the elaborate, multi-level system the Pentagon had employed for reviewing and analyzing the photographs and the risk posed by their release. First, the Army’s Associate Deputy General Counsel reviewed and sorted the photographs. Second, the Department of Defense’s Office of General Counsel (“OGC”) reviewed those determinations based on examination of a representative sample of the photographs. Third, the sample photos were provided to three senior uniformed officials, the Chairman of the Joint Chiefs of Staff, the Commander of U.S. Central Command, and the Commander of U.S. Forces-Afghanistan, to assess the risk attendant their release. Fourth, the General Counsel considered whether certification was appropriate, based on the pictures and the recommendations of the three military officials. Fifth, the General Counsel presented the photographs and the military officials’ recommendation to the Secretary, who made the ultimate decision. Id. at *4.
A separate affidavit by the Vice Director of Operations for the Joint Staff explained the danger of releasing the photos:
The photographs are susceptible to use as propaganda to incite a public reaction and could be used as recruiting material to attract new members to join enemy forces. . . . Significantly, ISIL [more popularly referred to as ISIS] has a particular interest in using imagery associated with U.S. detention practices as part of its propaganda and recruitment efforts.
Id. The District Judge remained unsatisfied. While noting that the Secretary need not personally review each photograph, the District Judge found numerous deficiencies in the review process. Id. at *5. First, the Secretary had to “explain the terms of his delegation” of the preliminary stages of the review to subordinates. Id. Second, the Secretary had to “demonstrate knowledge of the contents of the individual photographs,” not just the commanders’ conclusions. Id. Third, the Secretary needed to “describe the categories of objectionable content contained in the photographs, identify how many photographs fit into each category, and specify the type of harm that would result from disclosing such content.” Id. The government declined the opportunity to meet such rigorous standards, seeking appellate review instead. Id.
The 2015 Certification
Meanwhile, in November 2015, the Secretary of Defense renewed the certification for all but 198 pictures, after an even more elaborate process, which involved:
(1) an OGC attorney’s creation of a representative sample of photographs by examining and categorizing each photograph based on the potential to cause PNSDA-cognizable harms;
(2) independent review by commissioned officers on the staff of the Deputy Director for Special Operations, Counterterrorism and Detainee Operations;
(3) review of the results by three OGC attorneys and a uniformed attorney, each of whom independently assessed the likelihood of harm from each photograph;
(4) consultation among the first three levels of reviewers to reach consensus;
(5) OGC review of the judgments made at the first four levels of review;
(6) review of the samples by four senior commanders: the Chairman of the Joint Chiefs of Staff, and the Commanders of U.S. Central Command, U.S. Africa Command, and U.S. Forces, Afghanistan; and
(7) the Defense Secretary’s review of the recommendations, the 198 photographs recommended for disclosure, and the representative sample of the remaining pictures.
Id. The District Court was not satisfied, asserting that it could not conduct “adequate judicial review” because the Government had failed to submit information regarding the category criteria, the sampling methodology, the types of objectionable content, the quantity of photographs within each category, and the specific harm from disclosing the different photographs. Id. at *6. The court also concluded that the Secretary failed to demonstrate that the 2015 Certification was adequately individualized. Id.
The Second Circuit’s Rebuke of the District Court
The Second Circuit found it unnecessary to choose between de novo review and review under the “arbitrary and capricious” standard. In its view, even under a de novo standard, properly applied, the government should have prevailed. Under the de novo standard, affidavits laying out national security authorities’ judgments about the harms of releasing information, if “logical and plausible,” provided a sufficient basis to uphold those judgments. Id. at *6. (This “logical and plausible” standard is used in FOIA exemption 1 cases, which involve classified documents). The panel pointedly reiterated that given the “relative competencies of the executive and the judiciary . . . it is bad law and bad policy to second-guess predictive judgements made by the government’s intelligence agencies,” id. at *7 (citing American Civil Liberties Union v. Department of Justice, 681 F.3d 61, 70-71 (2d Cir. 2012) (quoting Center for National Security Studies v. U.S. Department of Justice, 331 F.3d 918, 927 (D.C. Cir. 2003)). The panel characterized the process the Department of Defense used as “thorough and robust,’ and noted that the first four levels of the 2015 review “would likely have sufficed to meet the Government’s burden.” Id. at *7-*8. The panel rebuked the District Judge for feeling “compelled to explore the details of the Secretary’s decision,” id. at *8, and closed by observing:
“Courts are not well-suited to evaluate the constantly evolving military conditions and national security challenges faced by U.S. forces and personnel. Judges do not abdicate their judicial role by acknowledging their limitations and deferring to an agency’s logical and plausible justification in the context of national security; they fulfill it.”
Id. at *8.
Reasons for Skepticism
Are there reasons the exhibit skepticism toward assertions of the need to continue to withhold Abu Ghraib photographic images almost 15 years after the abuse occurred? Of course. First, while release of the photos may increase the risks to Americans abroad (and at home for that matter) — ISIS’ social media strategy of provides a vivid example — the photographs are undoubtedly embarrassing to the Army. Neither the classification system, FOIA, nor, presumably, the PNSDA is designed to allow the Government to withhold information about wrongdoing merely to avoid embarrassment. Executive Order 13526, §1.7(a)(1)&(2), 75 Fed. Reg. 707 (Dec. 29, 2009); see Ray v. Turner, 587 F.2d 1187, 1201-02, 1209, 1222 (D.C. Cir. 1978)(Wright, J., concurring).
Second, the photos are not being withheld because they disclose plans or capabilities that could frustrate future initiatives, but because they could serve as propaganda for Islamic militants. This argument may be particularly relevant to photographic images, because as one District Judge has noted,“[a] picture may be worth a thousand words,” Judicial Watch v. U.S. Department of Defense, 857 F.Supp.2d 44, 48 (D.D.C. 2012). The Second Circuit seems to have cautiously accepted the proposition that material can be withheld under exemption 1 to prevent its use as propaganda by the Nation’s enemies, Center for Constitutional Rights v. C.I.A., 765 F.3d 161, 166-69 (2014). There are reasons to be cautious about too freely allowing the Government to withhold information from U.S. citizens about what it “is up to” because hostile foreign groups could use the material as anti-U.S. propaganda, as the Second Circuit acknowledged in Center for Constitutional Rights, id. After all, the line between propaganda and provision of information is not always clear.
In the few cases the Government has raised the propaganda value of information as a reason for withholding it, courts have largely been able to avoid deciding the issue, upholding the Government’s classification of the information on other grounds. American Civil Liberties Union v. U.S. Dept. of Defense, 628 F.3d 612, 624 (D.C. Cir. 2011); Judicial Watch, Inc. v. U.S. Dept. of Defense, 715 F.3d 937, 939, 943 (D.C. Cir. 2013)(death images of Osama bin Laden). However, one D.C. Circuit Judge upheld classification of photos because they could be altered or manipulated for propaganda purposes. International Counsel Bureau v. U.S. Department of Defense, 906 F.Supp.2d 1, 6-7 (D.D.C. 2012).
The Standard of Review
There is a strong argument that the review of the Secretary’s PNSDA certifications should be conducted under the “arbitrary and capricious,” rather than de novo, standard of review. While FOIA specifies de novo review, 5 U.S.C. §552(4)(A)(vii), including of the question of whether a document warrants classification (and thus can be withheld under exemption 1), the PNSDA does not. The whole point of exemption 3 is to recognize that some decisions regarding release of documents fall entirely outside the FOIA framework and presumptions. In enacting the PNSDA, Congress obviously made the decision to adopt a presumption that the photographs be withheld, and made the determination turn on a certification by the Secretary of Defense. It also mandated congressional notification, which is not required for classification decisions or invocation of FOIA exemptions. And of course, even under exemption 1, the de novo review has a significant element of deference to the government’s assessment of the potential harm attendant release of national security information, as the Second Circuit noted.
“Hard Look,” “Proceduralized” Review
Courts have struggled with providing meaningful judicial review when the relative institutional competencies heavily favor agencies. Courts faced this question when they began engaging in pre-enforcement review of complex and highly-technical regulations. Courts seemed to lack competence to review the scientific and technical judgments necessary in such cases. For different reasons, courts doubt their competence to second-guess military and diplomatic judgments. In circumstances where the relative competencies disfavor courts, judges have embraced two coping strategies. Under the first, courts have required agencies to expand the administrative record to discuss more fulsomely the basis for their conclusions, and more particularly how they address issues that suggest a contrary result. This provides the basis for “hard look” review. In the context of regulations based on scientific and technical judgment, this first approach has taken the form of requiring agencies to respond to comments in their “concise general statement” of the regulation’s “basis and purpose,”e.g., Automotive Parts & Accessories Ass’n v. Boyd, 407 F.2d 330, 338 (1968) (“on the occasion of this first challenge to the implementation of the new statute it is appropriate for us to remind the Administrator of the ever present possibility of judicial review, and to caution against an overly literal reading of the statutory terms ‘concise’ and ‘general’”); Natural Resources Defense Council, Inc. v. U.S. Nuclear Regulatory Commission, 547 F.2d 633, 646 (D.C. Cir. 1976); Portland Cement Ass’n v. Ruckelshaus, 486 F.2d 375, 393-94 (1973), cert. denied, 417 U.S. 921 (1974).
The second approach involves “proceduralizing” review, diverting attention away from the substance of an agency’s judgments to the robustness of its procedures. See Natural Resources Defense Council v. U.S. Nuclear Regulatory Commission, 547 F.2d 633, 643-45, 653-54 (D.C. Cir. 1976); id at 655-57 (Bazelon, J., concurring); Ethyl Corp. v. EPA, 541 F.2d 1, 66-68 (1976) (en banc) (Bazelon, C. J., concurring), cert. denied, 426 U.S. 941, (1976). As Judge Bazelon explained when concurring in NRDC v. Nuclear Regulatory Commission, “in highly technical areas, where judges are institutionally incompetent to weigh evidence for themselves, a focus on agency procedures will prove less intrusive, and more likely to improve the quality of decisionmaking, than judges ‘steeping’ themselves ‘in technical matters to determine whether the agency has exercised a reasoned discretion.’” This “coping mechanism” is not unique to administrative law. See U.S. v Morrison, 529 U.S. 598, 661-64 (2000) (Breyer, J., dissenting) (suggesting a similar approach in reviewing congressional determination of the scope of its Commerce Clause powers). However, the Supreme Court limited the explicit use of such an approach in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, 435 U.S. 519, 523-25 (1978)(holding that courts cannot mandate procedures not required by the APA, except in rare circumstances); Pension Benefits Guaranty Corp. v. LTV Corp., 496 U.S. 633, 655 (1990)(applying Vermont Yankee in the context of adjudication).
These two approaches are not entirely unrelated. Indeed, Judge Henry Friendly noted that they are virtually indistinguishable in terms of what they require of agencies, at least from the perspective of agency decision-makers. Henry J. Friendly, Some Kind of Hearing, 123 U. Pa. L. Rev. 1267, 1310-15 (1975).
The litigation over the Secretary of Defense’s certification of the Abu Ghraib photos illustrates “hard look” and “proceduralized” review. While recognizing United States v. Morgan, 313 U.S. 409, 420 (1941), the District Court seemed quite intrusive in terms of the manner in which the Secretary was permitted to rely on his subordinates. Even with the elaborate procedures for multiple levels of review, the Judge demanded more detail in terms of the scope of the Secretary’s delegation of authority to his subordinates. However, it is not entirely clear whether the impetus for the elaborate procedure was the District Judge or the Government itself. Sometimes in FOIA (and FOIA-related) cases, the description of elaborate procedures can satisfy a reviewing court that an agency has met its FOIA obligations. This is particularly true in the legion of cases involving the reasonableness of the search for documents responsive to a FOIA request. E.g., Iturralde v. Comptroller of the Currency, 315 F.3d 311, 315 (D.C.Cir.2003); SafeCard Services, Inc. v. S.E.C., 926 F.2d 1197, 1201 (D.C. Cir. 1991). Thus, “the adequacy of a FOIA search is generally determined not by the fruits of the search, but by the appropriateness of the methods used to carry out the search.” Iturralde, 315 F.3d at 315.
Nevertheless, in reviewing the Secretary’s PNSDA certifications, the District Court, at best, seems to have come perilously close to violating the Vermont Yankee principle of not imposing upon agencies procedures beyond those statutorily mandated; the PNSDA does not specify any particular procedures for the Secretary to employ in issuing the required certification.
And the detail required the District Court required regarding how elements of each picture could be tied to the potential for harm is suggestive of the “hard look” doctrine’s demands for extensive administrative records. Indeed, the burden placed on the government by the District Court is atypical of even for FOIA cases. Often the question of withholding is decided on the basis of a Vaughn index. Interestingly, the genesis of the Vaughn index is in the same period in which the D.C. Circuit was debating “hard look” and “proceduralized” review, which, at a minimum, required expansion of administrative records to facilitate judicial review. See Vaughn v. Rosen, 484 F.2d 820 (D.C. Cir. 1973). And it too was designed to address a two-fold competency advantage. The government’s ability to produce massive amounts of documents, and the limitations of an adversarial process in which one of the adversaries lacked access to the documents at issue, confronted the courts with the need to invest massive effort to effectively reviewing Government’s claims that the documents fell within FOIA’s exemptions. Id. at 824-26; see, Ray v. Turner, 587 F.2d at 1203-05 (Wright, J., concurring). Vaughn indexes describe records withheld and the basis for withholding them, so as “to permit adequate adversary testing of the agency’s claimed right to an exemption.” Nat’l Treasury Employees Union v. U.S. Customs Serv., 802 F.2d 525, 527 (D.C.Cir. 1986). There is no set format for a Vaughn index, because “the critical elements of the Vaughn index lie in its function, and not its form.” Kay v. FCC, 976 F. Supp. 23, 35 (D.D.C. 1997). Two cases that illustrate the flexibility of the Vaughn index and courts’ willingness to accept Vaughn indexes that are sufficient to understand the basics of the government’s claim of a privilege to withhold documents —Judicial Watch, Inc. v. HUD, 20 F.Supp.3d 247 (D.D.C. 2014), and Odland v. Federal Energy Regulatory Commission, 34 F.Supp.3d 3 (D.D.C. 2014).
The Saga Continues
In any event, we are less three months away from the deadline for the next triennial certification!
Post-script. This is a particularly fitting subject to discuss in the days following the passing of Senator John McCain. Among his many contributions to this country for which we owe an enormous debt of gratitude was his steadfast stance against torturing detainees, embodied in the Detainee Treatment Act, Pub. L. 109–148, div. A, tit. X, §§ 1001-1006, 119 Stat. 2680, 2739-44 (2005).