Public Employees for Environmental Responsibility v. U.S. Environmental Protection Agency

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA

PUBLIC EMPLOYEES FOR
ENVIRONMENTAL REPSONSIBILITY,

Plaintiff, Civil Action No. 18-cv-2219 (BAH)

v. Chief Judge Beryl A. Howell

U.S. ENVIRONMENTAL PROTECTION
AGENCY,

Defendant.

MEMORANDUM OPINION

The Environmental Protection Agency (“EPA”) uses an Integrated Risk Information

System (“IRIS”) program to analyze selected chemicals found in the environment to identify and

characterize hazards to human health, with the results of this analysis released in final public

assessments to guide EPA policymaking and inform the general public. In January 2018, in

testimony before the Senate Committee on the Environment and Public Works, then-EPA

Administrator Scott Pruitt made comments construed as confirming EPA’s preparation of a new

version of a long-delayed IRIS formaldehyde assessment, which, while not final, was ready to be

publicly released for comment and peer review, but was being “held up” and therefore had not

yet been released. Compl. ¶ 3, ECF No. 1. Prompted by this congressional testimony, plaintiff

Public Employees for Environmental Responsibility (“PEER”), “a non-profit organization

dedicated to research and public education concerning the activities of federal, state, and local

governments,” id. ¶ 2, submitted a request, pursuant the Freedom of Information Act (“FOIA”),

5 U.S.C. § 552, to EPA for, inter alia, this public-ready version of the formaldehyde assessment

referenced by then-Administrator Pruitt, and subsequently filed the instant lawsuit to obtain this
1
agency record. EPA has withheld in full the version of the IRIS formaldehyde assessment

referenced in the congressional testimony—the only document remaining at issue in this

lawsuit—as protected by the deliberative process privilege and therefore exempt from disclosure

under FOIA’s Exemption 5, id. § 552(b)(5).

Pending before the Court are the parties’ cross-motions for summary judgment. See

Def.’s Mot. Summ. J. (“Def.’s Mot.”), ECF No. 21; Pl.’s Cross-Mot. Summ. J. (“Pl.’s Mot.”),

ECF No. 23. For the reasons set forth below, EPA’s Motion for Summary Judgment is granted

and plaintiff’s Cross-Motion for Summary Judgment is denied.

I. BACKGROUND

Pertinent background underlying plaintiff’s FOIA request is described, followed by

review of the FOIA request and the initiation of the instant lawsuit.1

A. EPA’s IRIS Program

EPA’s IRIS program “analyzes selected chemicals found in the environment to identify

and characterize hazards to human health.” Def.’s Mem. Supp. Mot. Summ. J. (“Def.’s Mem.”)

at 1, ECF No. 21. According to EPA, the program’s primary purpose is to produce assessments,

which are documents that “[identify] relevant studies” and “synthesi[ze] . . . evidence from

identified studies across human, animal, and mechanistic lines of information” to identify human

health hazards. Def.’s Mot., Ex. 2, Decl. of Dr. Jennifer Orme-Zavaleta (“1st Orme-Zavaleta

Decl.”) ¶¶ 1, 15, ECF No. 21-2. By “identifying the human health hazard of a chemical, and the

1
In support of its Motion for Summary Judgment, EPA has submitted three separate declarations of Dr.
Jennifer Orme-Zavaleta, who has worked at EPA for more than 38 years and, since 2017, served as the Principal
Deputy Assistant Administrator for Science in EPA’s Office of Research and Development responsible for
administering the IRIS program. See Def.’s Mot., Ex. 2, Decl. of Dr. Jennifer Orme-Zavaleta ¶ 1, ECF No. 21-2;
Def.’s Reply Supp. Mot. Summ. J., Ex. 2, Second Decl. of Dr. Jennifer Orme-Zavaleta, ECF No. 27-2; Def.’s Suppl.
Br., Ex. 1, 3d Decl. of Dr. Jennifer Orme-Zavaleta, ECF No. 33-1. In support of its Cross-Motion for Summary
Judgment, plaintiff has submitted a declaration of Kevin Bell, who has been plaintiff’s staff counsel since September
2018 and whose responsibilities include “day-to-day management of” plaintiff’s FOIA requests and litigation. Decl.
of Kevin H. Bell ¶ 1, ECF No. 23-1.

2
accompanying evaluation of dose-response information for those hazards,” IRIS assessments

“provide the scientific foundation for decision-making to protect public health.” Id. ¶ 7. IRIS

assessments “are part of the EPA’s broader policy-making and decision-making process,” and

are “available for use by EPA’s program and regional offices, as well as other stakeholders, to

inform Agency decisions under relevant statutory authorities.” Def.’s Reply Supp. Mot. Summ.

J., Ex. 2, Second Decl. of Doctor Jennifer Orme-Zavaleta (“2d Orme-Zavaleta Decl.”) ¶ 8, ECF

No. 27-2. “Each IRIS assessment can cover a [single] chemical, a group of related chemicals, or

a complex mixture.” 1st Orme-Zavaleta Decl. ¶ 7. EPA is in the process of developing 16

different IRIS chemical assessments, including the formaldehyde assessment at issue in the

instant lawsuit. Id.

Completion of an assessment for a given chemical under the IRIS program comprises

seven distinct steps, from initial research, analysis and drafting, to internal and interagency

review, to public release of an interim version for public comment and peer review, to further

revision and review in light of such comments, to ultimate release of the final assessment, which

is the agency’s definitive statement on the health risks of the chemical under analysis.2 During

Step One (“Draft Development”), EPA first “undertakes internal scoping to identify EPA

program and regional office needs for an assessment.” Id. ¶ 8. Then, “[p]roblem formulation

frames the scientific questions that will be the focus of systematic reviews conducted as part of

assessment development.” Id. Next, “[d]raft development begins with a comprehensive search

and systematic review of the scientific literature,” after which “EPA provides preliminary

assessment materials to the public and an opportunity for public input on these materials.” Id.

At Step Two (“Agency Review”), “[s]cientists in EPA’s program offices and regions review the

2
“The IRIS process is not a formal regulatory development process;” rather, “EPA created the ‘Steps’ to
better inform the public of EPA’s progress in assessing a particular chemical.” 2d Orme-Zavaleta Decl. ¶ 3.

3
draft assessment,” and the draft assessment is revised as needed. Id. At Step 3 (“Interagency

Science Consultation”), “[o]ther federal agencies, including the Executive Office of the

President, review the draft assessment,” and the draft assessment is again revised as needed. Id.

Step Four (“Public Comment and External Peer Review”), provides that the draft assessment “is

released for public review and comment” as well as “external peer review.” Id. Next, at Step

Five (“Revise Assessment”), “[t]he IRIS Program revises the assessment to address peer review

comments” and “prepare[s] a written response-to-comment document.” Id. At Step Six (“Final

Agency Review”), the revised assessment is again “reviewed by EPA’s program offices and

regions, other federal agencies, and the Executive Office of the President.” Id. Finally, at Step

Seven (“Final Assessment”), “[t]he final IRIS assessment is posted to the IRIS website.” Id.

B. The IRIS Formaldehyde Assessment

The IRIS formaldehyde assessment at issue has been in process for well over a decade,

and, according to plaintiffs, dates back to 1997. See Pl.’s Mem. Supp. Cross-Mot. Summ. J. &

Opp’n Def.’s Mot. Summ. J. (“Pl.’s Opp’n”) at 2, ECF No. 23. In 2005, EPA announced that the

formaldehyde assessment would be sent “for external review during fiscal years 2006 and 2007,”

referencing IRIS Step Four, but that deadline was not met. Id. (omissions in original) (quoting

GOV’T ACCOUNTABILITY OFFICE, GAO-08-440, LOW PRODUCTIVITY AND NEW INTERAGENCY

REVIEW PROCESS LIMIT THE USEFULNESS AND CREDIBILITY OF EPA’S INTEGRATED RISK

INFORMATION SYSTEM 37 (2008), https://www.gao.gov/products/GAO-08-440).

In 2010, EPA released a draft formaldehyde assessment at Step Four of the IRIS process,

thereby providing the opportunity for public comment on and external peer review, and this

document remains publicly available on EPA’s website. 2d Orme-Zavaleta Decl. ¶¶ 4–5. In

2011, the National Academy of Sciences (“NAS”) conducted a peer review of the draft and

4
provided “significant comments” necessitating substantive revisions to the draft. Id. ¶ 6.3 In

response to those comments, IRIS “never finalized” “the 2010 [draft] [a]ssessment,” id., and

instead began working on a new version of the formaldehyde assessment to address the 2011

NAS peer review comments, returning the formaldehyde assessment to Step One of the IRIS

program, requiring EPA to “repeat Agency review and interagency science consultation before

being released” for public comment and renewed peer review. Id. (internal quotation marks

omitted).4 Since 2011, the new draft formaldehyde assessment has not yet undergone Steps Two

and Three—Agency Review and Interagency Science Consultation—and therefore has not yet

been advanced to IRIS Step Four or released for public review. See 2d Orme-Zavaleta Decl. ¶¶

6–7.

As of March 2019, EPA was “no longer working on the [IRIS] Formaldehyde

Assessment,” because “other chemicals ha[d] been identified as higher priorities for the IRIS

assessment process at th[at] time,” although a separate EPA office had begun working on “a risk

evaluation for formaldehyde under the Toxic Substances Control Act.” 1st Orme-Zavaleta Decl.

¶ 9; Def.’s Suppl. Br., Ex. 1, Third Decl. of Doctor Jennifer Orme-Zavaleta (“3d Orme-Zavaleta

Decl.”) ¶ 4, ECF No. 33-1. On March 26, 2021, however, EPA announced that “the IRIS

assessment of formaldehyde . . . ha[d] been unsuspended, and that a public milestone timeline”

3
NAS is a federally chartered, private corporation, 36 U.S.C. § 150301, whose statutory mandate is to, “[o]n
request of the United States Government, . . . investigate, examine, experiment, and report on any subject of science
or art,” id. § 150303. The NAS peer review “identified numerous significant recommendations for improving the
science that underlies the formaldehyde IRIS assessment.” Pl.’s Reply Supp. Cross-Mot. Summ. J. at 4, ECF No.
29.
4
EPA does not clarify when comments received at Step Four necessitate reworking an assessment from the
beginning, thereby returning the draft assessment to IRIS Step One, and when such comments may simply be
addressed in Step Five, when the “[t]he IRIS Program revises the assessment to address peer review comments,” 1st
Orme-Zavaleta Decl. ¶ 8. EPA notes, however, that the formaldehyde assessment is not the only draft assessment
that has been returned to IRIS Step One in response to comments received at Step Four. 2d Orme-Zavaleta Decl.,
Ex. A, IRIS Assessment Status at 2–3, ECF No. 27-2 (noting that IRIS assessments for acrylonitrile and n-butanol
“repeat[ed] agency review and interagency consultation before being released for additional public comment”
following comments received at IRIS Step Four).

5
for the formaldehyde assessment would be provided in June 2021. 3d Orme-Zavaleta Decl. ¶¶

3–4; see also id., Ex. 1, Mar. 11, 2021 Memo from Jennifer Orme-Zavaleta, Ph.D., to Acting

Assistant Administrators, Acting Regional Administrators, and Deputies of EPA, at 2 (“[EPA]

will unsuspend the IRIS assessment of formaldehyde and will add formaldehyde to the current

IRIS agenda. The next step for this assessment is agency review.”).

C. January 2018 Meeting of the American Chemistry Council and IRIS Staff

Meanwhile, more than a year before EPA’s suspension of the IRIS formaldehyde

assessment, while revisions were still underway following the NAS peer review, the

Formaldehyde Panel of the American Chemistry Council (“ACC”) met, on January 24, 2018,

with IRIS staff to discuss the status of the formaldehyde assessment. See Pl.’s Opp’n at 3; see

also Pl.’s Cross-Mot., Ex. 1, Decl. of Kevin Bell (“Bell Decl.”), ECF No. 23-1; 2d Orme-

Zavaleta Decl. ¶ 9. According to EPA, “[t]he IRIS Program . . . frequently [meets] with

members of the public to hear stakeholder perspectives,” such as “new or ongoing research

initiatives,” 2d Orme-Zavaleta Decl. ¶ 9, and requests for such meetings are publicly disclosed

on IRIS’s website, see id. ¶ 9 & n.8; IRIS Calendar: Meetings Requested by Specific Members of

the Public, U.S. ENVT’L PROTECTION AGENCY, https://iris.epa.gov/Events/#stakeholderMeetings

(last visited June 9, 2021). EPA “confirm[s]” that, although the draft assessment was discussed

“at th[e] [January 24, 2018] meeting[,] EPA did not share any copies of the [most recent draft

formaldehyde] Assessment or any portions of the draft with the [ACC] Panel.” 2d Orme-

Zavaleta Decl. ¶ 9.

Two days after the meeting with IRIS staff, an ACC panel member sent a letter to the

agency on behalf of the ACC, “stress[ing] the importance of producing a revised formaldehyde

IRIS assessment that fully implements and resolves scientifically the recommendations of the

2011 National Academy of Sciences . . . report” and noting that ACC “left the meeting very
6
alarmed and troubled” with the methodological approach IRIS was employing in revising the

formaldehyde assessment.” Letter from Kimberly Wise White, ACC, to Dr. Orme-Zavaleta at 1

(Jan. 26, 2018) (quoted in Pl.’s Opp’n at 4).5 In a further critique, the letter stated that “the

revised draft IRIS assessment has not revisited the science but instead will be a restructuring of

the [old] draft,” despite concern that “the previous draft relied on studies that ha[d] been shown

in recent years to have significant scientific and methodological issues.” Id.6 ACC insisted that

“a revised draft IRIS assessment must revisit all previous conclusions, demonstrate effective and

science-based integrations of all the lines of evidence and meet the standards of scientific

integrity and transparency requested by the NAS and the public.” Id.

D. Then-EPA Administrator Scott Pruitt’s Congressional Testimony

Just two days after the ACC sent its letter to the EPA concerning the IRIS formaldehyde

assessment, Scott Pruitt, at the time the EPA Administrator, testified, on January 30, 2018,

before the U.S. Senate Committee on the Environment and Public Works. During Pruitt’s

testimony, one member of the Committee voiced his “understanding that the EPA has finalized

its conclusion that formaldehyde causes leukemia and other cancers, and that that completed new

assessment is ready to be released for public review[,] [b]ut [the assessment] is still being held

up.” Def.’s Reply Supp. Mot. Summ. J. (“Def.’s Reply”) at 4, ECF No. 27 (quoting U.S. Senate

Comm. on Envt. & Pub. Works, S. Hr’g 115-325, Vol. 1, Oversight Hearing to Receive

Testimony from Environmental Protection Agency Administrator Scott Pruitt (“Senate Hr’g”)

5
A copy of the letter was obtained by the chemical-industry research company Chemical Watch and is
available online at http://files.chemicalwatch.com/ACC%20Formaldehyde%20Panel%20Letter%20to%20EPA%20-
%20January%2026%202018%20-%20Final%20(1).pdf.
6
EPA objects that ACC’s characterization of the revisions to the IRIS formaldehyde assessment “is
incorrect,” 2d Orme-Zavaleta Decl. ¶ 10, explaining that, although the current draft of the formaldehyde assessment
“does present information and conclusions in a new structure compared to” the 2010 IRIS Step Four formaldehyde
assessment, the most recent version “also incorporates new studies and updated analyses to reflect the latest science
on formaldehyde toxicity,” id.

7
Pruitt Testimony”) 278 (Jan. 30, 2018) (question of Sen. Edward Markey). He then asked Pruitt,

“Can you give us a status update as to the EPA’s handling of the formaldehyde issue and the

conclusion that it in fact does cause leukemias and other cancers?” Id. (quoting Senate Hr’g at

278).

In response, Pruitt stated that “[his] understanding is similar to [Markey’s], but [he

would] confirm that and provide the information to [Markey] from the [IRIS] program office.”

Id. (quoting Senate Hr’g at 278). Markey followed up, asking if Pruitt would “commit to

releasing that report, which is already completed, in a short period of time once [Pruitt] ha[d]

reviewed it, if in fact [it] meets the standards which [the] EPA staff has already established that it

does.” Id. (quoting Senate Hr’g at 278). Pruitt balked at this request, instead “commit[ting] . . .

that [he would] look into that and make sure [Markey’s] office is aware of what we have and

when we can release it.” Id. (quoting Senate Hr’g at 278).

On May 17, 2018, Senator Markey, among others, sent a letter to then-Administrator

Pruitt requesting that he follow through on his commitment to update Markey’s office about the

status of the IRIS formaldehyde assessment. See Pl.’s Reply Supp. Cross-Mot. Summ. J. (“Pl.’s

Reply”) at 3–4, ECF No. 29. EPA responded that “[t]he National Academy of Science . . .

identified numerous significant recommendations for improving the science that underlies the

formaldehyde assessment,” and that “the Agency is working to fully implement the NAS

recommendations in all IRIS assessments released moving forward.” Id. at 4.

E. Plaintiff’s FOIA Request and the Initiation of the Instant Lawsuit

Spurred by the ACC letter and Pruitt’s public acknowledgement that his understanding of

the status of the IRIS formaldehyde assessment was similar to that of Senator

Markey’s―namely, that EPA had finalized its conclusion that formaldehyde causes leukemia

and other cancers―and news coverage of the IRIS formaldehyde assessment, see Pl.’s Opp’n at
8
6–7 (citing Annie Snider, Sources: EPA Blocks Warnings on Cancer-Causing Chemical,

POLITICO (July 6, 2018), https://www.politico.com/story/2018/07/06/epa-formaldehyde-

warnings-blocked-696628), plaintiff submitted a FOIA request to EPA on July 9, 2018, see 1st

Orme-Zavaleta Decl., Ex. A, FOIA Request, ECF No. 21-2. The request referenced Pruitt’s

January 2018 Senate testimony and sought records related to the IRIS formaldehyde assessment,

including “[a]ny versions of the . . . draft [formaldehyde] health assessment.” Id. at 1–2.7

On September 25, 2018, plaintiff filed the instant lawsuit, seeking, in a one-count

complaint, to “compel” EPA “to disclose records wrongfully withheld in failing to respond” to

its FOIA request for the five above categories of documents “within the statutory deadline.”

Compl. ¶ 1. Thereafter, the parties negotiated a search plan to identify, process, and produce

records responsive to plaintiff’s request. See Bell Decl. ¶ 5. EPA then released, in three separate

batches in May, June, and July 2018, 138 records in full and 174 redacted records, and withheld

in full 99 records. Def.’s Mem. at 2; 1st Orme-Zavaleta Decl. ¶¶ 11–12. EPA also provided a

withholding index itemizing the withheld documents and explaining the reason each was

withheld. See Bell Decl., Ex. 2, Withholding Index, ECF No. 23-1; see also Def.’s Mem. at 2.8

7
Plaintiff’s FOIA request sought a total of five categories of records but, given the parties’ agreement to
focus on a single disputed record here, the other requested records are not at issue.
8
The parties quibble over whether the withholding index EPA provided to plaintiff qualifies as a Vaughn
index, which “describes the documents withheld or redacted [by the agency] and the FOIA exemptions invoked, and
explains why each exemption applies,” Prison Legal News v. Samuels, 787 F.3d 1142, 1145 n.1 (D.C. Cir. 2015);
see also Vaughn v. Rosen, 484 F2d 820 (D.C. Cir. 1973). Plaintiff views the withholding index as a Vaughn index,
see Pl.’s Mem. at 11 (“[Plaintiff] did on several occasions ask for an index of withholdings, and EPA provided
one.”), whereas EPA states that plaintiff “did not request a Vaughn index for the withholdings,” Pl.’s Mem. at 2, and
a “Vaughn index was not necessary because the parties agreed that the only outstanding issue with regard to EPA’s
response concerns a single document that EPA withheld in full,” id. EPA denies that “[t]he index provided by EPA”
to plaintiff was a Vaughn index because a Vaughn index “has particular requirements as set forth in Vaughn v.
Rosen,” Def.’s Reply at 5, without explicating which of those specific requirements are lacking in the index actually
provided to plaintiff. In any event, whether the index fully complies with the requirements of Vaughn is of no
moment, since, as explained in greater detail below, see infra Part II, an agency may justify its application of FOIA
exemptions through declarations or affidavits in addition to a Vaughn index. Here, the agency’s declarations
adequately describe the contents of the lone withheld record at issue here to permit a determination whether the
withholding was proper. See infra Part III.A.2.

9
Among the fully withheld records were all drafts of the IRIS formaldehyde assessment, which

EPA claimed were exempt from disclosure pursuant to the deliberative process privilege of

FOIA Exemption 5. Def.’s Mem. at 2; see also Bell Decl. ¶ 12.

Plaintiff objected to EPA’s application of the deliberative process privilege to the drafts

of the formaldehyde assessment and proposed, in September 2019, that “the parties narrow any

dispute for summary judgment briefing” by picking one representative version of the draft

formaldehyde assessment and focusing their briefing on that version “as a test case.” Bell Decl.

¶¶ 12–14. At plaintiff’s suggestion, the parties settled on a document with the filename

“Formaldehyde Main Text 102417 06-13-18.docx” (“2018 Version-Formaldehyde Assessment”)

as an appropriately representative document, primarily because this was the most recently dated

version of the formaldehyde assessment, “appeared to be of sufficient length (over 500 pages),”

and the filename did not suggest to plaintiff that the document “was an individual’s personal

comments or other clearly deliberative content.” Id. ¶¶ 15–16. Accordingly, the parties resolved

their dispute as to all the remaining partially and fully withheld documents, and informed the

Court that they “were unable to resolve their dispute over the application of FOIA exemption

5 . . . to a single document which EPA withheld in full,” namely the 2018 Version-Formaldehyde

Assessment. Joint Status Report & Proposed Briefing Schedule (“JSR”) ¶ 4, ECF No. 19. The

parties therefore stated that they “ha[d] agreed to narrow the subject of their dispute on summary

judgment to that document.” Id.

Per their agreement, the parties cross-moved for summary judgment on the applicability

of the deliberative process privilege to the 2018 Version-Formaldehyde Assessment. See Def.’s

Mot.; Pl.’s Cross-Mot. Following a lengthy briefing schedule, see Scheduling Order (Mar. 16,

2020); Amended Scheduling Orders (Apr. 28, 2020; June 30, 2020; July 30, 2020; Aug 4, 2020),

10
including Court-ordered supplemental briefing to address U.S. Fish & Wildlife Service v. Sierra

Club, 141 S. Ct. 777 (2021), an intervening Supreme Court decision on the application of the

deliberative process privilege to draft agency work product, see Min. Order (Apr. 16, 2021),

those cross-motions have been fully briefed and are ripe for resolution. See Def.’s Mem.; Pl.’s

Opp’n; Def.’s Reply; Pl.’s Reply; Pl.’s Suppl. Br. Concerning the U.S. Supreme Court’s

Decision in U.S. Fish & Wildlife Service v. Sierra Club & Mootness (“Pl.’s Suppl. Mem.”), ECF

No. 32; Def.’s Suppl. Br. (“Def.’s Suppl. Mem.”), ECF No. 33; Pl.’s Reply to Def.’s Suppl. Br.

(“Pl.’s Suppl. Reply”), ECF No. 36; Def.’s Reply to Pl.’s Suppl. Br. (“Def.’s Suppl. Reply”),

ECF No. 35.9 For the reasons set out below, defendant’s Motion for Summary Judgment is

granted and plaintiff’s Cross-Motion for Summary Judgment is denied.

II. LEGAL STANDARD

Under Federal Rule of Civil Procedure 56, “‘[a] party is entitled to summary judgment

only if there is no genuine issue of material fact and judgment in the movant’s favor is proper as

a matter of law.’” Soundboard Ass’n v. FTC, 888 F.3d 1261, 1267 (D.C. Cir. 2018) (quoting

Ctr. for Auto Safety v. Nat’l Highway Traffic Safety Admin., 452 F.3d 798, 805 (D.C. Cir. 2006));

see also Fed. R. Civ. P. 56(a). “‘In FOIA cases, summary judgment may be granted on the basis

of agency affidavits if they contain reasonable specificity of detail rather than merely conclusory

statements, and if they are not called into question by contradictory evidence in the record or by

evidence of agency bad faith.’” Aguiar v. DEA, 865 F.3d 730, 734–35 (D.C. Cir. 2017) (quoting

9
In its minute order directing the parties to address the impact of Sierra Club, the Court also noted a March
11, 2021 news article reporting that EPA was “resuming its years-delayed analysis of formaldehyde’s health
effects,” in turn suggesting that the IRIS formaldehyde assessment whose release plaintiff effectively seeks to
compel by way of a FOIA lawsuit could soon be released to the public at IRIS Step Four. Min. Order (Apr. 16,
2021) (quoting Pat Rizzuto, “Vanished" EPA Formaldehyde Health Effects Probe Reemerges, Bloomberg Law
(Mar. 11, 2021), https://www.bloomberglaw.com/bloomberglawnews/environment-and-energy/X4C0LD20000000).
Notwithstanding this development, the parties confirmed that plaintiff still sought the Draft Assessment via its FOIA
request and that the parties sought resolution of their pending cross-motions for summary judgment. See Pl.’s
Suppl. Mem. at 13–14; Def.’s Suppl. Mem. at 4–5.

11
Judicial Watch, Inc. v. U.S. Secret Serv., 726 F.3d 208, 215 (D.C. Cir. 2013)); see also Students

Against Genocide v. Dep’t of State, 257 F.3d 828, 833 (D.C. Cir. 2001) (“[A]n agency is entitled

to summary judgment if no material facts are in dispute and if it demonstrates ‘that each

document that falls within the class requested either has been produced . . . or is wholly exempt

from the Act’s inspection requirements.’” (omission in original) (quoting Goland v. CIA, 607

F.2d 339, 352 (D.C. Cir. 1978))). Most FOIA cases “can be resolved on summary judgment.”

Brayton v. Off. of U.S. Trade Representative, 641 F.3d 521, 527 (D.C. Cir. 2011).

FOIA was enacted “to promote the ‘broad disclosure of Government records’ by

generally requiring federal agencies to make their records available to the public on request.”

DiBacco v. U.S. Army (“DiBacco I”), 795 F.3d 178, 183 (D.C. Cir. 2015) (quoting Dep’t of

Justice v. Julian, 486 U.S. 1, 8 (1988)). Agencies are therefore statutorily mandated to

“make . . records promptly available to any person” who submits a request that “reasonably

describe such records” and “is made in accordance with [the agency’s] published rules.” 5

U.S.C. § 552(a)(3)(A). To balance the public’s interest in governmental transparency and

“‘legitimate governmental and private interests [that] could be harmed by release of certain types

of information,’” Judicial Watch, Inc. v. U.S. Dep’t of Def., 913 F.3d 1106, 1108 (D.C. Cir.

2019) (quoting FBI v. Abramson, 456 U.S. 615, 621 (1982)), FOIA contains nine exemptions, set

forth in 5 U.S.C. § 552(b), which “are ‘explicitly made exclusive’ and must be ‘narrowly

construed,’” Milner v. Dep’t of Navy, 562 U.S. 562, 565 (2011) (first quoting EPA v. Mink, 410

U.S. 73, 79 (1979); and then quoting Abramson, 456 U.S. at 630); see also Murphy v. Exec. Off.

for U.S. Att’ys, 789 F.3d 204, 206 (D.C. Cir. 2015); Citizens for Resp. & Ethics in Wash. v. U.S.

Dep’t of Justice (“CREW”), 746 F.3d 1082, 1088 (D.C. Cir. 2014). “[T]hese limited exemptions

12
do not obscure the basic policy that disclosure, not secrecy, is the dominant objective of the

Act.” Dep’t of Air Force v. Rose, 425 U.S. 352, 361 (1976).

FOIA authorizes federal courts “to enjoin the agency from withholding agency records

and to order the production of any agency records improperly withheld from the complainant.” 5

U.S.C. § 552(a)(4)(B). When an agency invokes an exemption to disclosure, district courts must

“determine de novo whether non-disclosure was permissible.” Elec. Priv. Info. Ctr. v. U.S. Dep’t

of Homeland Sec., 777 F.3d 518, 522 (D.C. Cir. 2015). The statute “places the burden ‘on the

agency to sustain its action,’ and the agency therefore bears the burden of proving that it has not

‘improperly’ withheld the requested records.” Citizens for Resp. & Ethics in Wash. v. U.S. Dep’t

of Justice, 922 F.3d 480, 487 (D.C. Cir. 2019) (first quoting 5 U.S.C. § 552(a)(4)(B); and then

quoting U.S. Dep’t of Justice v. Tax Analysts, 492 U.S. 136, 142 n.3 (1989)); see also U.S. Dep’t

of Justice v. Landano, 508 U.S. 165, 171 (1993) (“The Government bears the burden of

establishing that the exemption applies.”); DiBacco v. U.S. Dep’t of Army (“DiBacco II”), 926

F.3d 827, 834 (D.C. Cir. 2019) (“‘An agency withholding responsive documents from a FOIA

release bears the burden of proving the applicability of claimed exemptions,’ typically through

affidavit or declaration.” (quoting DiBacco I, 795 F.3d at 195)). This burden does not shift even

when the requester files a cross-motion for summary judgment because the agency ultimately

“bears the burden to establish the applicability of a claimed exemption to any records or portions

of records it seeks to withhold,” Am. Immigr. Laws. Ass’n v. Exec. Off. for Immigr. Rev., 830

F.3d 667, 673 (D.C. Cir. 2016), while “[t]he burden upon the requester is merely ‘to establish the

absence of material factual issues before a summary disposition of the case could permissibly

occur,’” Pub. Citizen Health Rsch. Grp. v. FDA, 185 F.3d 898, 904–05 (D.C. Cir. 1999) (quoting

Nat’l Ass’n of Gov’t Emps. v. Campbell, 593 F.2d 1023, 1027 (D.C. Cir. 1978)).

13
 An agency may prove the applicability of claimed exceptions through a Vaughn index or

supporting affidavits or declarations, or both, that “describe[] the justifications for withholding

the information with specific detail, demonstrate[] that the information withheld logically falls

within the claimed exemption, and [are] not contradicted by contrary evidence in the record or by

evidence of the agency’s bad faith.” DiBacco II, 926 F.3d at 834 (internal quotation marks and

citation omitted); see also, e.g., CREW, 746 F.3d at 1088; Poitras v. Dep’t of Homeland Sec.,

303 F. Supp. 3d 136, 150 (D.D.C. 2018) (“An agency may carry its burden of showing an

exemption was properly invoked by submitting sufficiently detailed affidavits or declarations, a

Vaughn index of the withheld documents, or both, to demonstrate that the government has

analyzed carefully any material withheld and provided sufficient information as to the

applicability of an exemption to enable the adversary system to operate.”). “‘Ultimately, an

agency’s justification for invoking a FOIA exemption is sufficient if it appears logical or

plausible.’” Judicial Watch, Inc. v. U.S. Dep’t of Def., 715 F.3d 937, 941 (D.C. Cir. 2013) (per

curiam) (quoting ACLU, 628 F.3d at 619).

III. DISCUSSION

The parties dispute whether EPA properly asserted Exemption 5’s deliberative process

privilege to withhold in full the 2018 Version-Formaldehyde Assessment. Application of the

deliberative process privilege to this document is discussed below, following explanation of the

legal standard for withholding a record pursuant to Exemption 5 and of the agency’s description

of the 2018 Version-Formaldehyde Assessment.

A. FOIA Exemption 5 Legal Standard

Exemption 5 protects from disclosure “inter-agency or intra-agency memorandums or

letters that would not be available by law to a party other than an agency in litigation with the

agency.” 5 U.S.C. § 552(b)(5). “‘Among th[e] privileges protected by Exemption 5 is
14
the . . . deliberative process privilege.’” Judicial Watch, Inc. v. U.S. Dep’t of Defense, 847 F.3d

735, 739 (D.C. Cir. 2017) (alteration and omission in original) (quoting Arthur Andersen & Co.

v. IRS, 679 F.2d 254, 257 (D.C. Cir. 1982)); see also Abtew v. U.S. Dep’t of Homeland Sec., 808

F.3d 895, 898 (D.C. Cir. 2015). “To protect agencies from being ‘forced to operate in a

fishbowl,’ the deliberative process privilege shields from disclosure ‘documents reflecting

advisory opinions, recommendations and deliberations comprising part of a process by which

governmental decisions and policies are formulated.’” Sierra Club, 141 S. Ct. at 785 (first

quoting Mink, 410 U.S. at 87; and then quoting NLRB v. Sears, Roebuck & Co. (“Sears”), 421

U.S. 132, 150 (1975)); see also Pavement Coatings Tech. Council v. U.S. Geological Survey, 995

F.3d 1014, 1021 (D.C. Cir. 2021) (noting that deliberative process privilege “was ‘intended to

protect not simply deliberative material, but also the deliberative process of agencies.’” (quoting

Montrose Chem. Corp. v. Train, 491 F.2d 63, 71 (D.C. Cir. 1974))). It “is rooted in ‘the obvious

realization that officials will not communicate candidly among themselves if each remark is a

potential item of discovery and front page news.’” Sierra Club, 141 S. Ct. at 785 (quoting Dep’t

of Interior v. Klamath Water Users Protective Ass’n, 532 U.S. 1, 8–9 (2001)); see also Judicial

Watch, Inc., 847 F.3d at 739 (noting that the deliberative process privilege is predicated on the

theory that “agencies craft better rules when their employees can spell out in writing the pitfalls

as well as the strengths of policy options, coupled with the understanding that employees would

be chilled from such rigorous deliberation if they feared it might become public”). The privilege

is intended “[t]o encourage candor, which improves agency decisionmaking,” by “blunt[ing] the

chilling effect that accompanies the prospect of disclosure.” Sierra Club, 141 S. Ct. at 785; see

also Machado Amadis v. U.S. Dep’t of State, 971 F.3d 364, 371 (D.C. Cir. 2020) (noting that the

deliberative process privilege is intended to “protect[] ‘debate and candid consideration of

15
alternatives within an agency,’ thus improving agency decisionmaking”) (quoting Jordan v.

Dep’t of Justice, 591 F.2d 753, 772 (D.C. Cir. 1978) (en banc))).

“To qualify for the deliberative process privilege, an intra-agency memorandum must be

both pre-decisional and deliberative.” Abtew, 808 F.3d at 898 (citing Coastal States Gas Corp.

v. Dep’t of Energy (“Coastal States”), 617 F.2d 854, 866 (D.C. Cir. 1980)); see also Pavement

Coatings Tech. Council, 995 F.3d at 1021; Hall & Assocs. v. EPA, 956 F.3d 621, 624 (D.C. Cir.

2020).10 The Supreme Court clarified the contours of these requirements in Sierra Club, 141 S.

Ct. 777, emphasizing that the privilege protects agency documents that “reflect [an agency’s]

preliminary view” rather than its “final decision” on a matter, noting that “[t]he privilege . . .

distinguishes between predecisional, deliberative documents, which are exempt from disclosure,

and documents reflecting a final agency decision and the reasons supporting it, which are not,”

id. at 785–86. “Documents are ‘predecisional’ if they were generated before the agency’s final

decision on the matter, and they are ‘deliberative’ if they were prepared to help the agency

formulate its position.” Id. at 786 (citing Sears, 421 U.S. at 150–52; Renegot. Bd. v. Grumman

Aircraft Eng’g Corp., 421 U.S. 168, 184–86, 190 (1975)); see also Judicial Watch, Inc., 847

F.3d at 739 (“Documents are ‘predecisional’ if they are ‘generated before the adoption of an

agency policy,’ and ‘deliberative’ if they ‘reflect[] the give-and-take of the consultative

process.’” (alteration in original) (quoting Pub. Citizen, Inc. v. Off. of Mgmt. & Budget, 598 F.3d

865, 874 (D.C. Cir. 2010))). “There is considerable overlap between these two prongs because a

document cannot be deliberative unless it is predecisional.” Sierra Club, 141 S. Ct. at 786.

10
At the outset, an agency can claim the deliberative process privilege only with respect to “inter-agency or
intra-agency memorandums or letters.” 5 U.S.C. § 552(b)(5). The 2018 Version-Formaldehyde Assessment has not
been shared outside the agency and “is a purely internal agency document,” 1st Orme-Zavaleta Decl. ¶ 13, and thus
satisfies Exemption 5’s threshold requirement that a record be an inter-agency or intra-agency communication.

16
 Sierra Club instructs that “a court must evaluate the documents ‘in the context of the

administrative process which generated them’” to decide whether a document represents an

agency’s final decision. Id. (quoting Sears, 421 U.S. at 138). “[D]etermining whether an

agency’s position is final for purposes of the deliberative process privilege is a functional rather

than formal inquiry,” id. at 788, that focuses on “whether [the record] communicates a policy on

which the agency has settled,” id. at 786. To answer this question, “courts must consider

whether the agency treats the document as its final view on the matter.” Id. (citing Sears, 421

U.S. at 161). “[O]nce cited as the agency’s final view, the document reflects ‘the consummation

of the agency’s decisionmaking process’ and not a ‘merely tentative position,’” and therefore

loses the protection of Exemption 5. Id. (quoting Bennett v. Spear, 520 U.S. 154, 177–78

(1997)). On the other hand, “a document that leaves agency decisionmakers ‘free to change their

minds’ does not reflect the agency’s final decision” and is exempt from disclosure. Id. (quoting

Grumman Aircraft Eng’g Corp., 421 U.S. at 189–90 & n.26). A record’s “real operative effect”

is probative of its finality, but is assessed by reference “to the legal, not practical, consequences

that flow from an agency action.” Id. at 787 (citing Sears, 421 U.S. at 159 n.25, 160). Thus, a

document that reflects an agency view “that [is] subject to change,” id. at 786, may be exempt as

predecisional and deliberative even if it in fact “has the effect of changing an agency’s course,”

id. at 788. “The government, not the requester, must identify the deliberative process to which

any record relates.” Ctr. for Investigative Reporting v. U.S. Customs & Border Prot., 436 F.

Supp. 3d 90, 101 (D.D.C. 2019) (citing 100Reporters LLC v. U.S. Dep’t of Justice, 248 F. Supp.

3d 115, 152 (D.D.C. 2017)).

B. Description of the 2018 Version-Formaldehyde Assessment

According to EPA, the 2018 Version-Formaldehyde Assessment at issue “is one working

draft version of the Formaldehyde Assessment that was being developed in Step 1 of the IRIS
17
process.” 1st Orme-Zavaleta Decl. ¶ 14. It “is a Microsoft Word document that contains

numerous comments and redlined edits from multiple staff members . . . that reflect

ongoing . . . deliberations as to the type of analysis EPA should be conducting, which scientific

studies should be used as part of the assessment, and the type, scope, and substantive content of

the agency’s proposed conclusions and recommendations.” Id. ¶ 15. Since the IRIS

formaldehyde assessment remains in IRIS Step One, this document has not been subject to IRIS

Steps Two and Three (i.e., Agency Review and Interagency Science Consultation), id., and, in

fact, “is a purely internal agency document” that “was not shared beyond” the EPA’s Office of

Research and Development, the particular office that administers the IRIS program, id. ¶ 13.

Furthermore, “since EPA staff continued to work on the Formaldehyde Assessment after the date

the [2018 Version-Formaldehyde Assessment] was created, there are later versions of the draft

Formaldehyde Assessment” that postdate the 2018 Version-Formaldehyde Assessment. Id. ¶ 14.

The process of drafting the 2018 Version-Formaldehyde Assessment began “with scoping

and problem formulation to identify the level of information needed,” and that process “reflects

the identification of relevant studies across multiple scientific disciplines through a structured

literature search; evaluation of study methods; analysis and synthesis of evidence from identified

studies across human, animal, and mechanistic lines of information; integration of the available

evidence; and hazard identification.” Id. ¶ 15. In the process of drafting the assessment, IRIS

may also need to identify one or more studies to “use[] as the basis to derive toxicity values” for

formaldehyde. Id. As EPA explains, “[t]he selection, organization, and analysis of factual

information is precisely the purpose of” the IRIS formaldehyde assessment. Id. As such,

according to the agency, the 2018 Version-Formaldehyde Assessment “reflects staff opinions,

18
analysis, and discussions relating to EPA’s decision-making processes about the potential risks

associated with formaldehyde inhalation.” Id.

C. The 2018 Version-Formaldehyde Assessment Is Predecisional and
Deliberative

As detailed below, the administrative context in which the 2018 Version-Formaldehyde

Assessment was produced, the substance of this document, and the fact that this document

predates the still-nonexistent final IRIS formaldehyde assessment, together, indicate that this

document “reflect[s] a preliminary view,” rather than a “final decision,” about the effects of

formaldehyde, Sierra Club, 141 S. Ct. at 786, and therefore is protected by the deliberative

process privilege.

First is “the obvious point” that EPA “identified th[is] document[] as [a] “‘draft[],’”

which is “by definition, a preliminary version of a piece of writing subject to feedback and

change.” Id. Of course, EPA’s application of the label “draft” is not dispositive, yet here, as in

Sierra Club, “administrative context confirms that the draft[] [is] what [it] sound[s] like: [an]

opinion[] that [is] subject to change.” Id. Specifically, an assessment at IRIS Step One must still

proceed through IRIS Steps Two and Three—Agency Review and Interagency Science

Consultation—during which the draft is substantively reviewed and, crucially, subject to revision

as needed. See 1st Orme-Zavaleta Decl. ¶ 8. Thus, the agency’s view of the 2018 Version-

Formaldehyde Assessment being at Step One of the IRIS program thus “specifically

contemplates further review by the agency after receipt of the draft, and with it, the possibility of

changes to the [assessment] after” IRIS Step One is completed, Sierra Club, 141 S. Ct. at 787

(emphasis in original). This possibility for subsequent changes indicates that the 2018 Version-

Formaldehyde Assessment is predecisional and deliberative. See id.; see also, e.g., Am. Soc’y for

Prevention of Cruelty to Animals v. Animal & Plant Health Inspection Serv., 19 Civ. 3112

19
(NRB), 2021 WL 1163627, at *12 (S.D.N.Y. Mar. 25, 2021) (finding exempt under Sierra Club

information that “would reveal the Agencies’ interim thoughts on courses of action that are

contingent and subject to change”).11

Second, the 2018 Version-Formaldehyde Assessment precedes, and “contribute[s]” to,

Morley v. CIA, 508 F.3d 1108, 1126–27 (D.C. Cir. 2007), the eventual final IRIS formaldehyde

assessment, and as such was “generated before the agency’s final decision on the” health effects

of formaldehyde, Sierra Club, 141 S. Ct. at 786. That final formaldehyde assessment will, once

all seven of the IRIS steps are completed with release to the public, represent the EPA’s official

statement on the health risks of formaldehyde, and as such will inform not only other EPA

programs but also the general public. See supra Part I.A. The fact that the 2018 Version-

Formaldehyde Assessment precedes that final assessment—indeed, no such final assessment yet

exists—further confirms that the 2018 Version-Formaldehyde Assessment is predecisional. See

Judicial Watch, Inc. 847 F.3d at 739 (“Documents are ‘predecisional’ if they are ‘generated

before the adoption of an agency policy[]’ . . . .” (quoting Pub. Citizen, Inc. v. Off. of Mgmt. &

Budget, 598 F.3d 865, 874 (D.C. Cir. 2010))). Furthermore, although the final formaldehyde

assessment will affect, inter alia, EPA policymaking, the 2018 Version-Formaldehyde

Assessment itself carries no consequences, either legal or practical, further supporting the

conclusion that the document is predecisional. See Sierra Club, 141 S. Ct. at 787 (emphasizing

“the legal, not practical, consequences that flow from an agency’s action” in assessing finality).

11
Plaintiff argues that the 2018 Version-Formaldehyde Assessment is not “subject to change” within the
meaning of Sierra Club, because, unlike the records reviewed in that case, the formaldehyde assessment would be
subject to change even after release to the public at IRIS Step Four. Pl.’s Suppl. Mem. at 4. Evidently, plaintiff’s
reasoning is that because this document is subject to change after Step Four, the fact that the document is subject to
change before release at Step Four is immaterial to its status as predecisional. See id. Plaintiff offers no support for
this myopic reading of Sierra Club, and nothing in the Court’s opinion in that case suggests that an agency’s
provisional, nonfinal position on a matter loses its status as predecisional simply because a later version is also
nonfinal. See Sierra Club, 141 S. Ct. at 786–88.

20
 Third, the 2018 Version-Formaldehyde Assessment not only is labeled a draft and

subject, pursuant to EPA’s IRIS policy, to internal revision before public release, but also

contains ordinary indicia of a draft document, bolstering the conclusion that this document is a

predecisional, deliberative document protected by the deliberative process privilege.

Specifically, as noted, the 2018 Version-Formaldehyde Assessment contains redlines, indicating

proposed changes, and comment bubbles, reflecting commenters’ opinions on and proposals

concerning specific portions of the draft. See 1st Orme-Zavaleta Decl. ¶ 15; see also Taylor

Energy Co. LLC v. U.S. Dep’t of Interior, 271 F. Supp. 3d 73, 95 (D.D.C. 2017) (holding that

draft versions of agency memorandum that contained “comments and redlines edits” were

properly withheld pursuant to deliberative process privilege). The reasons such redlines and

comments are significant are two-fold. These markings themselves comprise part of EPA’s

deliberative process, reflecting IRIS staff’s tentative ideas, proposals, and recommendations on

the substance of the 2018 Version-Formaldehyde Assessment, and plainly “reflect the personal

opinions of the writer rather than the policy of the agency,” Coastal States, 617 F.2d at 866; see

also 1st Orme-Zavalet Decl. ¶¶ 14–15. Relatedly, these redlines and comments also corroborate

the “administrative context,” Sierra Club, 141 S. Ct. at 786, in which the 2018 Version-

Formaldehyde Assessment is situated, highlighting that EPA treated this document as a work in

progress, not a final statement of agency policy, see id. Indeed, the fact that subsequent draft

versions of the document were created after the creation of the 2018 Version-Formaldehyde

Assessment at issue here, see 1st Orme-Zavaleta Decl. ¶ 14, only underscores that this document

is EPA’s “preliminary view,” rather than its “final decision,” on the health risks of

formaldehyde, Sierra Club, 141 S. Ct. at 786.

21
 Finally, even putting aside the existence of redlines and comments, the 2018 Version-

Formaldehyde Assessment is protected by the deliberative process privilege because the

document comprises the agency’s tentative selection and synthesis of facts and studies upon

which to base its assessment of the health effects of formaldehyde, which are quintessentially

deliberative. “[W]ell-established law in this Circuit [provides] that the deliberative process

privilege operates to shield from disclosure agency decision-making reflecting the collection,

culling and assessment of factual information or . . . data.” Ctr. for Biological Diversity v. EPA,

369 F. Supp. 3d 1, 20 (D.D.C. 2019) (collecting cases); see also Mapother v. Dep't of Justice, 3

F.3d 1533 (D.C. Cir. 1993) (holding that deliberative process privilege protected document

organizing facts relevant to decision to exclude foreign leader from United States); Ancient Coin

Collectors Guild v. U.S. Dep’t of State, 641 F.3d 504, 513 (D.C. Cir. 2011) (holding exempt

factual information “culled . . . from the much larger universe of facts” available because this

“reflect[ed] an exercise of judgment as to what issues are most relevant to the pre-decisional

findings and recommendations” (internal quotation marks omitted)); Goodrich Corp. v. EPA,

593 F. Supp. 2d 184, 189 (D.D.C. 2009) (“[E]ven if the data plugged into the model is itself

purely factual, the selection and calibration of data is part of the deliberative process to which

Exemption 5 applies.”); Reliant Energy Power Generation, Inc. v. FERC, 520 F. Supp. 2d 194,

205–06 (D.D.C. 2007) (finding “spreadsheets and tables that analyze raw data” exempt from

disclosure (alteration and internal quotation marks omitted)). As explained, the 2018 Version-

Formaldehyde Assessment reflects “identification of relevant studies across multiple scientific

disciplines . . . evaluation of study methods; analysis and synthesis of evidence from identified

studies . . . and hazard identification.” 1st Orme-Zavaleta Decl. ¶ 15. Such “selection,

organization, and analysis of factual information” “constitutes an exercise of judgment by EPA

22
staff.” Id. EPA further notes that because the 2018 Version-Formaldehyde Assessment’s

“[i]ntegration of evidence across studies inherently requires scientific judgment and

consideration of the strengths and weaknesses of the available studies,” disclosure “of even the

factual information contained within [the 2018 Version-Formaldehyde Assessment] would

expose the agency’s deliberative process, including exposing the agency’s preliminary

deliberations, thoughts, and analyses.” Id. ¶ 16. The 2018 Version-Formaldehyde Assessment,

which comprises the facts and data provisionally selected and relied upon by EPA, was therefore

properly withheld under Exemption 5.

D. Plaintiff’s Objections Are Unavailing

Notwithstanding that the 2018 Version-Formaldehyde Assessment is a work-in-progress

document preceding EPA’s final IRIS formaldehyde assessment and is, per the IRIS program,

expressly subject to further agency revision, plaintiff advances several arguments that the 2018

Version-Formaldehyde Assessment nevertheless falls outside the protection of the deliberative

process privilege. These contentions are considered in turn, and each is rejected.

1. Plaintiff’s Overreliance on Pruitt’s Senate Testimony

The first and most serious flaw with plaintiff’s position that the 2018 Version-

Formaldehyde Assessment is not protected by the deliberative process privilege is its insistence

that EPA has prepared a Step Four formaldehyde assessment—that is, a formaldehyde

assessment that has undergone Agency Review and Interagency Science Consultation and is now

ready for public release for comment and peer review at Step Four. Many of plaintiff’s

arguments are directed to whether that hypothetical document is protected by the deliberative

process privilege. See, e.g., Pl.’s Opp’n at 21–22, 31–33.

Such arguments are simply beside the point. To begin with, as already explained,

plaintiff has voluntarily narrowed the scope of the instant lawsuit to the single issue of whether
23
EPA properly withheld the 2018 Version-Formaldehyde Assessment pursuant to the deliberative

process privilege. See supra Part I.E; JSR ¶ 4. Thus, arguments about whether some other,

hypothetical document was properly withheld fall outside the agreed-upon scope of the instant

lawsuit. In an apparent effort to bypass that agreed-upon scope, plaintiff argues that its FOIA

request “did not seek one specific unfinished version of the Draft IRIS [Assessment], but any

version, with a specific aim towards the version that EPA reported years ago was in a final state

and ready for public scrutiny.” Pl.’s Opp’n at 18 (emphasis in original). That may well be an

accurate description of what plaintiff originally sought with its FOIA request, but disregards the

parties’ agreed-upon scope to focus the instant lawsuit on the 2018 Version-Formaldehyde

Assessment, and therefore misstates the issue before the Court.

Moreover, the record suggests that the hypothetical document regarding which plaintiff

spills much ink—again, an IRIS Step Four formaldehyde assessment ready to be released for

public comment—simply does not exist. According to the EPA, via declarations by the

highest-ranking career EPA employee responsible for overseeing the IRIS formaldehyde

assessment, and attested to under penalty of perjury, the IRIS draft formaldehyde assessment

was, after 2011, returned to IRIS Step One, never completed Steps Two and Three, and thus

never reached Step Four, such that a “clean” version ready for public comment and external peer

review has not been created. See 1st Orme-Zavaleta Decl. ¶ 9 (“[The 2018 Version-

Formaldehyde Assessment] has only proceeded through the first step of the IRIS process, Draft

Development.”); 2nd Orme-Zavaleta Decl. ¶¶ 3–4 (“The 2018 Version-Formaldehyde

Assessment has only proceeded through the draft development phase of the IRIS process. The

Agency review step (Step 2) was never initiated for the [2018 Version-Formaldehyde

Assessment], nor has the [2018 Version-Formaldehyde Assessment] gone through the

24
Interagency review step (Step 3).”); see also Def.’s Reply at 14 (affirming that “clean” or “Step

4” version of the formaldehyde assessment, other than the Step Four assessment released in

2010, “does not exist”).

Indeed, plaintiff concedes that EPA may not have produced the IRIS Step Four

formaldehyde assessment sought in the FOIA request. See Pl.’s Opp’n at 18 (acknowledging the

possibility that “no Step 4 Draft IRIS report ever existed”). Nevertheless, plaintiff presses the

argument that such a version of the formaldehyde assessment does exist—suggesting, by

implication, that EPA is dissembling before the Court in affirming otherwise—in reliance on

then-Administrator Pruitt’s January 2018 testimony before the Senate Committee on the

Environment and Public Works. As plaintiff characterizes this testimony, Pruitt “testified to the

U.S. Senate that he understood that ‘the EPA ha[d] finalized its conclusion that formaldehyde

causes leukemia and other cancers, and that that completed new assessment [was] ready to be

released for public review.’” Pl.’s Opp’n at 3 (alterations in original) (quoting Senate Hr’g at

278).

This is, at best, an incomplete description of Pruitt’s testimony. As described, see supra

Part I.D, Senator Markey, not Pruitt, stated during the hearing that “it is [his] understanding that

the EPA has finalized its conclusion that formaldehyde causes leukemia and other cancers, and

that that completed new assessment is ready to be released for public review[,] [b]ut [the

assessment] is still being held up.” Def.’s Reply at 4 (quoting Senate Hr’g at 278). In response

to that statement, and to Markey’s question whether Pruitt could provide an update concerning

EPA’s handling of the IRIS formaldehyde assessment, Pruitt stated that “[his] understanding is

similar to [Senator Markey’s], but [he would] confirm that and provide the information to

[Senator Markey]. Id. (quoting Senate Hr’g at 278). Thus, Senator Markey’s statement of the

25
situation contained three subsidiary claims: (1) EPA had finalized its conclusion that

formaldehyde causes cancer; (2) the assessment is ready to be released for public review; and (3)

the assessment was nevertheless being “held up.” Which of those three claims, or subset thereof,

to which Pruitt was agreeing when he stated that his understanding was “similar” to Senator

Markey’s is unclear.

More to the point, even accepting, arguendo, plaintiff’s characterization of this statement

as confirming that an IRIS Step Four formaldehyde assessment existed and was being held back

from public release, this brief testimony is too slender a reed to support plaintiff’s desired result

that such facts are established. This brief testimony by an agency head responsible for the

administration of an entire executive branch agency and all of its subsidiary offices, programs,

and initiatives, about the status of a particular report within a specific agency program, made in

the context of responding to a query at a Senate hearing—and couched with the caveat that the

information must be checked―is simply less credible than, and must give way to, EPA’s

affidavits in this case. Those affidavits, provided by the EPA official responsible for directing

the agency component overseeing the IRIS program, make clear that Pruitt’s understanding, as

stated at that hearing, was incorrect, and no such version of the formaldehyde assessment, ready

for public release, existed. Furthermore, Pruitt’s agreement with Senator Markey’s

understanding was tentative, and he expressly stated that he would need to confirm the status of

the formaldehyde assessment and would follow up with Markey’s office as needed. Follow-up

communications from EPA to Senator Markey did not confirm Pruitt’s testimony and instead

stated that EPA was still “working to fully implement the NAS recommendations in all IRIS

assessments released moving forward.” Pl.’s Reply at 4.12

12
As additional support for the claim that a new IRIS Step Four formaldehyde assessment exists, without
public release, plaintiff points to EPA’s own website, which, according to plaintiff, indicates that the IRIS

26
 Finally, plaintiff contends that EPA’s declarant herself confirmed, in congressional

testimony on March 27, 2019, that EPA has a new IRIS Step Four formaldehyde assessment not

yet released to the public. See Pl.’s Reply at 4–5 (citing House Science, Space, & Tech. Comm.,

Hearing: EPA’s IRIS Program: Reviewing its Progress and Roadblocks Ahead (“Hr’g Record”),

YOUTUBE (Mar. 27, 2019), https://www.youtube.com/watch?v=RwGG587A67U). Specifically,

in response to a question about the status of the IRIS formaldehyde assessment, EPA’s declarant,

Dr. Jennifer Orme-Zavaleta, stated, “[W]e do have a draft formaldehyde assessment,” and noted

that the draft would serve as a starting point for a separate EPA office’s work on a formaldehyde

risk evaluation under the Toxic Substances Control Act. Pl.s’ Reply at 5 (quoting Hr’g Record at

1:17:17). Plaintiff argues that, because this statement “does not refer to . . . a series of

increasingly marked up documents, or a jumble of redline comments,” the draft formaldehyde

assessment referenced by Dr. Orme-Zavaleta’s must be “complete enough to be considered the

definitive draft.” Id. This inference is unsupportable, and Dr. Orme-Zavaleta’s testimony, which

clearly referred to a draft formaldehyde assessment, is entirely consistent with her declarations in

the instant case that a Step Four formaldehyde assessment does not exist.

In sum, then, even putting aside the parties’ agreed-upon scope limitation in their

cross-motions for summary judgment to whether the deliberative process privilege was properly

applied to the 2018 Version-Formaldehyde Assessment, plaintiff’s various arguments that EPA

possesses an IRIS Step Four formaldehyde assessment that is ready for public release are

formaldehyde assessment is at Step Four of the IRIS process. See Pl.’s Opp’n at 9, 11; see also Pl.’s Reply at 4.
EPA’s website does refer to an IRIS Step Four formaldehyde assessment, but this reference is to the 2010 Step Four
formaldehyde assessment, which EPA moved back to Step One following NAS’s peer-review comments. See 2d
Orme-Zavaleta Decl. ¶ 5 (“That website refers to the 2010 Assessment, not the [c]urrent Draft Assessment . . . [T]he
2010 Assessment reached Step 4 of the IRIS process and was released for public comment and external peer
review[,] [and] . . . remains publicly available on EPA’s website today.”); see also supra Part I.B. Thus, plaintiff’s
reliance on EPA’s website as proof of a new Step Four formaldehyde assessment is simply wrong.

27
unpersuasive. Not for the first time, plaintiff’s factual contentions “totter between the trivial and

the speculative” and “do not undermine the reliability of the agency’s affidavits.” Pub. Emps.

for Envt’l Resp. v. U.S. Section, Int’l Boundary & Water Comm’n, 740 F.3d 195, 201 (D.C. Cir.

2014) (Kavanaugh, J.).

2. Plaintiff’s Remaining Arguments Against Withholding the 2018 Version-
Formaldehyde Assessment

Much of plaintiff’s briefing does not relate to the 2018 Version-Formaldehyde

Assessment, which is the withheld document that is now the agreed-upon focus of this lawsuit,

but instead on the speculative, nonexistent Step 4 Formaldehyde Assessment. See, e.g., Pl.’s

Opp’n at 21–22 (arguing “that the Step 4 Formaldehyde Assessment is not predecisional and

therefore should be released”); id. at 31–33 (arguing that hypothetical IRIS Step Four

formaldehyde assessment is final statement of agency policy and therefore not deliberative). To

the extent plaintiff advances arguments applicable to the 2018 Version-Formaldehyde

Assessment, those arguments are rejected.

First, plaintiff argues that the 2018 Version-Formaldehyde Assessment is not

deliberative, because it consists of “purely factual materials.” Pl.’s Opp’n at 17; see also id. at

24 (claiming that “IRIS Assessments are purely factual reports.”); Pl.’s Suppl. Mem. at 6–7. To

be sure, the deliberative process privilege does not protect “[u]nevaluated factual reports or

summaries,” Pl.’s Opp’n at 16 (emphasis added) (quoting Vaugh v. Rosen, 523 F.2d 1136, 1143

(D.C. Cir. 1975)), but, as noted, “the deliberative process privilege operates to shield from

disclosure agency decision-making reflecting the collection, culling and assessment of factual

information or . . . data.” Ctr. for Biological Diversity v. EPA, 369 F. Supp. 3d at 20 (collecting

cases). As explained, see supra Part III.A.3, the fact-selection at issue here plainly falls into the

second category. The determination of which facts, and supporting studies, to rely upon in the

28
formaldehyde assessment is the central deliberative question that the IRIS staff must resolve in

drafting the assessment. See 1st Orme-Zavaleta Decl. ¶ 15. Indeed, IRIS assessments are

important to the general public in part because they reflect EPA’s expert evaluation and synthesis

of the existing body of scientific study on a given chemical. As such, the 2018 Version-

Formaldehyde Assessment involves the sort of substantive “culling” of factual materials, Ctr. for

Biological Diversity, 369 F. Supp. 3d at 20, that qualifies it for protection under the deliberative

process privilege.

Plaintiff relies on an out-of-circuit district court decision holding that the National

Transportation Safety Board’s “selection of . . . data culled from hundreds of pages of data” was

not protected by the deliberative process privilege, Pl.’s Opp’n at 27 (quoting Lahr v. NTSB, 453

F. Supp. 2d 1153, 1187 (C.D. Cal. 2006)); see also Pl.’s Reply at 13–14, but Lahr is readily

distinguishable. In that case, the agency did no more to justify its claim that the data-selection

was deliberative than to “cursorily stat[e] that ‘without the protection provided by the exemption,

full and frank discussion of options and opinions so vital to the decision-makers would be

impossible,’” and did not provide “an explanation or description of the communicative or

evaluative procedures the [agency] followed in doing its ‘culling.’” Lahr, 453 F. Supp. 2d at

1187. Here, in contrast, EPA has explained in adequate detail the deliberative process

underlying the IRIS formaldehyde assessment, namely the need for the exercise of scientific

evaluation of extant formaldehyde studies to evaluate their methodological reliability, analysis of

the evidence and findings of sound studies, and the synthesis and integration of such evidence

and studies to reach an ultimate determination on the health effects of formaldehyde. See 1st

Orme-Zavaleta Decl. ¶¶ 13–15; see also supra Part III.B.

29
 Second, plaintiff argues that the 2018 Version-Formaldehyde Assessment is not

predecisional, because “there was no ultimate policy decision, or even a process of formulating a

policy,” to which the 2018 Version-Formaldehyde Assessment contributed and was antecedent.

Pl.’s Opp’n at 20 (emphasis in original); Pl.’s Reply at 14–17; see also, e.g., Sierra Club, 141 S.

Ct. at 786 (“Documents are ‘predecisional’ if they were generated before the agency’s final

decision on the matter . . . .” (quoting Sears, 421 U.S. at 150–152)); Petroleum Info Corp. v.

Dep’t of Interior, 976 F.2d 1429, 1435 (D.C. Cir. 1992) (“To fall within the deliberative process

privilege, materials must bear on the formulation or exercise of agency policy-oriented

judgment.” (emphasis in original)). This argument construes the scope of the deliberative

materials encompassed by Exemption 5 too narrowly. The D.C. Circuit has long held that

Exemption 5 “covers recommendations, draft documents, proposals, suggestions, and other

subjective documents which reflect the personal opinions of the writer rather than the policy of

the agency,” Coastal States, 617 F.2d at 866; see also Pub. Citizen, Inc., 598 F.3d at 875, and the

universe of covered agency decisions extends not only to official agency policies but also to

agency decisions more generally, see, e.g., Sierra Club, 141 S. Ct. at 786 (noting that

deliberative process privilege covers an agency’s “decisions” and “position[s]” in addition to

more formal “policies”). The eventual final IRIS formaldehyde assessment is such an agency

position to which the 2018 Version-Formaldehyde Assessment contributes. In other words, the

fact that the aim of the formaldehyde assessment is to provide information about the health risks

of formaldehyde, rather than to, for example, substantively regulate formaldehyde use, does not

remove the 2018 Version-Formaldehyde Assessment from the scope of the deliberative process

privilege, as the public release of an official agency statement of the health hazards of

formaldehyde is an agency decision covered by Exemption 5. See, e.g., Husch Blackwell LLP v.

30
EPA, 442 F. Supp. 3d 114, 122–23 (D.D.C. 2020) (holding “materials preparing officials for

congressional testimony and draft responses to Congress” exempt from disclosure under the

deliberative process privilege (internal quotation marks omitted)); Urban Air Initiative, Inc. v.

EPA, 271 F. Supp. 3d 241, 261 (D.D.C. 2017) (finding that agency decisions about study

preliminary to creation of “an updated emissions model” were “exactly the type of agency

judgments that the deliberative process privilege protects”).

Plaintiff further argues that, in assessing whether the 2018 Version-Formaldehyde

Assessment is predecisional, “the relevant decision is not the issuance of the final IRIS

Formaldehyde Assessment, but the publication of the public review draft at Step 4, because at

Step 4 the document would have become public regardless.” Pl.’s Opp’n at 20. Even accepting

plaintiff’s position that the relevant agency decision that the 2018 Version-Formaldehyde

Assessment precedes and contributes to is the eventual IRIS Step Four formaldehyde assessment,

rather than the final, IRIS Step Seven formaldehyde assessment, does not change the above

analysis. Put another way, a Step Four assessment ready for public release and peer review is an

agency decision for which antecedent deliberations are protected by the deliberative process

privilege, and the work product generated in the development of such an agency decision is

therefore predecisional, regardless of the fact that such a Step Four assessment would undergo

further revision before becoming the final formaldehyde assessment. A Step Four assessment is

still the agency’s final decision with respect to which provisional conclusions to present to the

public for feedback. See, e.g., Ecological Rights Found., 2021 WL 535725, at *15 (“[A]n

agency’s consideration of what information to present to external parties and how to present it is

a [protected agency] decision in itself[.]”); Am. Ctr. for Law & Justice v. U.S. Dep’t of Justice,

325 F. Supp. 3d 162, 172 (D.D.C. 2018) (“[I]f agency deliberations about public statements were

31
FOIA-able, then agencies would be hamstrung . . . .”). Accordingly, regardless of whether the

agency decision toward which the 2018 Version-Formaldehyde Assessment contributes is the

eventual Step Four assessment or the eventual final, Step Seven assessment, the 2018 Version-

Formaldehyde Assessment is predecisional.

E. Foreseeable Harm

Plaintiff next argues that even if the 2018 Version-Formaldehyde Assessment is

predecisional and deliberative, it is not protected by the deliberative process privilege because no

harm would result from its disclosure. The FOIA Improvement Act provides that “[a]n agency

shall withhold information . . . only if the agency reasonably foresees that disclosure would harm

an interest protected by” one of the nine FOIA exemptions. 5 U.S.C. § 552(a)(8)(A). This

provision requires agencies withholding information under an exemption to show not only that a

withheld record “falls within a FOIA exemption,” but also “that the agency ‘reasonably foresees

that disclosure would harm an interest protected by [the] exemption.’” Machado Amadis, 971

F.3d at 370 (alteration in original) (quoting 5 U.S.C. § 552(a)(8)(A)(i)(I)).

An agency successfully makes this second, “heightened” showing, Judicial Watch, Inc. v.

U.S. Dep’t of Com., 375 F. Supp. 3d 93, 100 (D.D.C. 2019), by “‘identify[ing] specific harms to

the relevant protected interests that it can reasonably foresee would actually ensue from

disclosure of the withheld materials’ and ‘connect[ing] the harms in [a] meaningful way to the

information withheld,’” Ctr. for Investigative Reporting, 436 F. Supp. 3d at 106 (third alteration

in original) (quoting Judicial Watch, Inc. v. U.S. Dep’t of Justice (“Judicial Watch II”), Civ. A.

No. 17-0832 (CKK), 2019 WL 4644029, at *5 (D.D.C. Sept. 24, 2019)); see also H.R. Rep. No.

114-391, at 9 (2016) (“An inquiry into whether an agency has reasonably foreseen a specific,

identifiable harm that would be caused by a disclosure would require the ability to articulate both

the nature of the harm and the link between the specified harm and specific information
32
contained in the material withheld.”). Agencies therefore “must provide more than ‘nearly

identical boilerplate statements’ and ‘generic and nebulous articulations of harm.’” Ctr. for

Investigative Reporting, 436 F. Supp. 3d at 106 (quoting Judicial Watch II, 2019 WL 4644029,

at *4–5).

“[T]he agency’s burden to demonstrate that harm would result from disclosure may shift

depending on the nature of the interests protected by the specific exemption with respect to

which a claim of foreseeable harm is made.” Ecological Rights Found., 2021 WL 535725, at

*32 (citing Rosenberg v. Dep’t of Def., 442 F. Supp. 3d 240, 259 (D.D.C. 2020); S. Rep. No.

114-4, at 8 (2015)). To demonstrate foreseeable harm with respect to exemptions under the

deliberative process privilege, “[t]he agency ‘cannot simply rely on generalized assertions that

disclosure could chill deliberations,’” but instead “must ‘provide context or insight into the

specific decision-making processes or deliberations at issue, and how they in particular would be

harmed by disclosure.’” Id. (first quoting Machado Amadis, 971 F.3d at 371; and then quoting

Ctr. for Investigative Reporting, 436 F. Supp. 3d at 107).

The D.C. Circuit in Machado Amadis considered the adequacy of an agency’s foreseeable

harm showing under the deliberative process privilege. The agency in that case produced, in

response to the plaintiff’s FOIA request, a series of “Blitz Forms,” documents used to adjudicate

FOIA appeals, with redactions under the deliberative process privilege. 971 F.3d at 369–71. In

support of the redactions, the agency’s affidavit stated that the withheld materials revealed “line

attorneys’ evaluations, recommendations, discussions, and analysis which are prepared for

senior-level review and decisionmaking,” id. at 370 (internal quotation marks omitted), and

asserted that disclosure of this information “would discourage line attorneys from candidly

discussing their ideas, strategies, and recommendations, thus impairing the forthright internal

33
discussions necessary for efficient and proper adjudication of administrative appeals,” id. at 371

(alteration and internal quotation marks omitted). The D.C. Circuit deemed this showing of

foreseeable harm sufficient, without any further factual proffer by the agency, because the

agency “specifically focused on the information at issue” and properly “concluded that

disclosure of that information would chill future internal discussions.” Id. (internal quotation

marks omitted); see also Ecological Rights Found., 2021 WL 535725, at *32.

EPA’s declarations in this case make a similarly adequate showing. As in Machado

Amadis, EPA has identified particular harms that are linked to “specific information contained in

the material withheld,” Judicial Watch II, 2019 WL 4644029, at *4 (internal quotation marks and

citation omitted). First, EPA explains that disclosure of the 2018 Version-Formaldehyde

Assessment would chill “frank analysis and evaluations” because “[i]f the staff working on this

Assessment or a similar assessment knew that all of their edits and comments would someday be

released . . . they would be less likely to freely discuss . . . scientific analysis.” Def.’s Reply at

16–17. This is not a hypothetical or speculative harm, since, as noted, fifteen other IRIS

assessments are currently in various stages of drafting. 1st Orme-Zavaleta Decl. ¶ 7. Thus,

release of the 2018 Version-Formaldehyde Assessment, and the accompanying realization by

IRIS staff that their internal deliberations and draft work product are subject to disclosure under

FOIA, would likely chill the agency’s work on those fifteen other chemical assessments.

Second, EPA explains that release of the 2018 Version-Formaldehyde Assessment would cause

“public confusion” because it may reveal “particular positions that may not represent the

agency’s decisionmakers’ final conclusions of the health effects of formaldehyde,” and would

disclose “staff opinions, which may or may not ripen into the final conclusions” of the agency.

34
1st Orme-Zavaleta Decl. ¶ 18.13 The agency thus affirmatively concludes that disclosure of the

2018 Version-Formaldehyde Assessment would harm an interested protected by the deliberative

process privilege. See Sierra Club, 141 S. Ct. at 785; see also Machado Amadis v. Dep’t of

Justice, 388 F. Supp. 3d 1, 18–19 (D.D.C. 2019) (citing Jordan, 591 F.2d at 772) (summarizing

the relevant interests of the privilege).

Plaintiff advances three arguments in response. First, plaintiff contends that no harm

would result from disclosure of the 2018 Version-Formaldehyde Assessment because the

document has already been selectively shared outside the agency with the ACC at the January

2018 meeting with IRIS staff. Pl.’s Opp’n at 23. Thus, according to plaintiff, any harm from

disclosure has already occurred, and the disclosure of the 2018 Version-Formaldehyde

Assessment in response to plaintiff’s FOIA request would result in no additional harm. This

argument is easily dispatched, as it rests on a misapprehension of the substance of the January

2018 meeting between IRIS staff and the ACC. As noted, although IRIS staff discussed with the

ACC its general methodological approach to the formaldehyde assessment, in light of the NAS’s

peer-review comments, the actual draft formaldehyde assessment—and, by extension, the

agency’s specific deliberations revealed therein—was not shared with the ACC. 2d Orme-

Zavaleta Decl. ¶ 9.

Second, plaintiff argues that EPA’s caution about the confusion caused by the release of

the 2018 Version-Formaldehyde Assessment is not a cognizable harm in this case because the

13
Plaintiff implicitly acknowledges this possibility for confusion, and the fact that the 2018 Version-
Formaldehyde Assessment will likely differ in substance from the IRIS formaldehyde assessment ultimately
released at Step Four, when stating that it seeks the 2018 Version-Formaldehyde Assessment because “after a new
[Step Four] assessment is released, it would be useful to compare it to the agency’s previous conclusions,” namely
those stated in the 2018 Version-Formaldehyde Assessment. Pl.’s Suppl. Mem. at 14. This precise scenario, in
which two substantively different versions of an agency document are publicly released, creating uncertainty over
which is the agency’s official statement, is a specific harm that the deliberative process privilege is intended to
ameliorate. See, e.g., Machado Amadis, 388 F. Supp. 3d at 18.

35
public is already “[t]erminally [c]onfused” about the state of the IRIS formaldehyde assessment.

Pl.’s Opp’n at 39–40. This argument falls far short of being persuasive. Any extant public

confusion does nothing to diminish the fact that the release of the 2018 Version-Formaldehyde

Assessment would likely only add to the confusion about the agency’s positions. In any event,

plaintiff equivocates over the relevant public confusion that EPA cites as a harm that would

result from releasing the 2018 Version-Formaldehyde Assessment. The agency’s concern is that

release of the 2018 Version-Formaldehyde Assessment would generate public confusion over

which document represents the agency’s Step Four IRIS assessment that has gone through

Agency Review and Interagency Science Consultation, and is therefore ready for public

comment and outside peer review. See Def.’s Reply at 18. Plaintiff, in contrast, posits that the

public is “already confused about why the completed [formaldehyde] Assessment was

abandoned.” Pl.’s Opp’n at 39. The circumstances of the delay of the IRIS Step Four

formaldehyde assessment are entirely distinct from the eventual substance of that Step Four

assessment, and thus any public confusion about the former in no way undercuts EPA’s

explanation that release of the 2018 Version-Formaldehyde Assessment would generate

confusion about the latter.

Third, citing the ACC letter’s characterization that EPA was simply “restructuring” the

2010 draft formaldehyde assessment, rather than “revisit[ing] the science,” plaintiff argues that

any revelation of the agency’s deliberative process arising from disclosure of the 2018 Version-

Formaldehyde Assessment’s release has, in fact, already occurred, with the release of the 2010

Step Four assessment. Pl.’s Opp’n at 24. This argument rests on a faulty factual premise. As

already noted, according to EPA, ACC’s characterization of the revisions being undertaken since

2011—namely that the assessment is merely being restructured rather than substantively

36
revised—“is incorrect,” and, to the contrary, although “the [2018 Version-Formaldehyde

Assessment] does present information and conclusions in a new structure compared to the

[p]revious [2010] [a]ssessment, it also incorporates new studies and updated analyses to reflect

the latest science on formaldehyde toxicity.” 2d Orme-Zavaleta Decl. ¶ 10; see also supra note

6.

In sum, then, the agency has identified two harms that the deliberative process privilege

is meant to prevent, chilling of agency deliberations and public confusion, see Machado Amadis,

388 F. Supp. 3d at 18–19 (summarizing the relevant interests of the privilege), and has

adequately linked those harms to the disclosure of the particular information contained in the

2018 Version-Formaldehyde Assessment. EPA has therefore satisfied the foreseeable harm

requirement.

F. Segregability

FOIA requires that “[a]ny reasonably segregable portion of a record shall be provided to

any person requesting such record after deletion of the portions which are exempt” from

disclosure. 5 U.S.C. § 552(b). Producing segregable information is essential for agencies’ FOIA

compliance, and “district courts cannot approve withholding exempt documents ‘without making

an express finding on segregability.’” Machado Amadis, 971 F.3d at 371 (quoting Morley, 508

F.3d 1108 at 1123); see also, e.g., Porup v. CIA, No. 20-5144, 2021 WL 2021615, at *11 (D.C.

Cir. May 21, 2021) (“[A] trial court must make a segregability finding if a federal agency has

redacted or withheld documents pursuant to FOIA exemptions.”); Stolt-Nielsen Transp. Grp. Ltd.

v. United States, 534 F.3d 728, 734 (D.C. Cir. 2008) (“[B]efore approving the application of a

FOIA exemption, the district court must make specific findings of segregability regarding the

documents to be withheld.” (internal quotation marks and citation omitted)); Sussman v. U.S.

Marshals Serv., 494 F.3d 1106, 1116 (D.C. Cir. 2007) (same).
37
 In evaluating segregability, “[a]gencies are entitled to a presumption that they complied

with the obligation to disclose reasonably segregable material.” Sussman, 494 F.3d at 1117.

Even under that presumption, “the agency must provide a ‘detailed justification’ for [the exempt

material’s] non-segregability,” but need not “provide so much detail that the exempt material

would be effectively disclosed.” Johnson v. Exec. Off. for U.S. Att’ys, 310 F.3d 771, 776 (D.C.

Cir. 2002) (quoting Mead Data Ctr., Inc. v. Dep’t of Air Force, 566 F.2d 242, 261 (D.C. Cir.

1977)). Affidavits attesting to the agency’s “line-by-line review of each document withheld in

full” and the agency’s determination “that no documents contained releasable information which

could be reasonably segregated from the nonreleasable portions,” in conjunction with a

description of the withheld record, suffice. Id. (internal quotation marks omitted); see also

Loving v. Dep’t of Def., 550 F.3d 32, 41 (D.C. Cir. 2008).

Contrary to plaintiff’s claim that EPA “made no attempt to segregate any purely factual

information, or to specify why it could not,” Pl.’s Opp’n at 41, EPA has made the necessary

segregability showing here. EPA explains that “[f]actual information contained in the [2018

Version-Formaldehyde Assessment] cannot be reasonably segregated from the deliberative

content,” since “any purely factual information contained within the [2018

Version-Formaldehyde Assessment] is so thoroughly integrated with agency deliberations that its

disclosure would reveal those deliberations.” 1st Orme-Zavaleta Decl. ¶ 16; see also Bell Decl.

¶ 17 (plaintiff’s declarant stating that he was informed by EPA, in December 2019, that agency

was assessing segregability of the 2018 Version-Formaldehyde Assessment and ultimately

determined, in January 2020, that the entire document was exempt from disclosure under

Exemption 5). More specifically, EPA notes that because the 2018 Version-Formaldehyde

Assessment’s “[i]ntegration of evidence across studies inherently requires scientific judgment

38
and consideration of the strengths and weaknesses of the available studies,” disclosure “of even

the factual information contained within the [2018 Version-Formaldehyde Assessment] would

expose the agency’s deliberative process, including exposing the agency’s preliminary

deliberations, thoughts, and analyses.” 1st Orme-Zavaleta Decl. ¶ 16. Such “inextricably

intertwined” language of similar specificity to that expressed by EPA is routinely found to be

sufficient for segregability purposes. See, e.g., Elec. Frontier Found. v. Dep’t of Justice, 739

F.3d 1, 12–13 (D.C. Cir. 2014); Ecological Rights Found., 2021 WL 535725, at *34 (relying on

similar “inextricably intertwined” language in agency’s declarations); Ctr. for Biological

Diversity, 369 F. Supp. 3d at 26 (same).

Nonetheless, plaintiff suggests that EPA can redact redline edits and comment bubbles

from the 2018 Version-Formaldehyde Assessment, and that “segregation of those changes and

comments would be relatively simple.” Pl.’s Opp’n at 16, 23, 31. In plaintiff’s view, apparently,

EPA could simply withhold those features of the document and produce the remainder of the

2018 Version-Formaldehyde Assessment as nonexempt and unprotected by the deliberative

process privilege. This argument erroneously presupposes that the only portion of the [2018

Version-Formaldehyde Assessment] exempt from disclosure is the redlines and comments. As

explained, see supra Part III.A.3, the entire 2018 Version-Formaldehyde Assessment, not just the

redline edits and comment bubbles, is protected by the deliberative process privilege, because the

entire document reflects agency judgment, deliberation, and decisionmaking with respect to

which facts and studies to feature in the IRIS formaldehyde assessment.

IV. CONCLUSION

For the foregoing reasons, EPA’s Motion for Summary Judgment is granted and

plaintiff’s Cross-Motion for Summary Judgment is denied. EPA has justified its withholding in

39
full of the 2018 Version-Formaldehyde Assessment under Exemption 5’s deliberative process

privilege, shown that foreseeable harm would result from further disclosures, and complied with

its segregability obligations.

An order consistent with this Memorandum Opinion will be entered contemporaneously.

Date: June 18, 2021
__________________________
BERYL A. HOWELL
Chief Judge

40