DOE et al v. TRUMP et al

Filing 237

MEMORANDUM OPINION. Signed by Judge Colleen Kollar-Kotelly on 9-13-2019. (lcckk3)

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA
JANE DOE 2, et al.,
Plaintiffs
Civil Action No. 17-1597 (CKK)

v.
MARK T. ESPER, et al.,
Defendants

MEMORANDUM OPINION AND ORDER
(September 13, 2019)
The Court has considered Plaintiffs’ [213-1] Motion to Compel Production of Documents
Improperly Withheld by Defendants.1 The parties present a myriad of discovery disputes ranging
from broad issues, such as whether or not certain topics are relevant to this lawsuit, to specific
issues, such as whether or not certain entries in Defendants’ Vaughn index are sufficiently
detailed. In an effort to narrow or reframe the issues requiring the Court’s resolution, the Court
will now resolve many of the overarching disagreements between the parties.
In summary, the Court has reached the following four conclusions. First, the Court rejects
Defendants’ argument that the level of deference to be applied to the Mattis Plan is already
conclusively established. On the current record, the level of deference to be applied to the Mattis

1

The Court’s consideration has focused on the following documents:
• Pls.’ Mot. to Compel Production of Documents Improperly Withheld by Defs., ECF No.
213-1 (“Pls.’ Mot.”);
• Defs.’ Res. to Pls.’ Mot. to Compel, ECF No. 218 (“Defs.’ Res.”); and
• Pls.’ Reply Mem. of Law in Support of their Mot. to Compel Production of Documents
Improperly Withheld by Defs., ECF No. 222 (“Pls.’ Reply”).

In an exercise of its discretion, the Court finds that holding oral argument in this action would
not be of assistance in rendering a decision at this time. However, after the parties further narrow
and reframe any continuing disputes based on the guidance in this Memorandum Opinion and
Order, oral argument may be required. See LCvR 7(f).

1

Plan remains unclear. Additional discovery is needed to determine if the Plan is the product of
considered military decision-making that reasonably and evenhandedly regulates the matter at
issue. Second, the Court concludes that the relevant time-period for discovery is the development
of the Mattis Plan. On the current record, Plaintiffs have not established a sufficient connection
between the development of the Mattis Plan and the earlier delay of the Carter Policy, the
President’s 2017 tweet, and the 2017 Presidential Memorandum to justify further discovery into
those earlier events. Third, the Court concludes that Defendants may not assert the deliberative
process privilege over documents that were used or considered in the Panel of Experts for the
Transgender Policy Review’s (the “Panel”) development of the Mattis Plan as those documents
go to the heart of Defendants’ intent and decision-making process. The Court further finds that,
even if the deliberative process privilege were to apply to this subcategory of documents,
Plaintiffs need for the information overcomes Defendants’ privilege. The Court acknowledges
that this conclusion will likely still require the Court to determine whether or not certain,
individual documents fall into this subcategory of documents which is not protected by the
deliberative process privilege. Finally, the Court concludes that the parties need to conduct
additional negotiations to determine whether or not Plaintiffs’ narrowed requests for the raw
data, personnel files, and field reports supporting the statistical summaries and conclusions of the
Mattis Report remain overly burdensome.
In resolving these overarching disagreements, Plaintiffs will be able to narrow their
discovery requests and Defendants will be able to respond to these narrowed requests with the
knowledge of what matters are subject to discovery in this case, either producing documents or
formulating refined objections. The parties should meet and confer to determine how they can
most efficiently proceed using the Court’s guidance contained herein. The Court notes that in

2

complying with the Court’s findings and conclusions regarding the nature and scope of discovery
in this case, the parties are not withdrawing their arguments and/or objections; all of which are
preserved for the record. The parties shall file a Status Report by OCTOBER 25, 2019,
indicating how they intend to proceed given the findings and conclusions in this Memorandum
Opinion.
I.

Level of Deference to be Applied to the Mattis Plan
The first overarching issue that the Court will address is the level of deference to be

applied to the Mattis Plan in this case. Defendants contend that “this Court must apply military
deference in reviewing the [Mattis Plan] and that discovery about the process behind the
development of the new policy is irrelevant to determining whether military deference applies.”
Defs.’ Res., ECF No. 218, 12 (emphasis in original). The Court disagrees.
Defendants argue that the D.C. Circuit’s judgment and concurring opinions reversing this
Court’s denial of Defendants’ motion to dissolve the preliminary injunction conclusively
establish that military deference is owed in this case. See generally Doe v. Shanahan, 755 F.
App’x 19 (D.C. Cir. 2019) (judgment); Doe v. Shanahan, 917 F.3d 694 (D.C. Cir. 2019)
(concurring opinions). Defendants quote language from the D.C. Circuit’s judgment stating that
“in a constitutional challenge ‘to decisions by the executive and legislative branches regarding
the composition and internal administration of combat-ready military forces[,]’ ‘courts must give
great deference to the professional judgment of military authorities.’” Defs.’ Res., ECF No. 218,
11 (quoting Doe, 755 F. App’x at 24). Defendants further explain that, because the Mattis Plan
involves issues of military readiness, the Court must be careful to avoid substituting its own
judgment for the military’s reasoned judgment. Defendants’ argument is particularly reliant on
the concurring opinion of Judge Williams. Defendants cite to Judge Williams’s rejection of

3

“‘plaintiffs’ contention, accepted by the district court, that deference to military decisionmaking
… depends on the actual exercise of independent military judgment.’” Defs.’ Res., ECF No. 218,
13 (quoting Doe, 917 F.3d at 729 (Williams, J., concurring)). Instead, Judge Williams explained
that “‘the Constitution itself requires deference to the military choices of the political branches.’”
Id. (quoting Doe, 917 F.3d at 730 (Williams, J., concurring)).
Under Defendants’ theory, deference is owed to the Mattis Plan regardless of the
decision-making process in which Defendants engaged. Because the decision-making process is
irrelevant to the Court’s standard of review, Defendants argue that discovery into the decisionmaking process is improper. The Court concludes that Defendants have misinterpreted the D.C.
Circuit’s judgment and are overly reliant on the concurring opinion of Judge Williams.
First, the D.C. Circuit’s judgment did not require that military deference be applied in
this case on the limited record before the Court. The D.C. Circuit explained that “[c]ourts ‘must
be particularly careful not to substitute our judgment of what is desirable for that of [the
executive and legislative branches], or our own evaluation of evidence for [their] reasonable
evaluation.’” Doe, 755 F. App’x at 24 (quoting Rostker v. Goldberg, 453 U.S. 57, 65 (1981)
(emphasis added)). Here, the D.C. Circuit does not require blind deference to any decision
labeled as “military.” Instead, the D.C. Circuit instructs courts to defer to the military’s
“reasonable evaluation.” In order to determine whether or not the military’s evaluation was
reasonable, some discovery into Defendants’ decision-making process is required.
In making its ruling, the D.C. Circuit relied heavily on three Supreme Court cases each
granting military deference. But, in all three of these cases, deference was granted after the Court
had already determined that reasoned decision-making had taken place. First, in Rostker, the
Supreme Court concluded that a sex-based draft registration statute, which excluded women

4

from registering, was permissible and should be granted military deference. Rostker, 453 U.S. at
79. In granting deference, the Court found that “Congress carefully evaluated the testimony” that
ran counter to the proposed policy. Id. at 82. The Court cited extensively to the evidence
considered by Congress and repeatedly referred to Congress’ “evaluation of that evidence.” Id. at
83. Next, in Goldman v. Weinberger, 475 U.S. 503 (1986), the Court rejected a First Amendment
challenge to Air Force uniform regulations which prohibited the plaintiff from wearing a
yarmulke. 475 U.S. at 509. Again, in upholding the military dress policy, the Court noted that it
must give deference “to the professional judgment of military authorities.” Id. at 507. Deference
was not granted automatically. Instead, the Court emphasized the “considered professional
judgment” which contributed to a 190-page document describing in minute detail requirements
for military dress. Id. at 508. Finally, in Trump v. Hawaii, 138 S.Ct. 2392 (2018), the Court
upheld an executive order placing entry restrictions on nationals from eight foreign states. 138
S.Ct. at 2403-05. In granting the President deference on matters of national security, the Court
emphasized the comprehensive nature of the world-wide study which led the President to place
restrictions on those specific foreign states. Id. Notably, the Court found that the process did not
begin with a preordained outcome. Instead, the Department of Homeland Security “collected and
evaluated data regarding all foreign governments.” Id. at 2405. Accordingly, in the Supreme
Court’s opinions granting military and national security deference, a precursor to that grant of
deference was the finding that reasoned decision-making had taken place.
While the D.C. Circuit’s Judgment did find that the Mattis Plan “plausibly” rests on
reasoned military judgment, this finding was sufficient only to make the preliminary injunction
unjustified. Doe, 755 F. App’x at 25. This finding of plausibility does not negate the parties’
material dispute of fact as to whether or not the Mattis Plan was based on the military’s reasoned

5

judgment. In his concurring opinion, Judge Wilkins explicitly refused to express a view on
whether or not “the Mattis Plan was a product of comparable ‘considered professional judgment’
as the policy in Goldman.” Doe, 917 F.3d at 704 (Wilkins, J., concurring).
Based on these Supreme Court cases as well as the D.C. Circuit’s judgment, the Court
finds that military deference is not guaranteed based on the mere fact that the challenged policy
concerns the military. Instead, military deference is granted to the military’s “considered
professional judgment” after the military has undertaken a “reasonable evaluation.” Goldman,
475 U.S. at 508; Rostker, 453 U.S. at 68.
Judge Williams’s concurring opinion in the judgment reversing this Court’s denial of
Defendants’ motion to dissolve the preliminary injunction does not change the Court’s analysis.
Defendants are correct that Judge Williams implies that deference applies in cases related to the
military regardless of whether or not the challenged decision was the result of reasoned decisionmaking. According to Judge Williams, deference applies to the decisions of military officials
“not because of their expertise—but because the military authorities have been charged by the
constitutionally responsible branches—that is, the Executive and Legislative Branches—with
carrying out our Nation’s military policy.” Doe, 917 F.3d at 730 (Williams, J., concurring)
(internal quotation marks omitted). According to Judge Williams, no further discovery should be
granted in this case because the Court should simply defer to the military’s judgment.
However, no other judge from the three-judge panel joined Judge Williams’s concurring
opinion. And, Judge Williams’s concurring opinion is not binding on this Court. Moreover,
another judge on the panel, Judge Wilkins, also filed a concurring opinion refuting some of
Judge Williams’s arguments. Judge Wilkins expressly disagreed with the notion that further
discovery in this case should be prohibited. Judge Wilkins explained that “Congress and the

6

Executive receive deference only where military policies are based upon the ‘considered
professional judgment’ of ‘appropriate military officials’ and only after finding that these
policies ‘reasonably and evenhandedly regulate’ the matter at issue.” Id. at 703-04 (Wilkins, J.,
concurring) (quoting Goldman, 475 U.S. at 509-10). According to Judge Wilkins, the standard of
review which should be used to assess the Mattis Plan depends on a variety of factors including
“whether the policy was motivated by animus,” “what military purposes are furthered by the
policy,” and “whether Congress or the Executive used considered judgment and accommodated
the servicemembers’ rights in a reasonable and evenhanded manner.” Id. at 704 (Wilkins, J.,
concurring). Additional discovery into the decision-making process of the Panel’s development
of the Mattis Plan is required to answer these questions.
Accordingly, the Court finds that Judge Williams’s concurring opinion fails to establish
the level of deference the Court must exercise in evaluating the Mattis Plan. The level of
deference to be applied depends, in part, on whether or not the Mattis Plan is a product of
reasoned decision-making and an even-handed evaluation of the evidence.
In sum, the Court concludes that, on the current record, neither party has established the
level of deference owed to the Mattis Plan due to the absence or presence of reasoned military
judgment. Additional discovery into the process of the Panel in developing the Mattis Plan is
needed before the Court can determine the level of deference owed to the Plan.
II.

Relevant Time Period for Discovery
The next broad issue presented in the parties’ discovery briefing involves the relevant

time period for discovery. Plaintiffs contend that Defendants have improperly withheld
documents relating to the decision to delay implementation of the Carter Policy and the preTweet review process. According to Plaintiffs, Defendants attempt to avoid claims of animus by

7

alleging that, even before the President’s 2017 Tweet, then-Secretary of Defense Jim Mattis had
already initiated an independent review of the Carter Policy. Plaintiffs contend that they are
entitled to documents relating to the scope and purpose of the pre-Tweet review of the Carter
Policy. The Court disagrees and concludes that Plaintiffs’ discovery requests should focus on the
development of the Mattis Plan, not on the development or delay of policies which came before.
The Court begins by addressing one of Plaintiffs’ primary arguments in support of
discovery into the pre-Tweet review of the Carter Policy. Plaintiffs contend that they “have a
specific basis for [requesting] information relating to the scope and purpose of [the pre-Tweet]
review, because information already produced by the government indicates that the object of the
pre-Tweet process was to justify reversing the Carter Policy.” Pls.’ Mot., ECF No. 213-1, 10.
However, in its September 13, 2019 Sealed Memorandum Opinion, the Court found that the
document on which Plaintiffs primarily rely is subject to multiple privileges and was
inadvertently produced by Defendants. See ECF Nos. 234, 235. Accordingly, Defendants have
successfully clawed-back this document, and Plaintiffs may not rely on it to further their
argument.
Besides this clawed-back document, Plaintiffs make the conclusory assertion that “[o]ther
documents also strongly support [the] inference” that Defendants had a pre-ordained goal of
reversing the Carter Policy. Pls.’ Mot., ECF No. 213, 11. However, in support, Plaintiffs cite
only one document from Heather Wilson, the Secretary of the Air Force, writing “‘with such
small numbers, it is hard to make the case that this change in policy [to transgender accessions]
would cause readiness problems that would lessen our ability to fight and win on the
battlefield.’” Id. at 11 n.5 (quoting Ex. N, ECF No. 214-15, 6). But, Plaintiffs omit the two
preceding lines of Ms. Wilson’s email where she wrote, “I accept that deployability is an issue. I

8

also accept that stability is an issue, particularly given the very high percentage of the
transgender individuals who return to their birth gender by early adulthood.” Ex. N, ECF No.
214-15, 6. Additionally, Defendants have provided Plaintiffs with the Air Force’s formal
recommendation to then-Secretary Mattis concerning the delay of the Carter Policy. See Ex. 7,
ECF No. 218-7 (filed under seal). Given the full context of Ms. Wilson’s statements as well as
the Air Force’s formal recommendation, the Court finds that, on this record, Plaintiffs’ cherrypicked statement does not support further discovery into the delay of the Carter Policy.
The Court further notes that the D.C. Circuit’s judgment reversing this Court’s denial of
Defendants’ motion to dissolve the preliminary injunction somewhat curtails Plaintiffs’ line of
argument. The D.C. Circuit concluded that, in developing the Mattis Plan, Defendants had taken
steps to cure the “procedural deficiencies” that had plagued Defendants’ prior attempts to
prevent transgender accessions in the military. Doe, 755 F. App’x at 23. These steps included
“the creation of a panel of military and medical experts, the consideration of new evidence
gleaned from the implementation of the policy on the service of transgender individuals
instituted by then-Secretary of Defense Ash Carter (“the Carter Policy”), and a reassessment of
the priorities of the group that produced the Carter Policy.” Id. Based on these new steps, the
D.C. Circuit found that it was error “to conclude that the Mattis Plan was foreordained.” Id. In
sum, the D.C. Circuit concluded that the Mattis Plan represents a new policy, which should be
judged on the process used for its own development, not on the processes which led to the
development or delay of prior policies, such as the delay of the Carter Policy.
The notion that the Mattis Plan should be judged on its own merits, rather than on the
merits of prior policies, is echoed in the Supreme Court’s recent decision in Trump v. Hawaii,
138 S.Ct. 2392 (2018). While Hawaii involved a policy which was found to be facially neutral,

9

the Court still finds the Supreme Court’s analysis instructive. In Hawaii, the Court considered an
executive policy barring entry to the United States by nationals from eight foreign countries, six
of which have a Muslim-majority population. The Court refused “to probe the sincerity of the
stated justifications for the policy by reference to extrinsic statements.” 138 S.Ct. at 2418.
Instead, the Court found that the policy itself “reflect[ed] the results of a worldwide review
process undertaken by multiple Cabinet officials and their agencies.” Id. at 2421. Finding the
process leading to the development of the policy sufficient, the Supreme Court upheld the policy.
Id. at 2423.
In line with the D.C. Circuit’s instruction as well as the Supreme Court’s opinion in
Hawaii, the Court finds that discovery into the decision to delay implementation of the Carter
Policy is not appropriate at this time. Plaintiffs have not provided sufficient evidence connecting
the decision-making process involved in the development of the Mattis Policy to the decision to
delay implementation of the Carter Policy. Between the delay of the Carter Policy and the
development of the Mattis Plan, Defendants engaged in a new process, involving the creation of
the Panel and the evaluation of new evidence. Absent a demonstrated connection, additional
discovery into the delay of the Carter Policy is not appropriate at this time. However, the Court
does not foreclose the possibility that Plaintiffs will, at some future time, introduce evidence
connecting the decision to delay the Carter Policy to the development of the Mattis Plan, which
could provide grounds for additional discovery into the Carter Policy. But, at this time, Plaintiffs
are precluded from further discovery on this issue.
III.

Use of the Deliberative Process Privilege
Third, the parties disagree on whether or not the deliberative process privilege can be

used to exclude from discovery documents related to the development of the Mattis Plan. The

10

deliberative process privilege is used to shield from discovery documents which are predecisional and deliberative such as documents “reflecting advisory opinions, recommendations
and deliberations comprising part of a process by which governmental decisions and policies are
formulated.” NLRB v. Sears, Roebuck & Co., 421 U.S. 132, 150 (1975) (internal quotation marks
omitted). Defendants attempt to shield from discovery numerous documents on the grounds that
they are protected by the deliberative process privilege. Plaintiffs argue that the deliberative
process privilege does not apply to documents both before and after the President’s 2017 Tweet
announcing a ban on transgender military accessions because Defendants’ deliberative process is
squarely at issue in this lawsuit. The Court agrees with Plaintiffs in part but finds that Plaintiffs
have gone too far in limiting the deliberative process privilege. For reasons that will be explained
below, the Court finds that Defendants cannot invoke the deliberative process privilege to shield
from discovery those documents which were used or considered in the development of the Mattis
Plan. Documents relating to the delay of the Carter Policy, the pre-Tweet process, the 2017
Presidential memorandum, or other events, are not encompassed by the Court’s decision.
Plaintiffs argue that, because Defendants’ decision-making process is squarely at issue in
this lawsuit, the deliberative process privilege is unavailable to Defendants. One of Defendants’
principal arguments throughout this lawsuit is that the Mattis Plan is entitled to deference. As the
Court has explained, the level of deference owed to the Mattis Plan depends, in part, on whether
or not the Plan is a product of “the considered professional judgment of appropriate military
officials” and whether or not the Plan “reasonably and evenhandedly regulate[s] the matter at
issue.” Doe, 917 F.3d at 703-04 (Wilkins, J., concurring) (internal quotation marks omitted).
However, the Court cannot ascertain whether or not the Mattis Plan is the product of “the
considered professional judgment of appropriate military officials” if Defendants restrict

11

discovery into the decision-making process resulting in the development of the Mattis Plan.
Because determining whether or not the Mattis Plan is the product of reasoned military judgment
requires insight into the process used to develop the Plan, the deliberative process privilege does
not apply to documents used or considered in the development of the Mattis Plan. See Recycling
Solutions, Inc. v. D.C., 175 F.R.D. 407, 408 (D.D.C. 1997) (explaining that privilege “cannot be
used both as a sword and as a shield”); Ideal Elec. Sec. Co. v. Int’l Fid. Ins. Co., 129 F.3d 143,
151 (D.C. Cir. 1997) (explaining, in the attorney-client privilege context, that privilege is
implicitly waived when “the client places otherwise privileged matters in controversy”).
Defendants disagree and claim that the deliberative process privilege is applicable to
documents that Defendants used or considered in developing the Mattis Plan. Defendants’
primary argument in support of the use of the deliberative process privilege is an argument that
the Court has already addressed. Defendants argue that “Plaintiffs’ contention that the
deliberative process may be probed to test whether deference is due would stand the very
doctrine of military deference on its head.” Defs.’ Res., ECF No. 218, 25. The Court disagrees.
As the Court previously explained, “Congress and the Executive receive deference only where
military policies are based upon the ‘considered professional judgment’ of ‘appropriate military
officials’ and only after finding that these policies ‘reasonably and evenhandedly regulate’ the
matter at issue.” Doe, 917 F.3d at 703-04 (Wilkins, J., concurring) (quoting Goldman, 475 U.S.
at 509-10). Accordingly, discovery into the deliberative process used to develop the Mattis Plan
is relevant and appropriate. Defendants cannot ask for deference, then proceed to invoke the
deliberative process privilege to prevent discovery into the evidence which would show whether
or not such deference is appropriate.

12

Defendants also question the applicability of the primary case relied on by Plaintiffs, In
re Subpoena Duces Tecum Served on the Office of the Comptroller of the Currency, 145 F.3d
1422 (D.C. Cir. 1998). In In re Subpoena, the D.C. Circuit explained that the deliberative
process privilege “was fashioned in cases where the governmental decision-making process is
collateral to the plaintiffs’ suit.” 145 F.3d at 1424. In cases like this, where the governmental
decision-making process is central to the suit, defendants cannot assert the deliberative process
privilege over documents which illuminate that decision-making process. The D.C. Circuit went
on to state that “[i]f the plaintiff’s cause of action is directed at the government’s intent, … it
makes no sense to permit the government to use the privilege as a shield. For instance, it seems
rather obvious to us that the privilege has no place … in a constitutional claim for
discrimination.” Id. Given that Plaintiffs in fact assert a constitutional claim for discrimination,
the Court finds that In re Subpoena is directly on point. Because Defendants’ decision-making
process in developing the Mattis Plan is central in this case, the deliberative process privilege
does not apply to documents used or considered by the Panel.
In addition to challenging Defendants’ decision-making process, Plaintiffs also challenge
Defendant’s intent in issuing the Mattis Plan. On the current record, Plaintiffs have provided
adequate justification for questioning Defendants’ intent in issuing the Mattis Plan based on the
fact that the Plan does not appear to be facially neutral and that the Plan represents an abrupt
departure from the Carter Policy, which allowed for transgender accession and service and had
been carefully reviewed and developed only a year before. Given that Defendants’ intent and
decision-making process are at the forefront of this case, it makes little sense to prohibit,
pursuant to the deliberative process privilege, discovery into these issues. See Elkins v. District
of Columbia, 250 F.R.D. 20, 27 (D.D.C. 2008) (explaining that the deliberative process privilege

13

did not apply when the plaintiffs challenged the intent and mental state of government
defendants); Waters v. U.S. Capitol Police Bd., 218 F.R.D. 323, 324 (D.D.C. 2003) (explaining
that “this Circuit does not permit the application of the deliberative process privilege to thwart
discovery of information in a case in which a plaintiff challenges governmental action as
discriminatory”).
The Court acknowledges Defendants’ argument that challenges to military policies are
different than challenges to other governmental policies. However, this difference does not alter
the Court’s conclusion that the deliberative process privilege should not be used to shield
discovery into Defendants’ decision-making process and intent when the extent and scope of that
decision-making process is a central issue in this lawsuit. See Adair v. Winter, 451 F. supp. 2d
202, 209 (D.D.C. 2006) (holding the deliberative process privilege inapplicable as to the
plaintiffs’ discovery requests into Selective Early Retirement board proceedings where plaintiffs
were attempting to show that the Navy discriminated against the plaintiffs on the basis of
religion), reconsidered on non-applicable statutory grounds sub nom. In re Navy Chaplaincy,
512 F. Supp. 2d 58 (D.D.C. 2007); Chaplaincy of Full Gospel Churches v. Johnson, 217 F.R.D.
250, 258 (D.D.C. 2003) (allowing discovery into the selection-board proceedings despite the
deliberative process privilege because the plaintiffs had provided some “reason to believe” that
the Navy discriminated against non-liturgical chaplains), rev’d in part, vacated in part on other
inapplicable statutory grounds sub nom. In re England, 375 F.3d 1169 (D.C. Cir. 2004).
Defendants further argue that In re Subpoena is not relevant because the governmental
intent behind the development of the Mattis Plan is not at issue in this lawsuit. Defendants rely
primarily on Trump v. Hawaii, 138 S. Ct. 2392 (2018) to argue that the intent of the
governmental decisionmakers is not at issue in this case. However, Hawaii is distinguishable

14

from the facts of this case. In Hawaii, the Supreme Court determined that the proclamation
barring entry to the United States by foreign nationals from certain countries was a facially
neutral policy reflecting “the results of a worldwide review process undertaken by multiple
Cabinet officials and their agencies.” 138 S.Ct. at 2421. In this case, the record is not sufficiently
developed for the Court to make any such finding. Moreover, in Hawaii, even given the facially
neutral policy and the defendants’ reasoned decision-making, the Supreme Court still considered
evidence that went “behind the face of the Proclamation.” Id. at 2420. Additionally,
constitutional protections of foreign nationals seeking admission into the United States are far
more limited than protections for United States citizens alleging governmental discrimination
based on a protected status. Id. at 2418-19. Accordingly, the Court concludes that the Supreme
Court’s decision in Hawaii does not render irrelevant the government’s intent or decisionmaking process in developing the Mattis Plan.
Accordingly, the Court concludes that Defendants may not assert the deliberative process
privilege over every document that was used or considered in the development of the Mattis
Plan. The Court understands that it likely will still have to consider whether or not individual
documents fall into this subcategory of documents to which the deliberative process privilege
does not apply. However, the Court’s conclusion should further limit the disputes between the
parties. Moreover, the fact that the deliberative process privilege does not apply to documents
which were used or considered in the development of the Mattis Plan does not bar Defendants
from asserting the deliberative process privilege over other documents. Beyond requesting
discovery into the development of the Mattis Plan, Plaintiffs also argue that the deliberative
process privilege should not be used to shield from discovery “materials relating to the decision
to delay implementation of the open accessions policy and the pre-Tweet process.” Pls.’ Mot.,

15

ECF No. 213-1, 14. As the Court has previously explained, on the current record, Plaintiffs have
not shown that discovery into the decision to delay implementation of the open accessions policy
and the pre-Tweet process is probative of the development of the Mattis Plan. Because Plaintiffs
have not produced evidence showing that such information would be probative of Defendants’
decision-making process in developing the Mattis Plan, the Court will not prevent Defendants
from asserting the deliberative process privilege over such information.
The Court has concluded that the deliberative process privilege does not apply to
documents that were used or considered in the development of the Mattis Plan. However, as an
alternative ground, the Court also finds that Plaintiffs’ need for these documents outweighs the
deliberative process privilege. The deliberative process privilege “is a qualified privilege and can
be overcome by a sufficient showing of need.” In re Sealed Case, 121 F.3d 729, 737 (D.C. Cir.
1997). In analyzing whether or not the privilege is overcome, courts consider (1) “the relevance
of the evidence,” (2) “the availability of the evidence,” (3) “the seriousness of the litigation,” (4)
“the role of the government,” and (5) “the possibility of future timidity by government
employees.” Id. at 737-38. Defendants concede that the third and fourth factors—the seriousness
of the litigation and the role of the government—weigh in Plaintiffs’ favor. Defs.’ Res., ECF No.
218, 35 n.12. However, the other three factors are disputed. In assessing these factors, the Court
finds that Plaintiffs’ need for documents which were used or considered in the Panel’s
development of the Mattis Plan overcomes Defendants’ assertion of privilege.
Before evaluating the relevant factors, the Court acknowledges the Ninth Circuit’s recent
decision in Karnoski v. Trump, 926 F.3d 1180 (9th Cir. 2019). In that decision, the Ninth Circuit
vacated the district court’s discovery order which had broadly prevented Defendants from
asserting the deliberative process privilege as to any documents. 926 F.3d at 1204. In vacating

16

the discovery order, the court explained that “the existing record is not adequate to evaluate the
relevance of all of the requested information, at least in terms of balancing production of
materials against the military’s countervailing confidentiality interest.” Id. at 1206. Instead, the
court recommended that “the district court should consider classes of documents separately when
appropriate.” Id. Here, the Court accepts the Ninth Circuit’s guidance and conducts the balancing
test as to a discrete subset of documents—documents which were used or considered in the
development of the Mattis Plan. See In re Sealed Case, 121 F.3d at 737 (explaining that the
determination that privilege is overcome “is to be made flexibly on a case-by-case, ad hoc
basis”).
Looking to the first disputed factor, the Court finds that the relevance of the evidence
weighs in Plaintiffs’ favor. As the Court has already explained, whether or not Defendants
exercised reasoned military judgment and developed the Mattis Plan in an evenhanded manner is
probative of the degree of deference owed to the Plan. As such, information concerning the
decision-making process leading to the development of the Mattis Plan is highly relevant to this
lawsuit. The documents used or considered in the development of the Mattis Plan are probative
of Defendant’s decision-making process in developing the Mattis Plan. As to this limited class of
documents, the Court finds that Plaintiffs have established the relevance and importance of the
evidence.
Next, the Court similarly finds that the availability of the evidence weighs in Plaintiffs’
favor. The evidence relating to the decision-making process in developing the Mattis Plan is
almost exclusively in Defendants’ control. Only Defendants have access to the material that was
used or considered in the development of the Mattis Plan.

17

Defendants have two primary arguments as to why their exclusive control over the
relevant documents should not result in the deliberative process privilege being overcome. First,
Defendants contend that, even without the withheld documents, Plaintiffs have had ample
discovery, including the production over 30,000 non-privileged documents. In support,
Defendants cite Hinckley v. United States, 140 F.3d 277 (D.C. Cir. 1998), in which the D.C.
Circuit found that the balancing test did not weigh in favor of disclosure, in part, because the
defendant “ha[d] already given [the plaintiff] access to a tremendous amount of information,
including all of the evidence that was before the Review Board as well as the Review Board’s
final decision and explanation for it.” 140 F.3d at 286. The Court is not persuaded that Hinckley
is relevant. Here, Defendants have used the deliberative process privilege to withhold and
attempt to claw-back evidence that was used or considered by the Panel in making the policy
recommendations that resulted in the Mattis Plan. For example, Defendants withheld, pursuant to
the deliberative process privilege, a Powerpoint presentation from the Merit Panel’s Kickoff
meeting which appears to show different assumptions and courses of actions considered by the
Panel. April 24, 2018 Response to Order of the Court, ECF No. 124. As such, this case is not like
Hinckley, in which the documents already produced rendered the withheld documents
superfluous.
Second, Defendants contend that Plaintiffs have failed to take depositions of key figures
such as the chair of the Panel of Experts which could also provide Plaintiffs with the requested
information. However, if Defendants are excluding from discovery documents related to the
evidence used or considered in the development of the Mattis Plan, there is little reason to think
that Defendants would not also bar those being deposed from answering questions about
evidence used or considered in the development of the Plan. For example, during the deposition

18

of Colonel Mary Krueger, the Assistant Deputy for Health Affairs in the Office of the Assistant
Secretary of the Army for Manpower and Reserve Affairs, Plaintiffs asked Ms. Krueger if the
Panel considered a return to the 2016 open service policy as an option. Defendants instructed
Ms. Krueger not to answer the question pursuant to the deliberative process privilege. Ex. Q,
ECF No. 214-18, 95: 3-11. Moreover, absent the production of documents that were used or
considered in the development of the Mattis Plan, it is not clear to the Court how Plaintiffs would
be able to engage in fulsome depositions of key decision-makers and develop a line of inquiry as
well as possibly refresh witnesses’ recollections and impeach witnesses with documentary
evidence.
Accordingly, the Court finds that the second disputed factor also weighs in favor of
disclosure. Defendants have exclusive control over evidence which is probative of Plaintiffs’
claims. Absent discovery, Plaintiffs have essentially no other avenue to obtain such information.
Finally, the Court considers the last disputed factor—the possibility of future timidity by
government actors. Plaintiffs contend that disclosure will not chill government deliberations “in
future policy discussions that follow appropriate procedures and are based on legitimate policy
concerns.” Pls.’ Mot., ECF No. 213-1, 19. Defendants contest Plaintiffs’ claim and contend that
the disclosure of documents relating to the deliberative process would greatly hinder frank and
independent discussion regarding contemplated policies and decisions in the future. See, e.g., Ex.
6, Dec. of Robert E. Easton, ECF No. 218-6, ¶ 24 (“release of DoD information protected by the
deliberative process privilege in this case would have a substantial and immediate chilling effect
on policy deliberation and development within DoD writ large”); Ex. 3, Dec. of Andreas Thum,
ECF No. 218-3, ¶¶ 12-13 (“[d]isclosure would … have a chilling effect on the Army’s personnel

19

when developing forthcoming policies, and thus an adverse effect on the quality and integrity of
the Army’s future decision making process”).
The Court acknowledges the seriousness of Defendants’ concern. However, the Court
notes that Defendants’ feared chilling effect can be somewhat assuaged by taking certain steps to
restrict the public disclosure of sensitive information. For example, the Court can issue a
protective order, Defendants can redact certain information, documents can be restricted to
attorneys’ eyes only, and the Court can conduct in camera review over any particularly sensitive
documents. Moreover, insofar as disclosure of the documents would reveal any governmental
misconduct, such as discrimination against transgender servicemembers, “‘the public’s interest in
honest, effective government’” would overcome the governmental interest in hiding such
misconduct. Cobel v. Norton, 213 F.R.D. 1, 4 (D.D.C. 2003) (quoting In re Sealed Case, 121
F.3d at 737-38). As such, while the Court finds that Defendants’ concerns of future government
timidity weigh against disclosure, the Court also finds that steps can be taken so that any chilling
effect is minimized.
Accordingly, the Court finds that out of the five factors to be weighed in determining if
Plaintiffs overcome the deliberative process privilege only one factor weighs in favor of
maintaining the privilege—the possibility of future timidity by government actors. However, this
one factor does not outweigh the strength of the other four factors which support disclosure.
Moreover, steps can be taken to shield particularly sensitive disclosed documents from the public
while still allowing the important information to be considered by the Court. As such, the Court
concludes that Plaintiffs have overcome the deliberative process privilege for documents which
were used or considered in the development of the Mattis Plan.

20

In sum, the Court concludes that the deliberative process privilege does not bar discovery
into documents which were used or considered in the development of the Mattis Plan. The Court
concludes that the deliberative process privilege does not apply to this subset of documents as
they are probative of Plaintiffs’ challenges to Defendants’ decision-making process and intent.
Moreover, even if the deliberative process privilege were to apply to this subset of documents,
Plaintiffs’ need for the documents outweighs Defendants’ privilege. Again, the Court recognizes
that it likely will still have to consider whether or not individual documents fall into this
subcategory of documents to which the deliberative process privilege does not apply.
IV.

Plaintiffs’ Request for Raw Data and Personnel Files
Finally, the Court will address Plaintiffs’ request for the raw data, personnel files, and

field reports supporting the statistical summaries and conclusions in the Mattis Report. Plaintiffs
explain that they are “challenging the factual justifications for the Mattis Plan—in particular, the
government’s claim that permitting transgender troops to serve openly under the Carter Policy
undermines military readiness, imposes burdensome medical costs, and disrupts unit cohesion.
Plaintiffs consequently are entitled to the documents and information that the government relied
on to generate the summary statistics and conclusions included in the report and accompanying
Mattis Plan, considered by the Review Panel, and presented to the Armed Services Committee.”
Pls.’ Mot., ECF No. 213-1, 8. Plaintiffs specify they are requesting the raw data, personnel files
that include performance evaluations and assessments, and reports from the field in addition to
updated data regarding the readiness and performance of active duty service members with
gender dysphoria.
Defendants have numerous arguments as to why such discovery should not be
permitted—some stronger than others. First, Defendants contend that providing medical and

21

personnel records for non-party servicemembers implicates privacy issues. However, these
concerns are easily remedied by redacting all personally-identifying information and by issuing a
protective order regarding the disclosure of sensitive information. As such, the Court does not
consider privacy concerns to be a reason not to disclose the requested data and records.
Next, Defendants contend that such information is not relevant as it would not assist the
Court in analyzing the Mattis Plan. Defendants argue that such discovery is adverse to military
deference principles and is not relevant to the Court’s analysis as such evidence would not “shed
light on the ‘operation and purpose’ of the policy.” Defs.’ Res., ECF No. 218, 17 (quoting Doe,
917 F.3d at 705-06 (Wilkins, J., concurring)). Defendants claim that Plaintiffs request this
information for the improper purpose of introducing their own “expert testimony that
contradict[s] the military experts about whether the policies at issue [are] justified under the
current circumstances.” Id. at 18 (quoting Doe, 917 F.3d at 706 (Wilkins, J., concurring)).
Even if the Court assumes for the purpose of this Section that the Mattis Plan is entitled
to deference, the Court must still ensure that Defendants’ evaluation of the evidence in
developing the Plan was reasonable. See Rostker, 453 U.S. at 68 (explaining that a court should
not “substitute … [its] own evaluation of evidence for a reasonable evaluation” by military
officials). Deference does not imply that “[s]imply labeling [a] decision ‘military’ …
automatically guide[s] a court to the correct constitutional result.” Id. at 70. The Court does not
seek access to the evidence underlying the Mattis Plan in order to determine if, on an
independent evaluation, the Court would reach the same conclusion. Instead, the Court requires
access to the evidence underlying the Mattis Plan in order to ensure that the Plan reasonably and
evenhandedly regulates the service and accession of transgender military personnel. See
Goldman, 475 U.S. at 510 (upholding military policy because “those portions of the regulations

22

challenged here reasonably and evenhandedly regulate dress in the interest of the military’s
perceived need for uniformity”). It is not the role of the Court to question the military’s stated
objectives. But, the Court cannot abdicate its duty to ensure that the chosen policy reasonably
and evenhandedly furthers those objectives within the bounds of the Constitution. Access to at
least some of the data underlying Defendants’ decisions will allow the Court to determine if “the
class [targeted by the Mattis Plan] is similarly situated to others” and if Defendants
“accommodated the servicemembers’ rights in a reasonable and evenhanded manner.” Doe, 917
F.3d at 704 (Wilkins, J., concurring). Accordingly, the Court finds that such information is
relevant regardless of the degree of deference owed to the Mattis Plan.
Finally, Defendants argue that the discovery of such information would be unduly
burdensome for Defendants to collect. Defendants explain that, to create their summaries, charts,
and documents presented to the Panel, they ran searches and extracted data from the MDR. The
MDR is a centralized data repository that receives data from over 260 healthcare facilities and
from the TRICARE Purchase Care program. Dec. of Terry Adirim, Ex. 5, ECF No. 218-5, ¶ 5.
The MDR does not contain actual medical records and no personnel medical records were
searched in the creation of the summaries, charts, and documents. Instead, searches were run
within the MDR and later analyzed. Id. at ¶ 6. In conducting the searches, the researchers
searched for active duty service members with a diagnosis of gender dysphoria and compared
them with active duty service members without such a diagnosis. Id. at ¶ 7. Defendants report
that the MDR was used rather than the actual medical records because collecting such medical
and personnel records would have been challenging and unnecessary. Id. at ¶ 10.
Similarly, Defendants contend that, in order to produce the underlying data, personnel
files, field reports and other back-up documentation that Plaintiffs request, Defendants would

23

“have to hire a third-party contractor at a cost of likely hundreds of thousands of dollars.” Id. at
¶ 12. This endeavor would require “extracting the electronic medical records from multiple,
separate data systems” and could take months or over a year. Id. After obtaining such data,
Plaintiffs would then need to review each record and upload it into a highly secure database so
that a statistical analysis could be performed. Id. at ¶ 13.
Prior to deciding the scope of permissible discovery, the Court requires additional
information. Plaintiffs’ initial discovery request is extremely broad and encompasses all “data
personnel files, and field reports that purportedly support the statistical summaries and
conclusions contained in the Mattis report … as well as any other data relating to the readiness
and ongoing service of active duty servicemembers diagnosed with gender dysphoria.” Pls.’
Mot., 213-1, 8. At this point, the number of records encompassed by such a request is unclear.
The timeline encompassed by Plaintiffs’ request is also unclear—for example how many years of
medical and other personnel records would Plaintiffs require? Given this lack of clarity, the
Court is especially concerned by Defendants’ contention that such a production would require
copious amounts of time and money.
In their Reply, Plaintiffs provide some narrowed specifics as to their request. For
example, Plaintiffs explain that they seek the medical records of service members with gender
dysphoria in order to differentiate between mental health visits required administratively and
those visits reflecting actual mental health issues in order to rebut Defendants’ allegedly false
claims that transgender service members show greater mental instability. Pls.’ Reply, ECF No.
222, 9. Plaintiffs also specifically seek “the additional deployability data that has become
available since the issuance of the Mattis Report” in order to show that transgender service
members do not have a high rate of nondeployability. Id. at 10. And, Plaintiffs ask for data

24

relating to suicide attempts and suicidal ideation by transgender and non-transgender
servicemembers for the purpose of showing that transgender service members do not have a
higher rate of suicidal ideation and suicide attempts. Id. at 11. Finally, Plaintiffs further request
any field reports which demonstrate that the presence of transgender service members
undermines unit cohesion. Id. Plaintiffs contend that the absence of such reports could show that
Defendants’ concerns about transgender service members undermining unit cohesion are
pretextual. Plaintiffs contend that this information will help to establish that the evidence
allegedly relied upon by Defendants to justify the Mattis Plan is misleading and was not
analyzed in an evenhanded manner.
Because these narrowed requests were presented in Plaintiffs’ Reply, Defendants did not
have the opportunity to respond. Moreover, Defendants indicate that the parties have not yet
properly met and conferred concerning Plaintiffs’ request for such information. Defs.’ Res., ECF
No. 218, 15 n.4. As such, it is not clear to the Court whether or not such information could be
produced from Defendants’ MDR system. Defendants also did not have the opportunity to
explain whether or not, if Plaintiffs agreed to limit themselves to these narrowed requests, such
discovery would be still be overly burdensome. Before the Court enters any Order on the
whether or not the discovery of such information should be permitted, the Court requires
additional information pertaining to the burdensomeness of such discovery. The Court suggests
that the parties meet and confer regarding Plaintiffs’ request for data underlying Defendants’
summary statistics and conclusions. When conferring, the parties should keep in mind this
Memorandum Opinion’s overarching guidance on the breadth and scope of discovery relevant to
this lawsuit.

25

V.

Conclusion
For the reasons provided above, the Court GRANTS IN PART AND DENIES

WITHOUT PREJUDICE IN PART Plaintiffs’ [213-1] Motion to Compel. The Court’s above
guidance has resolved many of the overarching discovery disputes between Plaintiffs and
Defendants. However, the Court anticipates that additional disputes may arise in carrying out the
Court’s findings and conclusions. Plaintiffs and Defendants should meet and confer in an attempt
to resolve or narrow remaining disputes. The parties shall file a Joint Status Report by
OCTOBER 25, 2019, indicating whether or not additional briefing on discovery issues is
necessary. Again, the Court notes that in complying with the Court’s guidance contained herein,
the parties are not forfeiting or waiving their arguments and objections which are preserved for
the record.
SO ORDERED.
/s/
COLLEEN KOLLAR-KOTELLY
United States District Judge

26