66. The right “derives from two independent sources: the common law and the First
Amendment.” Va. Dep’t of State
Police, 386 F.3d at 575.
Criminal proceedings and related documents are thus presumptively open to the
public. See Press-Enter. Co. v. Superior Court of Cal.,
478 U.S. 1, 9 (1986) (First
Amendment); Va. Dep’t of State
Police, 386 F.3d at 575 (common law). But because a
8
presumption may be overcome, the public’s right of access is qualified, not absolute. See,
e.g., Am. Civil Liberties Union v. Holder,
673 F.3d 245, 252 (4th Cir. 2011).
Moreover, the strength of the right of access varies depending on whether the
public’s right of access to the document or proceeding derives from the common law or
the First Amendment. “[T]he common law does not afford as much substantive protection
to the interests of the press and the public as does the First Amendment.”
Doe, 749 F.3d at
265 (quoting In re
Application, 707 F.3d at 290). Under the common law, there is a
rebuttable presumption that the public’s right of access applies as to “all judicial documents
and records.”
Id. at 266 (emphasis added). To overcome that presumption, “a court must
find that there is a ‘significant countervailing interest’ in support of sealing that outweighs
the public’s interest in openness.” In re
Application, 707 F.3d at 293 (quoting Under
Seal,
326 F.3d at 486).
The public’s right to access documents under the First Amendment is narrower in
scope but stronger in force. It applies “only to particular judicial records and documents.”
Doe, 749 F.3d at 266 (quoting Stone v. Univ. of Md. Med. Sys. Corp.,
855 F.2d 178, 180
(4th Cir. 1988)). But where the First Amendment applies, a “compelling government
interest” must justify denial of access, and closure must be “narrowly tailored to serve that
interest.”
Id. (quoting In re Wash.
Post, 807 F.2d at 390). Accordingly, a court presented
with a sealing request must first “determine the source of the right of access with respect
9
to each document, because only then can it accurately weigh the competing interests at
stake.”
Id. (quoting Va. Dep’t of State
Police, 386 F.3d at 576). 5
The First Amendment’s application turns on “two factors: historical tradition and
the function of public access in serving important public purposes.” In re Wash.
Post, 807
F.2d at 389. This is commonly known as the “experience and logic” test. See
Press-Enter.,
478 U.S. at 9. The first factor, historical tradition, asks whether the document or proceeding
in question was traditionally open to the public. In re Wash.
Post, 807 F.2d at 389. The
second factor asks whether public access would curb “prosecutorial or judicial misconduct”
and further “the public’s interest in understanding the criminal justice system.”
Id.
Our precedent largely supports applying the public’s First Amendment right of
access to the district court’s order. We have held that the First Amendment right applies to
sentencings and plea hearings, as well as to documents filed in connection with those
proceedings.
Id. at 390. And, in the context of a summary judgment motion, we have
concluded that the First Amendment right extends to a judicial opinion when the underlying
materials and evidence forming the substance of an opinion are already covered.
Doe, 749
F.3d at 267–68. Accordingly, the document at issue here—a court order denying a sentence
5
Here, the district court made no attempt to determine the source of the public’s right to
access its order denying Defendant’s § 3582 motion. Instead, the court summarily
concluded that Defendant had not overcome the presumption of public access. The district
court cited this Court’s opinion in Doe,
749 F.3d 246. But because that opinion concerned
both the First Amendment and the common law, a passing citation to it is no substitute for
the requisite analysis. Absent consideration of whether the public’s right of access derived
from the First Amendment or the common law, the district court could not have “accurately
weigh[ed] the competing interests at stake.”
Id. at 266 (quoting Va. Dep’t of State
Police,
386 F.3d at 576).
10
reduction—is likely encompassed by the First Amendment right of access. But see United
States v. Harris,
890 F.3d 480, 491–92 (4th Cir. 2018) (applying the common law right of
access in a case regarding access to a sentencing memorandum).
In any event, we need not, and do not, conclusively decide this issue. See
Holder,
673 F.3d at 252. Because Defendant’s compelling interests in sealing the district court’s
order outweigh the public’s interest in accessing it under either the common law standard
or the more rigorous First Amendment standard, we assume without deciding that the First
Amendment applies. See
id. (assuming without deciding that the First Amendment right of
access extended to qui tam complaints and related docket sheets).
B.
Even assuming the First Amendment right of access applies, however, “[t]he mere
existence of” such a right “does not entitle the press and public to access in every case.” In
re Wash.
Post, 807 F.2d at 390; see also
Holder, 673 F.3d at 252. Under the First
Amendment, the district court’s order denying Defendant’s motion for a sentence reduction
should be sealed “only if (1) closure serves a compelling interest; (2) there is a ‘substantial
probability’ that, in the absence of closure, that compelling interest would be harmed; and
(3) there are no alternatives to closure that would adequately protect that compelling
interest.” In re Wash.
Post, 807 F.2d at 392 (quoting
Press-Enter., 478 U.S. at 14); see also
id. at 393 n.9 (“Because we hold that the First Amendment right of access extends to
documents as well as to hearings, requests for the sealing or unsealing of documents must
also be evaluated under these constitutional tests.”).
11
Additionally, a court must base its decision on “specific factual findings” rather than
“conclusory assertions alone.”
Id. at 392. Here, the district court’s cursory order neither
engaged in the required three-part analysis nor based its decision on specific factual
findings. In fact, the district court made no specific findings at all.
We conclude that the district court erred. Defendant’s circumstances required the
court to seal its order referring to his cooperation because the order placed him in danger
and there were no narrower alternatives to sealing. In evaluating the risks Defendant faces,
we take judicial notice 6 of the findings contained in a 2016 report by the Committee on
Court Administration and Case Management of the Judicial Conference of the United
States (“CACM Report”), which evaluated the need to protect government cooperators in
federal prisons. 7 We also consider the Eastern District of North Carolina’s standing order
requiring the automatic sealing of motions related to a defendant’s substantial assistance.
See United States District Court, Eastern District of North Carolina, Standing Order No.
09-SO-02 (Feb. 12, 2010) (“E.D.N.C. Standing Order”). 8
6
We may take judicial notice of facts that are generally known, even where the district
court declined to do so. E.g., United States v. Lavender,
602 F.2d 639, 641 (4th Cir. 1979).
Moreover, we may take judicial notice of governmental reports. United States v. Cecil,
836
F.2d 1431, 1452 (4th Cir. 1988).
7
See Memorandum from the Comm. on Court Admin. & Case Mgmt. of the Judicial
Conference of the U.S. (June 30, 2016), in Meeting Agenda Book for the Advisory Comm.
on Criminal Rules 243–51 (Sept. 19, 2016),
http://www.uscourts.gov/sites/default/files/2016-09-criminal-agenda_book_0.pdf.
8
Available at http://www.nced.uscourts.gov/data/StandingOrders/09-SO-2.pdf.
12
1.
Defendant argues that sealing the district court’s order would serve the compelling
interest of protecting his safety. 9 “Courts have recognized that an interest in protecting the
physical and psychological well-being of individuals related to the litigation . . . may justify
restricting access.”
Harris, 890 F.3d at 492. And “as a general rule, the need to protect the
well-being of” a defendant “is even more elevated if judicial records suggest that the
defendant may have cooperated with law enforcement.”
Id.
Although Harris evaluated the public’s right of access under the common law, not
the First Amendment, the interest in protecting cooperators from harm is no less
compelling in the First Amendment context. We are aware of no authority stating
otherwise. To the contrary, several of our sister circuits have held that protecting
cooperating inmates serves a compelling interest under the First Amendment. See United
States v. Doe,
870 F.3d 991, 998 (9th Cir. 2017); United States v. Doe,
63 F.3d 121, 128
9
Defendant also argues that sealing the order would preserve the integrity of ongoing
investigations and would protect his family. Although we are in “complete agreement with
the general principle” that protecting ongoing investigations serves a compelling interest,
the party moving to seal must nonetheless provide “specific underlying reasons” describing
“how the integrity of the investigation reasonably could be affected by the release of such
information.” Va. Dep’t of State
Police, 386 F.3d at 579. Because Defendant has failed to
do so here, we cannot find that this is a compelling interest in his case. Moreover, protecting
family members is certainly a compelling interest, made all the more compelling when a
defendant cooperated with the government. See
Harris, 890 F.3d at 492. Although
Defendant did not raise the risk to his family before the district court, the Government did
not make a waiver argument, and we could excuse that defect. See Manning v. Caldwell,
930 F.3d 264, 271 (4th Cir. 2019). But because Defendant provides no further details about
the risk to his family, the E.D.N.C. Standing Order does not elaborate on such a risk, and
we otherwise find sealing to be necessary, we do not comment further on that potential
compelling interest.
13
(2d Cir. 1995); United States v. Corbitt,
879 F.2d 224, 235 (7th Cir. 1989); cf. Detroit Free
Press v. Ashcroft,
303 F.3d 681, 705–06 (6th Cir. 2002) (holding that “compelling interests
sufficient to justify closure” existed where “[d]ivulging the detainees’ identities may deter
them from cooperating” in a terrorist investigation because “[t]errorist groups could subject
these individuals or their families to intimidation or harm”); Wash. Post v. Robinson,
935
F.2d 282, 291 (D.C. Cir.), as amended (June 14, 1991) (leaving open the possibility that
“evidence that the release of a plea agreement may threaten . . . the safety of the defendant
. . . may well be sufficient to justify sealing a plea agreement in a different case”). 10
Moreover, harm to cooperators does not only affect those individuals. Rather, “[t]he
problem threatens public safety” and “interferes with the gathering of evidence, the
presentation of witnesses, and the sentencing and incarceration of cooperating defendants,
. . . pos[ing] a substantial threat to the underpinnings of the criminal justice system as a
whole.” CACM Report at 7. Therefore, the Court Administration and Case Management
Committee of the Judicial Conference concluded “that there is a compelling government
interest in addressing these issues.”
Id.
For its part, the government’s arguments against sealing focus on the likelihood of
harm to Defendant, not whether protecting cooperators is a compelling interest in the
10
Whether an interest is compelling may sometimes blend into the substantial-probability-
of-harm prong. The Ninth Circuit “address[ed] together the first two prongs” of the analysis
because it found “significant overlap” as to how those prongs “pertain[ed] to the facts of”
its case.
Doe, 870 F.3d at 998. We find it unnecessary to do so here given that our precedent
independently supports finding that protecting cooperators serves a compelling interest.
See
Harris, 890 F.3d at 492.
14
abstract. In these circumstances, we have no difficulty concluding that protecting
cooperators from harm is a compelling interest that can justify sealing.
2.
Where a compelling interest exists, sealing is nevertheless inappropriate unless a
court determines there is a “substantial probability” the interest will be harmed absent
sealing. In re Wash.
Post, 807 F.2d at 392. In his pro se motion to seal, Defendant asserted
that the district court’s order referring to his cooperation established an ongoing, significant
risk to his safety. He requested sealing before he was physically harmed.
Of course, such conclusory statements cannot alone establish a substantial
probability of harm.
Id. But the record demonstrates that Defendant faces a heightened risk.
To start, Defendant was involved with, and provided information about, members
of an interstate drug-trafficking organization and individuals committing home invasion
robberies. The information he provided was used in multiple investigations, as well as in a
federal sentencing. And Defendant spent a decade “[r]obbing high level drug dealers”—
individuals who, according to the district court, are more likely to “put a hit out on you”
than to “call the police” after a robbery. J.A. 44, 47.
Furthermore, Defendant notes that the district court’s reference to his cooperation
poses a threat to him because the reference was not to a mere willingness to cooperate or
to the fact of cooperation divorced from whether the court found it to be substantial. Rather,
in context, it is clear that Defendant (1) cooperated, (2) provided useful information, and
therefore (3) was rewarded for it with a reduced sentence—one significantly below his
original Guidelines range.
15
To be sure, the record does not reveal any direct threats to Defendant. But “direct
threats are not ‘a strict condition precedent to a district court’s granting of a closure
motion.’”
Doe, 870 F.3d at 999 (quoting
Doe, 63 F.3d at 130).
The Ninth Circuit’s Doe decision provides a useful comparator because, although
the record in that case did contain an explicit threat to the defendant’s family, there were
no such threats made to the defendant himself.
Id. Rather, like here, the record
demonstrated “why the risk to [the defendant] was heightened.”
Id. at 998 (internal
quotation marks omitted). In its substantial assistance memorandum, the government
explained that the defendant was involved with an international drug cartel before he lost
more than half a million dollars’ worth of methamphetamine.
Id. He had smuggled drugs
across the U.S.-Mexico border and distributed them across Southern California.
Id. Like
Defendant, the defendant in that case began cooperating with the government soon after
his arrest, implicating others who were also incarcerated.
Id. The Ninth Circuit concluded
that a direct threat was not necessary to establish a substantial probability of harm where
the record clearly revealed a heightened risk. See
id. at 999. We agree with that approach.
The E.D.N.C. Standing Order and the CACM Report add credibility to Defendant’s
fears. The Standing Order recognized that longstanding “concerns about the safety of
cooperating defendants . . . have been exacerbated by the subsequent development of
PACER and the growth of the internet.” E.D.N.C. Standing Order at 2. Indeed, the Eastern
District of North Carolina had received reports of “assaults, even torture, of [cooperating]
defendants by fellow inmates.”
Id. at 1. And the Eastern District noted that the problem
was not merely made worse by the fact that inmates could now readily search fellow
16
inmates’ dockets using PACER or online legal research tools. Rather, “such information is
often posted on internet web sites . . . and on social networking sites . . . for the apparent
purpose of encouraging reprisals against defendants who cooperate with authorities.”
Id.
Similarly, “[t]he Court has reviewed a video posted on the internet displaying the
cooperation provisions of defendants’ plea agreements and discussing substantial
assistance motions, apparently aimed at discouraging the cooperators from testifying.”
Id.
In 2010, therefore, the Eastern District adopted the Standing Order. The Standing
Order requires automatic sealing of substantial assistance motions for an extendable period
of two years. See
id. at 3–4. It also applies to “any responses to substantial assistance
motions, replies, orders granting or denying such motions, and any pleadings and orders
relating to the continuance of a hearing on a pending Rule 35 motion.”
Id. at 3 n.1. 11 The
Eastern District concluded these measures were necessary “to preserve the ‘higher value’
of preventing interference with the due administration of justice that results from reprisals
against witnesses.”
Id. at 4 (citation omitted) (quoting In re Wash.
Post, 807 F.2d at 390).
The Eastern District is by no means alone in taking steps to protect cooperators.
Id.
at 2 (discussing other districts’ policies). As of June 2016, more than one-third of federal
district courts had “adopted local rules or standing orders to make all criminal defendants
appear identical in the record to obscure cooperation information.” CACM Report at 3 n.3.
11
The rule is not, however, absolute. The Standing Order provides exceptions for where
the government is under a “case-related disclosure obligation” in another case or where
disclosure “is necessary to protect [the] . . . legal interests or safety” of “the defendant
whose assistance is the subject of the motion.”
Id. at 3. In those cases, the government,
defendant, or defense counsel may disclose the sealed document “without seeking a court
order.”
Id.
17
The CACM Report urged more courts to consider adopting methods to protect cooperators.
See
id. at 1, 3.
The conclusions of the CACM Report provide additional compelling data regarding
the risks to Defendant and other cooperators. We agree with the Ninth Circuit’s conclusion
that the Report “sheds considerable light on the dangers faced by those who agree to
cooperate with government
investigations.” 870 F.3d at 999.
First, the CACM Report explains that “[r]emote electronic access [has] dramatically
increased the potential for illicit use of case information regarding cooperators.” CACM
Report at 1–2. Indeed, the Report notes that “it is largely for this reason that the Judicial
Conference initially delayed public electronic access to criminal case files.”
Id. at 2.
Next, the CACM Report presents alarming findings, noting that “district judge
[survey] respondents reported 571 instances of harms or threats—physical or economic—
to defendants and witnesses between the spring of 2012 and the spring of 2015, including
31 murders of defendant cooperators.”
Id. Moreover, it reported “363 instances in which
court records were known by judges to [have been] used in the identification of
cooperators” and acknowledged “a particular problem in our prisons, where new inmates
are routinely required by other inmates to produce dockets or case documents in order to
prove whether or not they cooperated. If the new inmates refuse to produce the documents,
they are punished.”
Id. The Report noted the Committee’s belief that all of this was
indicative of “a systemic problem of court records being used in the mistreatment of
cooperators.”
Id. And “[b]ecause criminal case dockets are . . . compared in order to
identify cooperators, every criminal case is implicated.”
Id. at 3.
18
The CACM Report recognized judges’ duty “to vigilantly safeguard the public’s
right to access court documents and proceedings pursuant to the First Amendment and
under common law,” but noted that “there are many well-established restrictions on access
to criminal case information that address compelling government interests,” such as those
limiting access to presentence investigation reports.
Id. at 4 & n.7. The Report concluded
that “the harms to individuals and the administration of criminal justice in this instance are
so significant and ubiquitous” that action was necessary and justified, “consistent with each
court’s duty to exercise ‘supervisory power over its own records and files.’”
Id. at 4
(quoting Nixon v. Warner Commc’ns, Inc.,
435 U.S. 589, 598 (1978) (“It is uncontested . .
. that the right to inspect and copy judicial records is not absolute. Every court has
supervisory power over its own records and files, and access has been denied where court
files might have become a vehicle for improper purposes.”)).
The CACM Report thus issued recommendations regarding how courts handle plea
agreements, sentencing memoranda, transcripts of plea hearings and sentencings, Rule 35
motions, presentence reports, and judicial opinions involving cooperators—the last of
which, “where reasonably practicable, should avoid discussing or making any reference to
the fact of a defendant’s . . . cooperation.”
Id. at 9; see also
id. at 8–9.
In sum, the facts of Defendant’s case, the background facts found by the Eastern
District and the CACM Report, and the analysis by the Ninth Circuit in similar
circumstances in Doe lead us to conclude that Defendant faces a substantial risk of harm
in the absence of sealing. See
Doe, 870 F.3d at 999.
19
Seeking to overcome these facts, the government notes that Defendant’s sentencing
transcript—which has been available on the docket since September 2014—discusses his
cooperation. 12 Similarly, the district court’s text order noted that “[t]he court’s order has
been public since July 30, 2018,” implying that Defendant’s decision not to move to seal
it until November of that year suggested there was no real risk of harm. J.A. 14. 13 The
government argues that the fact of Defendant’s cooperation “has therefore lost its secret
character.” Response Br. at 12. (citing Va. Dep’t of State
Police, 386 F.3d at 579).
But the government’s cited authority is inapposite. In Virginia Department of State
Police, a police department sought to maintain a seal on certain records disclosing the
identity of a suspect in a brutal rape and murder—but it failed to move to keep under seal
other documents containing the same
information. 386 F.3d at 572–73. The suspect’s
identity became “a matter of public record” and was “extensively reported in the media.”
Id. at 573; see
id. at 578. The district court thus found the police department’s arguments
unpersuasive, and it unsealed the disputed records.
Id. at 573. We affirmed.
Id. at 580.
Unlike Virginia Department of State Police, here there is no reason to believe that
Defendant’s cooperation is “already a matter of public record.”
Id. at 573. Rather, the Ninth
Circuit’s decision in Doe is a more apt comparison. There, the Ninth Circuit rejected the
district court’s conclusion that “the § 5K1.1 departure ‘cat’ is typically out of the bag at a
defendant’s sentencing.”
Doe, 870 F.3d at 995. The Ninth Circuit noted that “no members
12
The government also notes that certain of Defendant’s pro se filings make reference to
his cooperation. We will not hold against Defendant that, proceeding pro se, he did not
move to seal his own filings.
13
However, the district court did not mention this fact in its written order.
20
of the media expressed interest in Doe’s sentencing, and no one has ever publicly sought
access to Doe’s court file or proceedings.”
Id. at 996. Similarly, here, although the fact of
Defendant’s cooperation has technically been available to the public, there is no suggestion
in the record that it has been reported in the media.
The government also raises practical considerations. It argues that ruling in
Defendant’s favor would make “every document in the entire federal criminal justice
system mentioning [§ 5K1.1] . . . subject to be sealed.” Response Br. at 10. Further,
“because [§ 5K1.1] motions are so common place,” that would cover “a tremendous
number of documents” and require “a huge undertaking” for district courts.
Id.
We are sensitive to the government’s concerns, but we find them unpersuasive. Cf.
Va. Dep’t of State
Police, 386 F.3d at 574 (rejecting the idea that “the district court’s
decision to unseal the documents will have a dramatic effect beyond this case” because
“the issue is actually quite narrow and . . . its resolution is dictated by well-established
precedent and the unique circumstances of this case”). To the extent that automatic sealing
applies to a specific document, the Eastern District has already concluded, after weighing
such arguments, that “[c]ase-by-case review would not work.” E.D.N.C. Standing Order at
4. 14 Where automatic sealing does not apply to a specific category of filings, courts have
14
No party has raised a facial challenge to the Standing Order, and accordingly its validity
is not before us. “But nothing in our precedent prevents district courts from adopting some
variation of the practices recommended by the [CACM] Report, as long as district courts
decide motions to seal or redact on a case-by-case basis.”
Doe, 870 F.3d at 1002; see also
Bacon, 950 F.3d at 1295 & n.7. As noted, the Standing Order already allows for disclosure
of sealed documents without a court order in certain circumstances. E.D.N.C. Standing
Order at 3. Moreover, where a party raises a question as to a sealing decision, the court
must make a determination considering the specific facts of the case. The court may, of
21
no choice but to engage in the analysis discussed herein upon motion by the defendant or
another interested party. But protecting cooperating defendants in federal custody is in the
interest of the government as well as the defendant. E.g., CACM Report at 7. If the
government finds it burdensome to litigate every sealing motion filed by a defendant, it
need not do so.
3.
Finally, although Defendant has established a substantial probability that a
compelling interest will be harmed absent sealing of the district court’s order, he must also
show that there are no adequate alternatives to sealing in this case. In re Wash.
Post, 807
F.2d at 392. To that end, Defendant contends that sealing the order, rather than redacting
it, “is the narrowest means for serving the interests at stake” in his appeal. Opening Br. at
12. He argues that redacting references to § 5K1.1 would only cause other inmates to
conclude that the removed material involved cooperation. Likewise, the Eastern District
has found that alternatives to automatic sealing “would not work” to protect inmates who
have provided substantial assistance to the government. E.D.N.C. Standing Order at 4.
Common sense confirms this conclusion. Although this Court has used redactions
to protect the identities of a defendant’s family members, e.g.,
Harris, 890 F.3d at 492,
redaction cannot protect Defendant himself because redacting the order would merely “flag
the filings in his case. The publicly available and remotely accessible filings would
immediately look different from the filings in non-cooperators’ cases.” Doe, 870 F.3d at
course, take into consideration the background facts that led to the adoption of any relevant
standing order in the first place.
22
1001. For that reason, the CACM Report recommends that “courts restructure their
practices so that documents or transcripts that typically contain cooperation information”
are automatically placed in a sealed supplement. CACM Report at 3. We agree with
Defendant that sealing the order at issue is the narrowest means of protecting the
compelling interest in this case.
IV.
Because sealing the order referring to Defendant’s substantial assistance is
necessary to protect Defendant from harm, we reverse the district court’s decision denying
his motion to seal. On remand, the district court shall seal its order denying Defendant’s §
3582 motion for at least two years from the date the court enters its sealing order. 15
Of course, our opinion does not require other districts to implement the sort of
automatic sealing procedures recommended by the CACM Report or adopted by the
Eastern District of North Carolina. Courts can and should continue to consider possible
alternative responses to the violence cooperators and their families face in federal prisons
and elsewhere. But where, as here, the risks facing cooperators have already led a district
to take blanket measures, courts within the district should act consistently with the concerns
underlying that policy.
REVERSED AND REMANDED
15
We recognize that, as a general matter, it is preferable for district courts to make sealing
determinations in the first instance. See Va. Dep’t of State
Police, 386 F.3d at 581.
However, given the ongoing safety issues at stake, we find it appropriate in this instance to
remand with directions. See
Doe, 870 F.3d at 1002.
23
RICHARDSON, Circuit Judge, dissenting:
The public enjoys a weighty right of access to the criminal courts that is secured by
the First Amendment and the common-law tradition. See Doe v. Public Citizen,
749 F.3d
246, 265–69 (4th Cir. 2014); see also Richmond Newspapers, Inc. v. Virginia,
488 U.S.
555, 563–81 (1980) (plurality opinion) (Burger, J.). Even so, I wholeheartedly agree with
the majority that the protection of defendants and their families—not to mention the need
to safeguard ongoing investigations—generally presents a compelling interest that should,
in most cases, overcome the public’s right of access and permit sealing information related
to a defendant’s cooperation with law enforcement. But that interest alone cannot justify
a categorical rule that demands sealing without regard to the specifics of a given case. See
Press-Enterprise Co. v. Superior Court of California for Riverside County,
464 U.S. 501,
509 (1984); see also In re Knight Publishing Co.,
743 F.2d 231, 234–35 (4th Cir. 1984). 1
And here, the specifics lead me to conclude that the district court appropriately declined to
seal an order that referenced the Defendant’s cooperation.
Whether we focus on when the exercise of the inherent judicial Power (or something
else) requires sealing or when the First Amendment and the common law merely permit it,
the Defendant’s motion is fatally underinclusive—failing to satisfy a requirement shared
1
The district court’s standing order governing motions addressing cooperation is
not at issue here. So while a categorical rule barring consideration of a case’s specifics
might well raise concerns, compare Globe Newspaper Co. v. Superior Ct. for Norfolk Cty.,
457 U.S. 596, 607–08 (1982) (state statute that categorically required closed courtrooms
during minor victim testimony failed to consider the circumstances of a particular case),
with Am. Civ. Liberties Union v. Holder,
673 F.3d 245, 254 (4th Cir. 2011) (initial 60-day
sealing period for False Claim Act claims permissible), we need not address them here.
24
by the varied points in our somewhat Delphic doctrinal constellation. 2 That is because the
relief it seeks fails to protect whatever “secret characteristic” of the information remains
after all these years. Virginia Dep’t of State Police v. Washington Post,
386 F.3d 567, 579
(4th Cir. 2004) (internal quotations omitted) (First Amendment); see In re Knight
Publishing
Co., 743 F.2d at 235 (common law). Even after today’s order granting the
motion (which seals one document for two years in the middle of the Defendant’s long
prison term), the Defendant’s cooperation remains detailed in other publicly available court
records. Indeed, to find a substantial probability of harm absent sealing, the majority
itself—with a straight face—relies on that same publicly available information to bolster
the Defendant’s mere “conclusory statements.” Majority Op. 15; see also
id. at 3. That
information, more specific than the passing reference at issue today, has long been
accessible to the public and will remain so. Compare United States v. Doe,
870 F.3d 991,
994 (9th Cir. 2017) (addressing motion “to seal all documents related to, or disclosing the
existence of,” the defendant’s cooperation), with Majority Op. 23 (leaving documents
referencing the cooperation in the public domain).
And so, with great respect for the Majority’s desire to guard against the real and
substantial harms that may result from disclosing law-enforcement cooperation, I would
affirm the district court based on the specifics of this case.
2
This underinclusiveness obviates the need to determine the floors and ceilings
presented by the common law, the First Amendment, and the exercise of the judicial Power.
And it permits us to avoid the more complex twists and turns that are raised when we are
asked to review—instead of the conventional decision to grant a motion to seal (with a
challenge asserting a First Amendment and common law right of access)—a denial of a
motion to seal (with a challenge asserting, apparently, some other type of right).
25