Courthouse News Serv. v. Corsones

131 F.4th 59

Opinion

21-3098 Courthouse News Serv. v. Corsones

UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT

August Term, 2022

(Argued: April 10, 2023 Decided: March 11, 2025)

Docket No. 21-3098

_____________________________________

COURTHOUSE NEWS SERVICE; VERMONT PRESS ASSOCIATION, INC.; NEW ENGLAND FIRST AMENDMENT COALITION; GRAY MEDIA GROUP, INC., DBA WCAX-TV; GANNETT VERMONT PUBLISHING, INC., DBA Burlington Free Press;

SAMPLE NEWS GROUP, LLC, DBA Barre-Montpelier Times Argus, DBA Rutland Herald; VERMONT JOURNALISM TRUST, LTD.; VTDIGGER, a project of Vermont Journalism Trust, Ltd.; DA CAPO PUBLISHING, INC., DBA Seven Days; VERMONT COMMUNITY NEWSPAPER GROUP, LLC, DBA Stowe Reporter, News & Citizen, South Burlington Other Paper, Shelburne News, and The Citizen,

Plaintiff-Appellees,

v.

TERI CORSONES, in her official capacity as the State Court Administrator of the Supreme Court of the State of Vermont; AMANDA STITES, in her official capacity as Clerk of Court for Addison, Bennington, and Rutland Counties;

MARGARET VILLENEUVE, in her official capacity as Clerk of Court for Caledonia, Essex, Orleans, and Washington Counties; CHRISTINE BROCK, in her official capacity as Clerk of Court for Chittenden County; GAYE PAQUETTE, in her official capacity as Clerk of Court for Franklin, Grand Isle, and Lamoille Counties; ANNE DAMONE, in her official capacity as Clerk of Court for Orange, Windham, and Windsor Counties,

Defendant-Appellants. ∗

∗ The Clerk of Court is respectfully directed to amend the official case caption as set forth above.

_____________________________________

Before: LEVAL, CHIN, and SULLIVAN, Circuit Judges.

Defendants, administrators and clerks of the Vermont Superior Court, appeal from the judgment and permanent injunction of the United States District Court for the District of Vermont (Christina Reiss, J.) in favor of Plaintiffs, news and related media organizations. Beginning in 2020, when Vermont courts transitioned to electronic filing, the Superior Court adopted a policy of denying public access to newly filed civil complaints until a court clerk had reviewed them to ensure that they were signed, they did not contain unredacted confidential information, that they complied with technical requirements under the court’s rules, and that they did not show unredacted filers’ notes. Plaintiffs sued, challenging this practice as a violation of their First Amendment right of access to court documents. After a bench trial, the district court issued judgment in Plaintiffs’ favor, holding that Vermont’s pre-access review process violated Plaintiffs’ First Amendment right of access to judicial documents, and issued a permanent injunction barring Defendants from withholding complaints until completion of a pre- access review. We agree with the Plaintiffs and the district court that Vermont’s practice in the period reviewed by the trial court violated Plaintiffs’ First Amendment right of access, but we agree with the Defendants that the terms of the permanent injunction were not supported by the court’s findings. We accordingly AFFIRM the district court’s judgment to the extent it found that the practice it reviewed violated the First Amendment, but we VACATE the permanent injunction to the extent that it barred the Defendants from engaging in any review for unredacted confidential information before permitting access to the complaints. We REMAND for further proceedings.

JUDGE SULLIVAN concurs in part and dissents in part in a separate opinion.

JONATHAN E. GINSBERG (William J.

Hibsher, Glenn B. Coleman, on the brief), Bryan Cave Leighton Paisner LLP, New York, NY, for Plaintiff-Appellees.

DAVID BOYD, Assistant Attorney General, Office of the Attorney General, Montpelier, VT, for Defendant-Appellants.

Nolan L. Reichl, Peter J. Guffin, Ariel A. Pardee, Laura M. O’Hanlon, Pierce Atwood LLP, Portland, ME, for Amicus Curiae Conference of Chief Justices in support of Defendant-Appellants.

Katie Townsend, Bruce D. Brown, The Reporters Committee for Freedom of the Press, Washington, DC, for Amici Curiae The Reporters Committee for Freedom of the Press and Twenty-Eight Media Organizations in Support of Plaintiff-Appellees.

LEVAL, Circuit Judge:

Defendants, administrators and clerks of the Vermont Superior Court (“the Superior Court”) (we sometimes refer to Defendants collectively as “Vermont”), 1 appeal from the judgment of the United States District Court for the District of Vermont (Christina Reiss, J.) in favor of Plaintiffs, including a permanent injunction. The Plaintiffs, led by Courthouse News Service (“CNS”), are primarily press entities that cover the work of the Vermont courts. 2 In 2020, the Superior Court transitioned to receiving filed documents in electronic form. At that time, it adopted a policy of delaying the release to the public of newly filed civil complaints until a court clerk had reviewed them to ensure that they did not contain unredacted confidential information that might be misused to commit fraud and the like, that they were signed and complied with technical requirements under court rules, and that they did not reveal filers’ notes. This pre-access review process caused delays in making

1 The Defendants are Teri Corsones, Amanda Stites, Margaret Villeneuve, Christine Brock, Gaye Paquette, and Anne Damone. 2 The Plaintiffs also include a voluntary association that promotes the interests of the press.

Plaintiffs are: Courthouse News Service; Vermont Press Association, Inc.; New England First Amendment Coalition; Gray Media Group, Inc.; Gannett Vermont Publishing, Inc.; Sample News Group, LLC; Vermont Journalism Trust, Ltd., VTDigger; Da Capo Publishing, Inc.; and Vermont Community Newspaper Group, LLC.

5

complaints publicly available. Plaintiffs brought this suit, alleging that these delays in releasing complaints to the public violated the Plaintiffs’ First Amendment right of access to judicial documents. After a bench trial, the district court ruled that Vermont’s pre-access review process violated Plaintiffs’ First Amendment right and issued a permanent injunction barring Defendants from delaying release to the public of newly filed complaints pending review by court staff.

We agree with the Plaintiffs and the district court that Vermont’s practice, as reflected in the trial evidence, violated Plaintiffs’ First Amendment right of access to judicial documents. We therefore affirm the judgment to that extent. Nonetheless, we agree with the Defendants that the terms of the relief granted to the Plaintiffs, forbidding any and all review of complaints before releasing them to the public, were not justified by the evidence or the applicable law and were therefore not within the district court’s permissible discretion. We therefore vacate the injunction and remand for further proceedings, including reconsideration of the terms of an injunction.

6

BACKGROUND 3

The Superior Court is the basic trial court of the State. It functions through 14 units, one for each county of the State. 4 Reporters employed by the media Plaintiffs review new complaints filed in the Superior Court to identify newsworthy cases, on which the Plaintiffs then distribute news reports. The complaint alleges that the Plaintiffs and the public are harmed by delays imposed on them resulting from the Superior Court’s policy of delaying the release of new complaints.

I. The Superior Court’s Transition to Electronic Filing and Pre-Access Review Process

Prior to March 2020, the Superior Court did not accept electronic filings. Members of the public and the media could review newly filed paper complaints in person in courthouses. When someone sought access to a complaint, court staff would conduct a “quick file audit” to confirm the absence of confidential information and would remove any confidential information from the file before allowing access. 5

3 Unless otherwise indicated, facts are drawn from the district court’s findings of fact, Courthouse News Serv. v. Gabel, No. 21-CV-000132, 2021 WL 5416650, at *2–8 (D. Vt. Nov. 19, 2021), which Defendants do not challenge on appeal. 4 See Court Divisions, VT. JUDICIARY, https://www.vermontjudiciary.org/court-divisions (last

visited July 7, 2023). 5 In some cases, complaints were made publicly available before they were docketed.

7

Beginning in March 2020, the procedures changed as the Vermont judiciary transitioned to an electronic case management system known as Odyssey. By March 15, 2021, virtually all documents filed in the Superior Court were required to be filed electronically. Generally, the press and public can review electronically filed documents only at designated display terminals located in courthouses and judiciary offices during regular business hours, though the Court Administrator has the discretion to make certain civil case records accessible remotely.

A. Pre-Access Review Process In anticipation of the transition to electronic filing, the State adopted the 2020 Vermont Rules for Electronic Filing and amended the Vermont Rules for Public Access to Court Records, which together govern electronic filings in the Superior Court.

Pursuant to the pre-access review process adopted under these new rules, before allowing public access to a new complaint, a court clerk would check it for any unredacted confidential personal or financial information exempt from disclosure, such as social security numbers or financial account numbers, and also would verify that the complaint included a signature, that

8

the right filing codes, filing fee, and case type were designated, and that there were no comments left by the filer.

Under Vermont’s rules, the filer of a complaint bears primary responsibility for protecting confidential information from public disclosure and must certify to having verified that the document does not contain any such information. Vt. R. Pub. Access Ct. Recs. 7(a)(1)(A)–(B). If a court clerk finds that a filing is not compliant with the public disclosure rules, she may withhold public access, redact the filing, or reject it entirely. Vt. R. Pub. Access Ct. Recs. 7(a)(4)(A). She may also refer the matter to a judge who may impose sanctions or refer the matter to the State’s Professional Responsibility Board for disciplinary review. Vt. R. Pub. Access Ct. Recs. 7(a)(4)(B). This system appears to have had considerable success in guarding, through redaction or otherwise, against release of inappropriate information. The district court noted that Superior Court staff rejected only three exhibits related to two complaints, out of 4,156 during the relevant period on account of revelation of confidential information—a rejection rate of 0.048%. However, across all types of filings (not limited to complaints), 138 filings were rejected for revelation of inappropriate information.

9

B. Delays

Trial began on October 25, 2021 and concluded on October 26, 2021.

The district court reviewed the delays resulting from Vermont’s pre-access review process in releasing to the public the 4,156 complaints filed in the Superior Court during the approximately 16 months between the units’ transition to electronic filing 6 and August 6, 2021 (the last day for which data was produced prior to trial). Those delays were as follows:

54.8% of complaints were made available to the public on the day of filing;

22.6% were made available on the day after filing;

4.6% were made available two days after filing;

6.7% were made available three days after filing; and 11.4% were made available four or more days after filing.

The delays were longer in some county units of the Superior Court than others: for example, over the entire period, the percentage of newly filed civil complaints made available on the day they were filed ranged from 3.8% in

6The units for Orange, Windham, and Windsor Counties began accepting electronic filings in April 2020; the units for Addison, Bennington, Chittenden, and Rutland Counties in October 2020; and the units for Caledonia, Essex, Franklin, Grand Isle, Lamoille, Orleans, and Washington Counties in March 2021.

10

Essex County to 81.6% in Windham County. And delays fluctuated from week to week: for some weeks, no newly filed civil complaints were made available the day they were filed, while in one week, 91.7% were made available the same day as filing.

During the transition to Odyssey, the Superior Court faced staffing and equipment challenges engendered by the COVID-19 pandemic, and it continued to experience high employee attrition rates and hiring difficulties, which are likely partially to blame for the delays.

C. Centralization Starting on July 12, 2021, Vermont began a pilot program to centralize the pre-access review of civil cases. In lieu of each county unit’s own staff reviewing new complaints filed in that court, a centralized team would review all complaints filed in the Superior Court statewide. Between July 12 and July 19, 2021, five units 7 transitioned to the centralized review team. The

7The Chittenden, Essex, and Caledonia units shifted to the centralized review team on July 12, 2021, and the Rutland and Orleans units were then added on July 19, 2021.

11

centralization process was still in its early stages when the trial evidence closed. 8 Early results of centralization were promising. In a supplemental post-

trial submission, Defendants’ expert stated that, between July 26, 2021 (two weeks after the central review team was established) and September 26, 2021, 9 although only partial conversion to centralization had been accomplished, 67% of complaints were made available the day they were filed and 95% were made available within one business day, see App’x 487, 505, as compared to 77.4% released within a day in the period of the trial evidence, see Gabel, 2021 WL 5416650, at *7, an improvement of nearly 23%. 10 The submission did not reveal when the remaining 5% of complaints were made public.

8 While it is unclear whether the trial evidence of release delays in any degree reflected the new consolidated procedures, it is clear that the new procedures had at most a small effect on those statistics, as the consolidated procedures were still in their infant stages and had operated only briefly when the trial evidence closed. 9 Of the 768 complaints covered by the Defendants’ post-trial submission, approximately

75% were new data not included in the trial evidence. App’x 505. 10 Defendants’ expert reported “the percentage of initial filings that were reviewed on the

same day or workday after the submitted date.” App’x 493. Defendants treat this as the percentage of complaints made public within one business day. See Appellants’ Br. at 25–26. Plaintiffs do not dispute this. They do note that the “next business day” metric means that a Friday filing would not be available until Monday and that the 95% figure represents an average that does not reflect day-to-day and court-to-court variations. See Appellees’ Br. at 47 n.25.

12

II. This Litigation Plaintiffs filed this suit on May 20, 2021, alleging that the Defendants’

policy of withholding newly filed civil complaints from public access until Defendants completed their pre-access review process violated Plaintiffs’ First Amendment rights. In November 2021, after a bench trial, the district court ruled that the policy infringed on Plaintiffs’ First Amendment right of access to judicial documents and enjoined Defendants “from prohibiting public access to newly filed civil complaints which have not been designated confidential by the filer until the Vermont Superior Court has completed a pre-access review process.” 11 Gabel, 2021 WL 5416650, at *18. Defendants appeal.

STANDARD OF REVIEW

On appeal from a judgment issued after a bench trial, we review the district court’s findings of fact for clear error and its conclusions of law de novo. Citibank, N.A. v. Brigade Cap. Mgmt., LP, 49 F.4th 42, 58 (2d Cir. 2022). We

11The district court denied Plaintiffs’ request for declaratory relief, finding it “would serve no purpose which the court’s permanent injunction has not achieved.” Gabel, 2021 WL 5416650, at *18.

13

review the district court’s issuance of a permanent injunction for abuse of discretion, which is deemed to include application of incorrect law. Springfield Hosp., Inc. v. Guzman, 28 F.4th 403, 415 (2d Cir. 2022).

DISCUSSION

I. The First Amendment Right of Access As detailed in Press-Enterprise Co. v. Superior Court of California for Riverside County, 478 U.S. 1 (1986) (“Press-Enterprise II”), Plaintiffs have a presumptive First Amendment right of access to newly filed complaints, which attaches upon a court’s receipt of such complaints. This does not mean that the public or the press has an unconditional right of instantaneous access. The presumption of access that attaches upon a court’s receipt of a filing merely requires that delays in granting the public access to that filing be persuasively justified by the party causing the delay. With respect to the period covered by the trial evidence, we affirm the district court’s finding that Defendants did not meet this burden.

14

A. Framework for Determining Whether Plaintiffs Have a First Amendment Right of Access to New Civil Complaints

The Supreme Court outlined the framework for evaluating right-of-

access cases in Press-Enterprise II. In that case, the Court assessed whether a First Amendment right of access applied to the transcript of a preliminary hearing relating to a criminal prosecution. To answer this question, the Court applied a two-step framework. The same test applies to the public’s right of access to a complaint.

First, to determine whether there exists a First Amendment right of access to a document or proceeding, a court begins by applying what has become known as the “experience and logic” test. Under this test, a court must ask “whether the place and process have historically been open to the press and general public” (the experience prong) and “whether public access plays a significant positive role in the functioning of the particular process in question” (the logic prong). Id. at 8. If the proceeding or document in question passes both prongs of the test, there exists a qualified, presumptive First Amendment right of public access to the proceeding or document. Id. at 9.

Second, if such a presumptive right exists, the proponent of closure must establish through “specific, on the record findings . . . that ‘closure is

15

essential to preserve higher values and is narrowly tailored to serve that interest.’” Id. at 13–14 (quoting Press-Enter. v. Super. Ct. of Cal. for Riverside Cnty., 464 U.S. 501, 510 (1984) (“Press-Enterprise I”)).

In Bernstein v. Bernstein Litowitz Berger & Grossmann LLP, we applied the “experience and logic” test to determine that the First Amendment provides a presumptive right of access to civil complaints. 814 F.3d 132, 141 (2d Cir. 2016). We first noted that “[c]omplaints have historically been publicly accessible by default, even when they contain arguably sensitive information.” Id. We concluded that this history satisfied the experience prong of the “experience and logic” test. Turning then to the logic prong, we concluded that logical considerations amply supported public access to complaints because such access “allows the public to understand the activity of the federal courts, enhances the court system’s accountability and legitimacy, and informs the public of matters of public concern.” Id. Because both prongs of the test were met, we held that there existed a presumptive First Amendment right of access to complaints.

In Lugosch v. Pyramid Co. of Onondaga, we made clear that this presumption is of “immediate public access.” 435 F.3d 110, 126 (2d Cir. 2006)

16

(discussing documents filed in support of or in opposition to a motion for summary judgment). For the right of access to justify its purpose, it must be a right of timely access. Delays of access to complaints can defeat the purposes of disclosure recognized in Bernstein, as courts are less likely to be held accountable and the public is less likely to be informed when access to complaints is substantially delayed. See id. at 127.

These precedents, however, do not establish an unconditional First Amendment right to instantaneously access any newly filed complaint prior to any human review. Determination whether the presumptive right of access matures into an actual right of access depends on whether the party imposing delays succeeds in showing justification under the second step of the Press- Enterprise II test. Under this second step, the entity seeking to withhold or delay public access must establish that its reason for doing so “is essential to preserve higher values and is narrowly tailored to serve that interest,” Press- Enterprise II, 478 U.S. at 13–14 (quoting Press-Enterprise I, 464 U.S. at 510), a conclusion that must be justified by “specific, on the record findings,” id. at 13.

17

B. Whether Vermont’s Pre-Access Review Process Violated Plaintiffs’ First Amendment Rights

We find no error in the district court’s determination that, with respect to the period examined in the trial evidence, the First Amendment gave Plaintiffs a right of access to complaints newly filed in the Superior Court and that the Defendants’ pre-access review process in that period violated that right. Our court in Bernstein has already established that the “experience and logic” test of Press-Enterprise II provides the public with a presumptive right of access to civil complaints. The press, as the agency through which the public’s right of access is usually realized, enjoys the same right. See Lugosch, 435 F.3d at 120; Courthouse News Serv. v. Schaefer, 2 F.4th 318, 326 n.5 (4th Cir. 2021) (“The media’s rights of access are ‘co-extensive with and do not exceed those rights of members of the public in general.’” (quoting In re Greensboro News Co., 727 F.2d 1320, 1322 (4th Cir. 1984))). In Lugosch, we further ruled that the presumptive right of public access to a judicial document attaches at the time the document is filed, 435 F.3d at 126, thus placing the burden on a party that would withhold access to justify doing so under the standards of

18

Press-Enterprise II, id. at 124. 12 The question is whether the Defendants satisfied their burden of demonstrating, as required by the second step of Press-Enterprise II, that their review process was (i) “essential to preserve higher values” and (ii) “narrowly tailored to serve” those higher values. 478 U.S. at 13–14.

The first of those questions is whether the Defendants’ pre-access review policy was essential to serve a higher value. This of course can depend on which of the Defendants’ several objectives in delaying disclosure is evaluated. Press-Enterprise II does not explain what it meant by “higher value.” The Supreme Court has never defined the term and, in subsequent cases, has used it interchangeably with “overriding interest.” Waller v. Georgia, 467 U.S. 39, 48 (1984). ”Higher value” could mean simply an elevated value — one that contributes a substantial benefit. More likely the use of the comparative implies a comparison — higher than what? Any delay engenders

12Defendants argue that Plaintiffs are seeking to assert a novel “instantaneous access right” and should therefore establish that such a right satisfies the “experience and logic” test. See Appellants’ Br. at 30–31. The argument misunderstands the Press-Enterprise II framework. The purpose of its first test, the experience and logic test, is to determine whether there exists a presumptive First Amendment right of access to a particular material or proceeding. See Newsday LLC v. Cnty. of Nassau, 730 F.3d 156, 164 (2d Cir. 2013). If that test yields an affirmative answer, as here, its second test, inquiring into the value sought to be protected through the delay and whether the delay is narrowly tailored to protect that value, places the burden on the party that delays access to justify the delay.

19

a conflict between the value sought through delay or denial of disclosure and the value inhering in the public’s right of access. This suggests that the comparative calls for comparison of the benefit sought through the delay and the harm to the First Amendment interest in disclosure resulting from the delay imposed. The greater the benefit achieved and the lesser the delay caused, the more likely that the value preserved is a “higher value.” By the same token, the lesser the social value of what is achieved by imposing the delay and the greater the harm to First Amendment interests resulting from the delay, the weaker the argument that the delay preserved a “higher value.”

Courts have found that higher values include protecting the confidentiality of grand jury proceedings, 13 protecting minor victims of sex crimes, 14 protecting a defendant’s Sixth Amendment right to a fair trial, 15 protecting significant and substantial privacy interests, such as the physical safety of litigants, witnesses, or third parties, 16 preventing danger to persons

13 In re Grand Jury Subpoena, 103 F.3d 234, 242–43 (2d Cir. 1996). 14 Globe Newspaper Co. v. Super. Ct. for Cnty. Norfolk, 457 U.S. 596, 607 (1982). 15 ABC, Inc. v. Stewart, 360 F.3d 90, 100 (2d Cir. 2004).

16 United States v. Doe, 63 F.3d 121, 127, 129–30 (2d Cir. 1995). In Doe, the objective was to

protect the physical safety of the defendant, whose cooperation with the government against organized crime defendants would be revealed by trial evidence. The inference can perhaps be drawn from this and other precedents that other significant and substantial privacy interests might also qualify for protection as higher values.

20

or property, 17 and maintaining “the integrity of significant activities entitled to confidentiality, such as ongoing undercover investigations or detection devices.” 18 The principal benefit sought through the delay in this case is the protection against harms that can result from disclosure of confidential information. Civil complaints, and the attachments to them, can contain personal information about parties or other persons, such as social security numbers and financial account numbers. The disclosure of such information to the public creates a risk of misuse that can cause serious harm through frauds, identity thefts, and other malicious mischief. The undertaking by Vermont to prevent public disclosure of such confidential information so as to avoid such abuses, unquestionably serves a substantial value. Cf. United States v. Doe, 63 F.3d 121, 127 (2d Cir. 1995) (“[T]he privacy interests of individuals may also warrant [courtroom] closure orders in certain circumstances.”). Misuse of such confidential identifying information can bring ruin and misery not only to litigants but to persons who are strangers to the litigation and

17 In re Application of The Herald Co., 734 F.2d 93, 100 (2d Cir. 1984). 18 Id.

21

therefore have no way to control the court’s publication of documents containing such information about them.

We have no doubt that guarding against such abuses by redacting sensitive confidential information is a high value. Whether the value is not only high, but a “higher value” requires a comparison of intangibles that is difficult to make. 19 And different persons might reach different conclusions based on the weight they personally accord to different interests. Given the enormity of the harms that can easily result from disclosures of sensitive information, we recognize that Vermont has strong arguments to support its contention that guarding against such disclosures does protect a “higher

19The district court questioned the importance of the objective of pre-access review in observing that reliance on filers to eliminate or redact confidential information had been “overwhelmingly effective,“ so that of the complaints reviewed in the trial evidence, only two were found to contain unredacted confidential information. Gabel, 2021 WL 5416650, at *15–16. We do not entirely agree. Reliance on scrutiny by the filers was not as successful, and the problem of improper revelations not as trivial, as the district court suggested. While only two complaints were found to be problematic, 138 filings across all filing types were rejected during the relevant period because they contained confidential information, which suggests that reliance on filers was not as successful as the district court suggested. A single disclosure, furthermore, can cause enormously destructive consequences for the individual whose information is made public. See Erika Harrell, U.S. DEP’T OF JUST., BUREAU OF JUST. STAT., Victims of Identity Theft, 2018, at 10, 12 (2021). We have no doubt that Vermont’s efforts to prevent such harms served a high and substantial value.

22

value” and thus satisfies the first prong of the second Press-Enterprise II test. 20 However, we have no need to reach a definitive answer to that question at this time because, in any event, even assuming that this would qualify as a higher value, Vermont has failed to satisfy the second prong of the second Press-Enterprise II test.

The second prong of the second step obligates the Defendants to demonstrate that the pre-access review process was “narrowly tailored” to achieve that higher value without unduly imposing on the Plaintiffs’ presumptive right of access. Even if the delay is in service of a higher value, it nonetheless may fail to satisfy the “narrowly tailored” requirement if, for example, the process could reasonably be designed to impose significantly less delay. The proponent of delay is not required to show that it has reduced the

20In contrast, to the extent that Defendants devote time in pre-access review procedures to determining whether complaints have been properly signed and are otherwise in proper form or reveal filers’ notes, we doubt that this undertaking qualifies as a “higher value.” It is not that conforming to such technical requirements is not important. The question is rather whether this checking needs to be done prior to releasing a submitted complaint to the public. Assuming that failure to sign the complaint and put it into the required form would disqualify it from serving as a valid complaint, we see no reason why the invalidation of the complaint for such reasons could not be done after its release to the public. It does not appear that any harm to Vermont’s interests would result from the release to the public of complaints submitted for filing because they later are determined to be ineligible for filing on account of technical defects. And if a filer has negligently left notes in the document upon filing, we are not persuaded that the Superior Court has any significant interest in protecting the negligent filer from any embarrassment or disadvantage that might result from their revelation. This issue is further discussed below.

23

delay to the minimum possible. Perfection is not required. The demand for narrow tailoring is a reasonable demand. Vermont has failed to demonstrate that the delays it imposed during the trial evidence period could not have been reasonably shortened to a significant degree without impairing the higher value sought to be protected. The Defendants’ showing was deficient in a number of ways.

(i) First, it is not as if disclosure was delayed only to ensure the absence of confidential information. It is undisputed that the disclosure of complaints was also delayed while Defendants verified whether the complaints were signed and conformed to technical legal requirements and whether filers’ notes remained visible. See supra p. 22 n.20. While a complaint’s conformity with Vermont’s technical requirements is undoubtedly important, and certain complaints may well have been justifiably vulnerable to dismissal or rejection by reason of failures to conform to those requirements, those considerations do not in any way justify withholding the complaints from visibility to the press and the public. Revelation to the public of a complaint later adjudged to be unsuitable for filing because of its failure to conform to technical requirements would not cause harms of the sort that could justify the

24

Defendants in delaying the disclosure. Notwithstanding the undoubted importance of ensuring that complaints conform to legal requirements, the Defendants have failed to show that delaying their disclosure until legal conformity has been verified serves a higher value. This interference with the public’s right of access is unnecessary to ensuring that technical legal requirements will be met. Curative measures for failure to conform to such technical requirements can perfectly well be taken subsequent to public disclosure of the complaints without any harms ensuing from the earlier disclosure. A system (such as Vermont’s) designed to delay disclosure of filed complaints until they had been checked by clerk’s office personnel for compliance with technical requirements was not “narrowly tailored” to preserve higher values without undue interference with First Amendment rights.

The Defendants also delayed disclosure of complaints to check whether the filers of the complaints had inadvertently left notes not constituting part of the complaint visible. Delaying disclosure to protect sloppy filers from whatever disadvantages or embarrassments they might suffer resulting from their sloppiness in failing to ensure that they had left no confidential notes

25

visible before filing does not serve a higher value. One may wonder in what circumstances it is the job of the State to protect filers of complaints from their own carelessness in disclosing matters that they ought to have deleted, but that is not our question. More to the point, at least unless the carelessly disclosed matter risks causing serious harms, granting such protection to careless filers of complaints does not serve a higher value than the press’s entitlement to reasonably prompt disclosure. A system that delays public disclosure of complaints while state officials check complaints against such inadvertent revelation in filers’ notes (unless those notes reveal confidential information of the protected class) is not “narrowly tailored” to protect higher values, as required by Press-Enterprise II.

(ii) The next consideration that demonstrates the Defendants’ failure to satisfy the narrowly tailored requirement calls for focus on the reasons for some of the delays imposed by Vermont. The Defendants’ arguments have stressed how rapidly they released a large percentage of the filed complaints. Their argument is essentially, assuming that checking for improper revelation of confidential information appropriately serves a higher value, for some complaints, for example long complex complaints about financial matters that

26

are loaded with financial information, the job will be more difficult and time consuming than for others, so that it is reasonable for the disclosure of such complaints to be more delayed. Court clerks furthermore can do only one task at a time, so that dealing with some complaints must wait while clerks verify others. In some circumstances those might be persuasive arguments. But it is evident here that some of the delays came from other causes.

During the evidence period, each county of Vermont handled its complaints independently. Some counties cleared their complaints far more rapidly than others. For example, while Windham County released 81.6% of its complaints on the day of filing, Essex County released only 3.8% of its complaints on the day of filing. If there are locality-based explanations for such disparities, the Defendants have not advanced them. Such statistical disparities suggest that at least a part of the delays in releasing complaints was not because some complaints were more difficult to check than others or because some necessarily waited while the court clerks checked others. These statistics suggest that at least some of the delays were attributable to some of Vermont’s courts not giving the same priority, or the same efficiency, as other county courts to achieving reasonably prompt disclosure. Bringing the

27

laggard counties up to, or at least closer to, the speed of the more efficient counties is not asking that they perform at maximum possible speed. To the contrary, the statistics suggest that, if the complaints of certain Vermont counties were handled with greater recognition of the First Amendment obligation to achieve a reasonably prompt public disclosure, Vermont will have come a lot closer to achieving its constitutional obligation.

(iii) Next, a point that goes hand in hand with the county-by-county disparity in speed of performance. The evidence that the Defendants presented to the district court in a post-trial supplemental submission showed that, subsequent to the evidence period, Vermont abandoned the county-by- county system for pre-disclosure review of complaints, adopting instead a centralized procedure for pre-disclosure review. Although the conversion was only partially accomplished (with substantial further centralization still to be achieved), the Defendants already substantially improved their speed and efficiency, significantly reducing delays in the release of complaints, as compared with the period represented by the trial evidence. While during the trial evidence period, 77.4% of Vermont’s complaints were released within one business day after filing, by the time of the post-trial supplemental filing

28

that number had increased to 95%. See Gabel, 2021 WL 5416650, at *7. Although we recognize the possibility that the improvement might have been attributable to time-specific factors, the Defendants did not make a persuasive demonstration in their post-trial offer of these statistics that this was so. The improved numbers give some likely support to the proposition that what was accomplished through procedural changes made just after the trial evidence period could reasonably have been accomplished during the trial evidence period. Defendants have not disputed that they were reasonably capable of accomplishing pre-access review with considerably more speed than as experienced in the trial evidence period.

(iv) Finally, we note the incompleteness of the Defendants’ showing.

While they emphasize the brevity of the time lapses by which they released 89% of the complaints filed, as for the remaining 11%, the only information given by the Defendants was that these were released four or more days after filing. Four or more days is open-ended. It gives no information about how long this 11% was delayed. Because Vermont has not shown how long were the incremental delays attributable to the slowest 11%, we cannot conclude that this delay was too insignificant to be of constitutional significance.

29

All of these reasons contribute to our conclusion that the Defendants failed to satisfy their burden of demonstrating that the review processes considered at trial were “narrowly tailored” to achieve their justifying purposes.

C. Whether Vermont’s Practice Is Best Evaluated as a “Time, Place, or Manner” Restriction

Defendants argue that, while outright denials of access are evaluated under Press-Enterprise II standards, which they liken to strict scrutiny, restrictions that resemble “time, place, or manner” restrictions should be reviewed under intermediate scrutiny. They base their argument in substantial part on dictum in a footnote in the Supreme Court’s opinion in Globe Newspaper Co. v. Superior Court for the County of Norfolk, 457 U.S. 596 (1982). The Court there suggested that “limitations on the right of access that resemble ‘time, place, and manner’ restrictions on protected speech would not be subjected to . . . strict scrutiny.” Id. at 607 n.17. The case that the Supreme Court cited in support of this proposition discussed, in a plurality opinion (and also in dictum in a footnote), the propriety of a judge’s imposition of “reasonable limitations on access to a trial” to preserve the “quiet and orderly setting” of the courtroom or when a courtroom lacked capacity to

30

accommodate all who wished to enter. Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 581 n.18 (1980) (plurality opinion). The Defendants seek support from the Supreme Court’s observation that, when access cannot be granted to all, a court may take reasonable steps to fairly distribute its limited resources by, for example, prioritizing seating for members of the press who can in turn inform members of the public about the court’s business. See id.

For several reasons, we find no merit in the Defendants’ argument.

First, and most important, their argument that strict scrutiny is rendered inappropriate by the Globe and Richmond precedents is irrelevant. The district court did not impose strict scrutiny on the Defendants’ efforts to justify the delay. And neither do we. What was imposed on the Defendants was merely the burden of demonstrating that their pre-access review process was essential to the achievement of higher values and was narrowly tailored to do so, as the Supreme Court required in Press-Enterprise II. This is not strict scrutiny. The strict scrutiny standard was well established by 1986, when Press-Enterprise II was decided. If the Supreme Court had intended policies that burden the right of access to be reviewed under strict scrutiny, it would have said so. Strict scrutiny is far stricter than the Press-Enterprise II standard.

31

Applying strict scrutiny to a state practice or policy generally means its death knell. It would be extraordinarily rare for an instance of race discrimination to survive a strict scrutiny challenge. In contrast, the Press-Enterprise II standard will permit courts to uphold many practices and policies. If the state’s objective is sufficiently valuable to qualify as a higher value, and the delay imposed to achieve it is reasonable, it has a high likelihood of passing muster. While Press-Enterprise II scrutiny is demanding, it is not strict scrutiny. See Courthouse News Serv. v. Planet, 947 F.3d 581, 596 (9th Cir. 2020) (“[T]he Press- Enterprise II ‘balancing test’ is ‘rigorous,’ but not strict, scrutiny.”). Defendants’ argument that strict scrutiny is inappropriate is irrelevant.

Second, the Globe and Richmond precedents on which the Defendants rely predate Press-Enterprise II, in which the Supreme Court crystallized the framework for evaluating the right of access to court documents. If there is incompatibility between the earlier pronouncements and Press-Enterprise II, the later formulation superseded the earlier and is controlling.

Third, we do not believe the delays in releasing court documents such as complaints to the public are properly described as “time, place, and manner” restrictions. The characterization of a restriction as one of “time,

32

place, and manner” applies when the achievement of a significant governmental objective necessitates some imposition on a claimed right that is largely immaterial to the exercise of the right, whether that imposition occurs at one or another time or place, or in one or another relatively comparable manner. In other words, “time, place, and manner” restrictions must “leave open sufficient alternative avenues of communication to minimize the ‘effect on the quantity or content of th[e] expression.’” Vincenty v. Bloomberg, 476 F.3d 74, 88 (2d Cir. 2007) (alteration in original) (quoting Ward v. Rock Against Racism, 491 U.S. 781, 802 (1989)). Delays in providing the press with access to complaints filed in court (if more than trivial) are significantly different. For one thing, where a court withholds public access to a complaint, there may be no alternative channel for the public to become aware of the complaint and its substance. Furthermore, news is a perishable commodity. See Int'l News Serv. v. Associated Press, 248 U.S. 215, 235 (1918) (“The peculiar value of news is in the spreading of it while it is fresh . . . .”). There is a very substantial difference, from the perspective of the press, between receiving access to a complaint when it is newsworthy and receiving

33

access only a few days later if in the meantime the newsworthiness of the complaint has dissipated, as it might for many different reasons.

Fourth, even under intermediate scrutiny, the Defendants would need to show that their review process was “narrowly tailored to serve a significant governmental interest.” Carew-Reid v. Metro. Transp. Auth., 903 F.2d 914, 916 (2d Cir. 1990). As we have outlined above, Defendants have failed to show that their review was narrowly tailored to their interest in preventing the disclosure of confidential information. Accordingly, their argument would fail even under their proposed standard.

In short, we find no merit in the Defendants’ argument and conclude, based on the standards established in Press-Enterprise II, that they have failed to show that there was error in the district court’s ruling that Vermont’s pre- access review process during the period covered in the trial evidence violated the Plaintiffs’ right of access guaranteed by the First Amendment. II. The Permanent Injunction Imposed on Defendants Turning to the permanent injunction issued by the district court, we must address whether its terms were justified. The injunction barred the Defendants “from prohibiting public access to newly filed civil complaints which have not been designated confidential by the filer until the Vermont

34

Superior Court has completed a pre-access review process.” Gabel, 2021 WL 5416650, at *18. Here, we agree with Defendants that no rule of law justified such an injunction on the basis of trial evidence or any of the trial court’s findings. Accordingly, we hold that imposition of a permanent injunction on those terms was not a reasonable exercise of the court’s discretion.

The injunction essentially requires the instantaneous release of new civil complaints without permitting any delay whatsoever for pre-access review, no matter how rapid, efficient, and well justified by its objectives it may be. But the trial evidence and the findings of the district court at trial did not support a conclusion that the Superior Court would be unable to develop pre-access review processes that would be consistent with the requirements of the First Amendment. The evidence and the findings did support the conclusion that the processes being followed in the period covered by the trial evidence violated the First Amendment. The district court could have appropriately enjoined the Superior Court from continuing to follow those

35

processes of delay. 21 But we see no justification for a ruling absolutely barring the Superior Court from instituting a substantially improved practice, especially given that at the time of the trial, the Superior Court had already undertaken a substantial program of improvement in the speed and efficiency of its review practices and, by the time the court entered judgment, had already achieved considerable improvements. The district court recognized that improvements had been made but dismissed their significance on the ground that the Superior Court might return to its past practices if, for example, staffing problems made it difficult to maintain the new centralized and streamlined procedures. Id. at *15. While the court’s concerns about the Superior Court’s ability to adequately staff and fund its central team are reasonable, there was no indication that the judges or administrators of the Superior Court had any intention or desire to return to the practices that the district court had found to be unconstitutional. The potential risk of possible

21 The district court’s opinion suggests that it may have intended a narrower injunction than it imposed. The court said in its opinion that it addressed “only the specifically challenged pre-access review process and le[ft] internal procedures to the Vermont Superior Court[].” Gabel, 2021 WL 5416650, at *17. Nonetheless, the language of the injunction was far broader. It barred the Superior Court “from prohibiting public access to newly filed civil complaints which have not been designated confidential by the filer until the Vermont Superior Court has completed a pre-access review process.” Id. at *18. It left the Superior Court no possibility of developing a review process that would accomplish its important goal without excessive harm to First Amendment interests.

36

backsliding was an inadequate reason to prevent the Defendants from operating under a procedure that would satisfy the governing test.

The First Amendment does not preclude any and all delays in making filed court documents publicly available. On the contrary, the standards adopted by the Supreme Court in Press-Enterprise II expressly contemplate that delays will be permissible if they serve “higher values” and are “narrowly tailored” to advance those goals. The terms of the injunction prevent the Vermont court from protecting higher values, no matter how respectful its processes may be of the Plaintiffs’ First Amendment interests, no matter how minimally and briefly its processes would impose on those interests, and no matter how important the protections of the higher values may be in a particular instance.

While the district court relied substantially on language in Lugosch for invoking the importance of rapid accessibility of court documents—and the Lugosch opinion indisputably contains passages proclaiming that importance—its ruling is in fact incompatible with the terms of the injunction. At every turn of its analysis, the Lugosch opinion envisions courts imposing delays while they inquire into complicated issues before reaching a decision.

37

Lugosch recognized that before ordering release of a document to the public, a court must go through the Press-Enterprise II “experience and logic” test, as we have done, to determine whether there exists a presumptive First Amendment right of access to the document in question and, if so, whether sealing the document or delaying its disclosure would be narrowly tailored to preserve higher values. See Lugosch, 435 F.3d at 120. Engaging in this analysis to assess a party’s claim that the First Amendment entitles it to access before ordering disclosure is incompatible with the proposition that the First Amendment forbids any pre-access delay. Lugosch additionally recognized that in some circumstances, before ordering a disclosure, courts will need to consider specialized arguments not raised in this case, such as whether a privilege might bar disclosure, and, if so, whether the document was proffered by the party protected by the privilege, with the consequence that the privilege may have been waived, see id. at 125, or whether the document is subject to a confidentiality order prohibiting disclosure. Id. at 125–26. These can be complicated questions that can take time to resolve.

In Lugosch, with respect to the dispute before the court, we acknowledged that we were not in a position to determine whether the

38

presumption of immediate access was overcome by countervailing factors, and we remanded to the district court to make specific findings. Id. at 113. The district court ultimately ordered the disclosure of certain documents four months after our decision was issued. See Lugosch v. Congel, No. 1:00-CV-0784, 2006 WL 6651777, at *8 (N.D.N.Y. May 5, 2006). Likewise, in Grove Fresh Distributors, Inc. v. Everfresh Juice Co., which Lugosch cited with approval, the Seventh Circuit remanded the matter to the district court with instructions to consider the arguments raised by journalists who sought access to documents that were covered by the court’s protective order. 24 F.3d 893, 898 (7th Cir. 1994), superseded by rule on other grounds as stated in Bond v. Utreras, 585 F.3d 1061, 1068 n.4 (7th Cir. 2009). Considering the full range of Lugosch’s reasoning, rather than focusing only on its abstract proclamations of the importance of speedy access, Lugosch cannot be understood to mean, at least where the defendant has shown that its delays are in the service of a higher value, that the First Amendment requires a State court to allow public access to newly filed complaints immediately upon filing, without allowing the court to make any inspection whatsoever to guard against public revelation of confidential information contained in the document. Indeed, the Lugosch

39

opinion expressly acknowledged that “any right of access is not absolute” and that courts need time in individual cases to assess whether a particular demand falls within the scope of the right. Lugosch, 435 F.3d at 120 n.4.

The fact that the Superior Court’s processes reviewed by the court at trial were properly found to violate the First Amendment does not support a conclusion that the Superior Court would be incapable of developing pre- access review processes that would be consistent with the requirements of the Constitution. Had the district court entered a declaratory judgment that the pre-access process followed by the Superior Court violated the First Amendment, or imposed an injunction prohibiting the Defendants from continuing to employ those or similar processes, we have no doubt we would have found such a ruling to be a proper exercise of the district court’s discretion. But the Superior Court may well develop new, well-tailored, pre- access screening processes that will provide sufficiently speedy access so as to be in compliance with the First Amendment’s demands.

While we have expressed no conclusion on the question whether Vermont’s screening objective to avoid disclosures of confidential information conducive to serious abuses, such as frauds and identity thefts, served a

40

higher value, this was because it was unnecessary to adjudicate that question in view of the fact that Defendants had failed to satisfy the concomitant test of showing that the review process was “narrowly tailored.” We in no way imply that reasonable delay so as to avoid such harmful disclosures would not serve a higher value. While the district court expressed no view whether this objective pursued a “higher value,” the district court expressly recognized that the objective was “an important one.” Gabel, 2021 WL 5416650, at *15. The district court might conclude on remand that this objective (as opposed to Vermont’s other stated objectives) does serve a higher value, and to find that a narrowly tailored procedure designed to serve that higher value would be consistent with Press-Enterprise II.

For the reasons explained, we vacate the injunction entered by the district court and remand for the court to enter a new ruling that is not inconsistent with this opinion. III. Abstention and Mootness Defendants raise two additional arguments that they contend require reversal: first, that considerations of federalism and comity mandate that the federal courts abstain from adjudicating this case and, second, that this case

41

was moot when the district court issued its judgment. We reject both arguments.

A. Abstention Defendants argue that the district court should have abstained in light of principles of comity, equity, and federalism. We review the district court’s “‘essentially’ legal determination” regarding abstention de novo. Disability Rts. N.Y. v. New York, 916 F.3d 129, 133 (2d Cir. 2019) (quoting Diamond “D” Constr. Corp. v. McGowan, 282 F.3d 191, 197–98 (2d Cir. 2002)). We conclude that the district court did not err in refusing to abstain.

The Supreme Court has repeatedly admonished that federal courts have a “strict,” Quackenbush v. Allstate Ins. Co., 517 U.S. 706, 716 (1996), and “virtually unflagging obligation” to exercise the jurisdiction conferred upon them by Congress, Colorado River Water Conservation Dist. v. United States, 424 U.S. 800, 817 (1976). Accordingly, courts may abstain only in a few “carefully defined” circumstances, and abstention “remains ‘the exception, not the rule.’” New Orleans Pub. Serv., Inc. v. Council of City of New Orleans, 491 U.S. 350, 359 (1989) (quoting Hawaii Hous. Auth. v. Midkiff, 467 U.S. 229, 236 (1984)).

42

Defendants argue that one of those limited circumstances, outlined by the Supreme Court in O’Shea v. Littleton, 414 U.S. 488 (1974), applies here and mandates abstention. We disagree.

O’Shea extended the existing doctrine of abstention under Younger v.

Harris, 401 U.S. 37 (1971), beyond the confines of that case. In Younger, the Supreme Court held that absent “unusual situations,” federal courts should abstain from cases that would interfere with pending state prosecutions. 401 U.S. at 54. Younger abstention was later expanded to certain civil proceedings, but the Supreme Court has reiterated that courts may invoke Younger abstention only in these “exceptional circumstances.” Sprint Commc’ns, Inc. v. Jacobs, 571 U.S. 69, 78 (2013). In O’Shea, the Supreme Court extended Younger abstention to situations in which the relief sought amounts to “nothing less than an ongoing federal audit of state criminal proceedings which would indirectly accomplish the kind of interference that Younger v. Harris and related cases sought to prevent.” O’Shea, 414 U.S. at 500 (citation omitted). Defendants argue, as they did below, that the remedy Plaintiffs seek would amount to such an ongoing federal audit of state proceedings.

43

As the district court concluded, the remedy sought by Plaintiffs is a bright-line rule invalidating the Superior Court’s pre-access review process as in violation of Plaintiffs’ First Amendment rights. This remedy is “more akin to a bright-line finding than an ongoing monitoring of the substance of state proceedings.” Gabel, 2021 WL 5416650, at *12 (alterations adopted) (quoting Courthouse News Serv. v. Planet, 750 F.3d 776, 791 (9th Cir. 2014) (“Planet I”)). The mere possibility that a finding of unconstitutionality of the originally challenged State procedure may be followed by a further challenge to the subsequently developed, ameliorative State procedure does not make it “an ongoing federal audit” of State procedures. O’Shea, 414 U.S. at 500. If it did, it would mean that federal courts could rarely consider constitutional challenges to State procedures. The complaint challenging the State procedures did not unreasonably intrude upon the Vermont judiciary’s autonomy or risk the sort of “monitoring of the operation of state court functions that” O’Shea prohibits. 414 U.S. at 501.

Another factor weighing against abstention here is the significance of the First Amendment right at issue. We have previously held in cases involving the right of access that “the weight of the First Amendment issues

44

involved counsel[ed] against abstaining.” Hartford Courant Co. v. Pellegrino, 380 F.3d 83, 100 (2d Cir. 2004). The district court here was presented with a situation in which over 11% of new civil complaints were not made publicly available until an open-ended period of more than three days after they were filed, giving no maximum upper limit. Such delays are highly suspect and warrant federal court review.

In holding that this case does not mandate abstention, we join the majority of circuits. Four other circuit courts have held that abstention from First Amendment challenges to state courts’ procedures for making certain filings publicly accessible is not warranted. Compare Courthouse News Serv. v. N.M. Admin. Off. Cts., 53 F.4th 1245, 1257–63 (10th Cir. 2022), and Courthouse News Serv. v. Gilmer, 48 F.4th 908, 913–15 (8th Cir. 2022), and Schaefer, 2 F.4th at 324–25 (4th Cir. 2021), and Planet I, 750 F.3d at 789–92 (9th Cir. 2014), with Courthouse News Serv. v. Brown, 908 F.3d 1063, 1071–75 (7th Cir. 2018). We find those courts’ reasoning persuasive and agree that “[a]s long as there is continuing attention given to these ‘delicate issues of federal-state relationships,’ the case can move forward.” Gilmer, 48 F.4th at 915 (quoting Rizzo v. Goode, 423 U.S. 362, 380 (1976)).

45

B. Mootness Defendants also argue that Plaintiffs’ claim was moot by the time the district court issued its ruling. They note that the Superior Court was in the process of transitioning to a centralized review model whereby a central team would review all new filings, as opposed to staff for each county court reviewing only their own filings. They point out that by September 26, 2021, although only partial conversion to centralization had been accomplished, the centralized review team was able to make 95% of initial civil filings available within one business day. See App’x 487, 505. This being so, they argue that the case was moot because the centralized team no longer caused unconstitutional delays and because there was no reasonable expectation that Vermont would reverse course. We disagree.

The Defendants’ argument is deficient on its face. Under the second step of the Press-Enterprise II test, withholding of court documents from public access must pass two tests. It must be designed to serve a “higher value,” and the process must be narrowly tailored to achieve that higher purpose without excessive interference with the public’s right of access. The Defendants’ argument assumes that if 95% of complaints are released within one business day, the process must be narrowly designed. While there are

46

some obvious fallacies in that assumption, the more important failing lies in the Defendants’ failure to show that the objectives that cause the delay serve a higher value. We have concluded that, to the extent that Vermont delays public access to complaints to ensure that the complaints have been signed, that they conform to technical legal requirements, and that they do not show filers’ notes, that undertaking does not serve a higher value. To the extent that release of complaints is delayed to await satisfaction of these non-qualifying objectives, the process is not narrowly tailored within the meaning of Press- Enterprise II. If, for example, Vermont would achieve a significantly higher rate of release on the day of filing, simply by eliminating release delays in the service of these non-qualifying objectives, the Plaintiffs would be entitled under the Press-Enterprise II test to have the better tailored process put into effect. While Defendants’ post-trial submission of the partial results of centralization of the process does show enormous improvement, we cannot agree that it moots the case.

Furthermore, if partial completion of the conversion to centralization achieved such significant improvements, it is reasonable to infer that full

47

conversion might well bring further improvements. Defendants did not succeed in demonstrating that the dispute had become moot. IV. Declaratory Judgment The district court declined to enter a declaratory judgment in the Plaintiffs’ favor on the ground that such a judgment would serve no purpose over and above the injunction that the court ordered. As the injunction granted by the district court is now vacated by reason of its overbreadth, the district court might now wish to consider entering a declaratory judgment while contemplating the parties’ submissions addressed to the terms of the new injunction. 22

CONCLUSION

For the foregoing reasons, the district court’s judgment is AFFIRMED to the extent that it ruled that the Vermont procedures it reviewed violated Plaintiffs’ First Amendment right to access judicial documents. The injunction granted by the district court is VACATED for overbreadth. The case is REMANDED for further proceedings consistent with this opinion.

22The district court’s attention is directed to FED. R. CIV. P. 58, directing that “[e]very judgment . . . must be set out in a separate document . . . .”

RICHARD J. SULLIVAN, Circuit Judge, concurring in part and dissenting in part:

Although I agree with the majority that Plaintiffs have a presumptive First Amendment right of access to newly filed complaints, I cannot agree that Defendants’ pre-access review process violated that right in this case. Because, in my view, the pre-access review process passes muster under the standard set forth by the Supreme Court in Press-Enterprise Co. v. Superior Court of California for the County of Riverside, 478 U.S. 1 (1986) (“Press-Enterprise II”), I would reverse the district court’s judgment and vacate its injunction. 1 As explained by the majority, Press-Enterprise II established a two-part test for evaluating First Amendment right-of-access claims. Under the first step of this framework – also referred to as the “experience and logic” test – a court must determine whether a right of access attaches to a particular type of judicial proceeding or record by considering (1) whether it “ha[s] historically been open to the press and general public,” and (2) whether “public access plays a significant positive role in the functioning of the particular process [or record] in question.”

1I concur in the majority’s conclusions that the district court did not err in declining to abstain from hearing this case or to dismiss this case as moot. For the reasons set forth by the Tenth Circuit in Courthouse News Service v. New Mexico Administrative Office of Courts, 53 F.4th 1245, 1255– 63 (10th Cir. 2022), I agree that abstention was not required in this case. I also agree that this case is not moot, for the reasons set forth by the Fourth Circuit in Courthouse News Service v. Schaefer, 2 F.4th 318, 323–24 (4th Cir. 2021).

Press Enterprise II, 478 U.S. at 8. If the particular record or proceeding passes the “experience and logic” test, “a qualified First Amendment right of public access attaches.” Id. at 9. Notably, however, even when this right attaches, “it is not absolute.” Id. A court must therefore proceed to the second step of the Press- Enterprise II framework and “determine whether the situation is such that [countervailing interests] override the qualified First Amendment right of access.” Id. An “overriding interest” will overcome the presumption of access when there are specific findings that “closure is essential to preserve higher values and is narrowly tailored to serve that interest.” Id. (internal quotation marks omitted).

Here, I agree with the majority and the district court that a qualified First Amendment right of access applies to the newly filed complaints at issue in this case. See Bernstein v. Bernstein Litowitz Berger & Grossmann LLP, 814 F.3d 132, 141 (2d Cir. 2016) (concluding that “[e]xperience and logic both support access” to civil complaints). But I cannot agree with the majority’s conclusion that Defendants failed to “demonstrate that the pre-access review process was ‘narrowly tailored’ to achieve [a] higher value without unduly imposing on . . . Plaintiffs’ presumptive right of access.” Maj. Op. at 22.

As an initial matter, it seems obvious to me that Defendants’ pre-access review process serves a “higher value” within the meaning of Press-Enterprise II. See Courthouse News Serv. v. N.M. Admin. Off. of Cts., 53 F.4th 1245, 1271 (10th Cir. 2022) (discussing state courts’ “valid” interest in ensuring “the orderly administration of justice” when courts reviewed complaints for confidential information); Courthouse News Serv. v. Planet, 947 F.3d 581, 596 (9th Cir. 2020) (same); see also Courthouse News Serv. v. Brown, 908 F.3d 1063, 1069 (7th Cir. 2018) (suggesting that “privacy concerns” weigh against “the interest in public disclosure” under Press-Enterprise II). Indeed, Defendants’ asserted interest in “[p]rotecting the privacy interests of litigants and third parties,” Defs. Br. at 47 (internal quotation marks omitted), is a well-established and significant governmental interest. See U.S. Dep't of Just. v. Reps. Comm. for Freedom of Press, 489 U.S. 749, 767 (1989); Seattle Times Co. v. Rhinehart, 467 U.S. 20, 35 (1984). To the extent the pre-access review process includes checks for information arguably unrelated to this privacy interest (such as whether the complaint is properly signed or contains filers’ notes), these checks are incidental to Defendants’ more searching review for confidential information. Nothing in the record suggests that these additional checks for compliance with ministerial requirements has had any

4

discernible impact on the time it takes an individual complaint to pass through the pre-review process. See Courthouse News Serv. v. Gabel, No. 21-cv-132, 2021 WL 5416650, at *7 (D. Vt. Nov. 19, 2021) (finding that the “majority of the pre-access review process is devoted to a manual review of the complaint for confidential information” and that this is the “only portion of the pre-access review process which is not duplicated by . . . software”); see also App’x at 82–83 (explaining that, even before the advent of electronic filing, clerks would conduct a “cursory” review of a paper complaint and “check[] for a signature” before making the complaint available to the press).

Nor am I persuaded that the relatively minimal delay between the receipt and docketing of new complaints in this case constitutes an infringement on Plaintiffs’ First Amendment rights. The record reflects that, during the relevant period, 54.8% of complaints were made available to the public on the day of filing and an additional 22.6% of complaints – 77.4% total – were made available within one day. Notably, Defendants were able to make complaints available on this timeline despite the fact that the Vermont court system experienced significant COVID-19-related staffing, equipment, and training challenges during the relevant period, see Gabel, 2021 WL 5416650, at *3, which almost certainly

5

contributed to the delays at issue in this case. The record confirms that Defendants responded to these challenges by making improvements to the pre-access review process such that, by September 2021, approximately 95% of complaints were available within one business day. See App’x at 505.

Although the majority states that some “comparison of intangibles” is required to determine whether the benefit obtained from the delay serves a “higher value” than the imposition of the delay on Plaintiffs’ right of access, Maj. Op. at 21, I believe the delay here is sufficiently minimal and the benefit is sufficiently important to constitute a relatively “higher” value. See Brown, 908 F.3d at 1066, 1070 (characterizing “a delay of no more than one business day in access to the vast majority of electronically filed complaints” as “minimal”); see also N.M. Admin. Off. of Cts., 53 F.4th at 1271–73 (recognizing that the presumption in favor of timely access to newly filed civil complaints must be evaluated in the context of practical restraints on the court system); Courthouse News Serv. v. Schaefer, 2 F.4th 318, 328–29 (4th Cir. 2021) (same). The cases in which other Circuits reached a contrary conclusion involved more significant delays than those presented in this case, resulting in a much greater imposition on the First Amendment right of access. See Planet, 947 F.3d at 597–98 (noting that “between 2012 and 2014, it took

6

two or more court days for [Plaintiff] to access one-fifth to two-thirds of newly filed complaints”); Schaefer, 2 F.4th at 322 (explaining that, of the two sample circuit courts discussed by the district court, one circuit court made “only 19% of the complaints available on the day of filing, and 22% of the complaints were not available until two or more court days after filing” and the other circuit court “only made 42.4% of the complaints available on the day of filing and 41.5% of the complaints were not available until two or more court days after filing”). Given that Plaintiffs typically only provide same-day media coverage of 42% to 46% of newly filed state-court complaints nationwide, I struggle to see how Plaintiffs’ interest in prompt disclosure was impermissibly infringed in this circumstance.

I also disagree with the majority’s suggestion that the pre-access review process was not narrowly tailored because Defendants failed to show that their pre-access review could not reasonably have been accomplished more rapidly and that they were reasonably capable of accomplishing their pre-access review with considerably more speed than occurred during the “trial evidence period.” 2 Maj. Op. at 28. It is indisputably true that Defendants have adjusted and centralized

2As used in the majority opinion, the term “trial evidence period” appears to mirror the term “Designated Period” referenced in the district court’s opinion, which spanned from the date the various Vermont superior courts transitioned to electronic filing until August 6, 2021. See Sp. App’x at 12; see also Maj. Op. at 9–10 (delineating what data the district court reviewed at trial).

their pre-access review processes to be even more efficient than they were during the trial period: by the time of the district court’s ruling, Defendants were making approximately 67% of complaints available on the same day and 95% of complaints available within one business day. But this fact alone does not mean that the prior review process was not “narrowly tailored”; it simply leads to the obvious conclusion that new statewide processes took time and resources to implement effectively, particularly when faced with the logistical challenges that accompanied a global pandemic. Although the majority specifically disclaims any suggestion that it is applying strict scrutiny in this case, it essentially concludes that – because the initial implementation of the pre-access review process was not the least restrictive or least intrusive means of protecting private information when specifically compared to a subsequent, more efficient iteration of that process – the initial implementation was not narrowly tailored. That approach is flawed as a matter of law and logic. See Ward v. Rock Against Racism, 491 U.S. 781, 798–99 (1989) (explaining that a regulation “need not be the least restrictive or least intrusive means” to be considered narrowly tailored). Furthermore, and as a practical matter, I fear that this aspect of the majority’s reasoning will seriously hinder state courts’ ability to innovate, pilot new procedures, or methodically roll

8

out new statewide processes in stages – as Defendants have endeavored to do here – unless those courts are fully assured that there will not be (even temporary) growing pains from doing so.

For these reasons, I respectfully dissent from the majority’s opinion and would reverse the district court’s judgment as to Plaintiffs’ First Amendment claim.