Wise Words from Third Circuit Judge Stephanos Bibas as to Access to Court Records

Dear Members of the Advisory Committee on Civil Rules:

I write to express my strong support for proposed Rule 5.3, a much-needed recapitulation of the procedural standards governing sealing of judicial records.

The public has a presumptive right to access judicial records. This right serves several important societal functions. By exposing the records of private wrongdoing to the public eye, this right deters further wrongdoing. By forcing the government to litigate out in the open, it empowers the public “to keep a watchful eye on the workings of public agencies.” Nixon v. Warner Commc’ns, Inc., 435 U.S. 589, 598 (1978). And by enabling the public to dig into the bases for judicial decisions, it enables the public to assure itself of the strength of judicial reasoning, the cornerstone of judicial legitimacy. This right thus not only “diminishes possibilities for injustice, incompetence, perjury, and fraud,” but also provides “the public with a more complete understanding of the judicial system and a better perception of its fairness.” In re Avandia Mktg., Sales Prac. & Prod. Liab. Litig., 924 F.3d 662, 672 (3d Cir. 2019) (internal quotation marks omitted).

On paper, this right looks strong. In the U.S. Court of Appeals for the Third Circuit, where I sit, a district court issuing an order to seal “must articulate the compelling, countervailing interests to be protected” and “make specific findings on the record concerning the effects of disclosure.” Id. (internal quotation marks omitted). But that standard is a paper tiger—mighty in appearance, but toothless in practice. Time and again, we see cases on appeal where the parties asked to seal large parts of the record below, and the district court, with little discussion, agreed to do so.

Take one recent motion to seal in a Third Circuit case. The district court had certified a multi-state class action against a car company. On appeal, the company asked us to seal ten volumes of appendices, totaling nearly three thousand pages of material. The company told us that sealing was necessary to prevent competitors from copying its design choices, stealing its engineering solutions, and exploiting its sales strategy.

But it took only a cursory glance at the appendices to see how overbroad the company’s request was. Hundreds of pages, for example, covered deposition testimony from one of the plaintiff’s lay witnesses, a mechanic who had worked on the defendant’s cars and had commented on their alleged defects. Hundreds of other pages dealt with one of the plaintiffs’ experts, who had opined on the veracity of the defendant’s public comments and sales catalogues. But if the mechanic based his testimony on his day-to-day work on the defendant’s cars, and the expert had based his testimony on the company’s public statements, how would these materials reveal competitive secrets? So I denied the request without prejudice, asking the company to come back with a narrower request if it so chose. But why had the record below been sealed in the first place?

It should come as little surprise that motions to seal almost always go unopposed. The harm of overbroad sealing falls on future victims of repeat tortfeasors, future litigants in related cases, and the public at large. It does not fall on the parties to the case.

Indeed, the parties generally have an active interest in secrecy. The party who moves to seal records fears that the information produced in the case could lead to bad press coverage, competitive harms, or even follow-on litigation. The other party presumably does not share those same concerns about broad disclosure, but it does not share the public’s concerns about overbroad sealing, either.

So in the give-and-take game of litigation, where it can be important to reserve your capital for fighting the fights that matter to your client, opposing a motion to seal is a waste of that capital. At the very least, it is a waste of time and money. For lawyers in an active dispute—so used to viewing themselves as only agents of their clients, and not also officers of the court—submitting an opposition to an overbroad motion to seal might not seem worth the cost of the paper to print it.

This leaves it to district-court judges to police these motions and protect the public’s presumptive right of access to judicial records. But what my experience as a judge on the court of appeals has taught me, Professor Engstrom et. al’s empirical work now confirms: Most district courts rubber-stamp these requests.

This might come as a surprise to some members of the Judicial Conference. After all, the law of every circuit requires district courts to reject requests to seal that are too broad to be justified by their supposed rationale.

So why do district courts frequently fail to apply that law? One problem might be the lack of developed precedent in this area. Because motions to seal are almost never subject to full treatment in district court opinions, let alone appellate review, there are few decisions applying legal standards to specific facts, teasing out what interests suffice to justify sealing, and how broad is too broad. Without clearly factually analogous precedent to rely on, district courts might be more hesitant to deny a motion to seal as overbroad, even though the operative legal standard compels that result.

Another problem is rooted in the lack of adversarial presentation. With no one on the other side, inflated claims of confidentiality go unchallenged.

Other problems might boil down to the mechanics of busy district court dockets. Perhaps, for example, when a district court grants a protective order, that decision generates an inertial force that skews the outcome of the subsequent motion to seal, even though the two decisions are subject to different legal standards. Or perhaps a motion to seal, because it is unopposed, slips under the radar, resembling an administrative matter to be rubber-stamped rather than a legal decision to be carefully scrutinized.

But whatever the reason, the results are clear: For most motions to seal, “the parties agree, the busy district court accommodates, and nobody is left in the courtroom to question whether the decision satisfied the substantive requirements.” Binh Hoa Le v. Exeter Fin. Corp., 990 F.3d 410, 417 (5th Cir. 2021).

This proposed rule will push parties and courts alike to do better. First, this proposed rule will ensure that the parties know what their obligations are when they submit a motion to seal. I often sit by designation on the District of Delaware, where I regularly receive motions to seal with no accompanying explanation or justification, putting the onus on me to push the parties for more. Under Proposed Rule 5.3(c), parties will be reminded that, as our circuit has regularly held, “[b]road allegations of harm, bereft of specific examples or articulated reasoning, are insufficient.” In re Avandia Mktg., 924 F.3d at 673 (internal quotation marks omitted).

Second, this proposed rule will act as a process-forcing mechanism for busy district court judges. District courts lack the benefit of adversarial litigation over motions to seal. And without a party opposing a motion, it is easy to think that no reasoning is needed for an order—if no party loses, the thinking goes, no party needs a basis for the court’s decision.

Proposed Rule 5.3(h) makes clear that even without an adversarial party, the district court must act as the steward of the public interest in accessing court records, and must reason through its decision. The decision to seal is for the court to make by applying the correct legal standard, not for the parties to make by means of consent. Proposed Rule 5.3(i) provides a checklist that a district court can use in reaching that decision.

Finally, the proposed rule will remind parties and district courts alike that sealing is not an all-or-nothing question. Proposed Rule 5.3(d)(3) and (g) make clear that, where viable, the default approach should be targeted redactions. The choice is not between disclosing a list of social security numbers and sealing an entire document. A happy medium is often readily available.

This proposed rule does not make new law. Sunlight is the best disinfectant for many societal ills, yet sealing “insulates the participants” in the legal process, “masking impropriety, obscuring incompetence, and concealing corruption.” Brown & Williamson Tobacco Corp. v. F.T.C., 710 F.2d 1165, 1179 (6th Cir. 1983). Judicial reasoning is the basis for judicial legitimacy, yet sealing “withdraws an element of the judicial process from public view,” making “the ensuing decision look more like fiat.” Matter of Krynicki, 983 F.2d 74, 75 (7th Cir. 1992) (Easterbrook, J.). For these reasons, the courts of appeals have consistently held that the public has a presumptive right to access judicial records, one that should only be abrogated where necessary, and even there only to the extent necessary. This proposed rule simply makes that right harder to neglect.