The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Free Speech

Why We Should Care About Pseudonymity in Litigation

Like other features of legal procedure—such as the jury trial, the mechanism for appointing judges, the availability of appeal—pseudonymity both deeply affects the fairness of litigation and, often, the substantive outcomes.


A commenter asks, perfectly fairly:

I started reading Volokh a year or so ago. None of these seemingly dozens of pseudonymity posts have seemed interesting enough to read, including this one. IANAL. Is pseudonymity actually a serious legal issue or just a weird obsession of Volokh's?

Well, yes, it is an obsession of mine; I wrote a 100+-page article on the subject, and am planning to put together an electronic treatise on the Law of Sealing and Pseudonymity in Litigation. And one reason I'm doing that is precisely that academics and others generally haven't focused much on the subject, so I suppose it is indeed a "weird" obsession.

But I think everyone should focus on it! (The hallmark belief of a weird obsession, to be sure.) Like other features of legal procedure—such as the jury trial, the mechanism for appointing judges, the availability of appeal—pseudonymity both deeply affects the fairness of litigation and, often, the substantive outcomes. Pseudonymity isn't just a matter of civil procedure; it should also matter to people who care about privacy, free speech, public supervision of courts, and more.

Here's my pitch, adapted from the Introduction to my article:

One defining question about any system of procedure is: Public or secret? American juvenile justice is secret. Criminal justice, generally public. Bar discipline, mostly secret in many states. Internal employer and university disciplinary proceedings, generally secret. Arbitration, generally secret. Civil justice, public.

The answer to the public-or-secret question of course affects the level of public supervision of the system, as well as the likely public confidence in the system. But the answer can also sharply affect the shape of litigation within the system:

  • the incentives to bring or not bring various kinds of cases,
  • the incentives to settle (or plea bargain),
  • the likely settlement values,
  • which witnesses testify,
  • and more.

Indeed, the implicit threat of publicity is common in many prefiling negotiations, though it may need to be kept implicit to avoid negotiations being treated as criminal extortion.

The follow-up question, of course, is: When a system is generally public, what provisions still allow some degree of secrecy? In particular, within our civil justice system, how do courts decide what can or must be sealed or redacted, and when parties can proceed pseudonymously? This too can sharply affect what cases get filed, what cases get dropped, and on what terms cases settle.

Yet the Federal Rules of Civil Procedure, unlike some state court rules, say little to answer this question. I want to push these questions—especially the one about pseudonymity—to their rightful place in our discussions about civil procedure.

This question has become especially important because court records are more visible than ever, including to casual Internet searchers. For many litigants these days, one of the most important questions is: Can I keep my name, and its connection to the case and its facts, out of Google search results?

Before, a typical employment lawsuit, for instance, would rarely be reported in newspapers. But now, Googling people's names will often find many of the cases in which they have participated, even if no reporter has ever written about those cases.

And many litigants would love pseudonymity. That's particularly obvious for defendants, most of whom are being sued over alleged misconduct. Say someone sues you for alleged embezzlement, fraud, or sexual assault, or even malpractice or breach of contract. You'd surely prefer that your friends, neighbors, and prospective clients and business partners not know about it. And while some defendants simply want to hide their misdeeds, others are innocent and don't want to be linked to incorrect accusations—whether temporarily, pending the trial and verdict, or perhaps forever.

Many plaintiffs would want pseudonymity, too; to offer a few examples:

  • Sexual assault plaintiffs may not want to be publicly identified.
  • Libel plaintiffs may not want to further publicize the allegedly libelous allegations over which they are suing.
  • Employment law plaintiffs who were fired for alleged misconduct, but are claiming that this was a pretext, may not want a Google search for their names to lead to those allegations (however forcefully denied).
  • People suing over politically controversial behavior (for example, an employee fired for allegedly racist or unpatriotic statements) or suing using legal theories that some might condemn or mock may not want to be publicly shamed or humiliated.
  • Even ordinary employment law or housing law plaintiffs may not want future employers or landlords to reject them as dangerously litigious.

Yet for good reason, most lawsuits are nonetheless litigated in the parties' own names. That is obviously true of adult criminal cases, even though nearly all criminal defendants would much prefer pseudonymity. And it's true of civil cases—our legal system generally calls for public proceedings and publicly filed documents, and the names of the parties are viewed as part of the information that needs to be kept public.

Such openness is viewed as important for letting the public (usually through the media) supervise what happens in courtrooms that are publicly funded and exercise coercive power in the name of the people. Many major stories and some scandals have been broken in part because of the availability of civil court records. The Boston Globe's investigation of the Catholic Church's coverup of sexual abuse by priests, dramatized in the film Spotlight, is just one especially noted example. And of course this is true of more minor stories as well; my Shenanigans: Internet Takedown Edition article discusses various frauds that I uncovered in large part because of public access to court records. And even for the many cases that go largely unnoticed, the possibility of public review helps deter misbehavior.

Some cases conclude that the First Amendment itself thus secures a presumptive right of the public to know litigants' names, as it has been held to secure a presumptive right of public access to court records. And more broadly, this openness is a matter of free speech and the public right to know (whether constitutionally secured or not).

How then are these interests reconciled? It turns out that the law is largely unsettled, for instance with regard to:

  • whether plaintiffs alleging sexual assault can proceed pseudonymously;
  • whether plaintiffs may proceed pseudonymously to avoid disclosure of their mental illnesses;
  • whether pseudonymity is more justified in lawsuits against governmental defendants or less justified;
  • when defendants may proceed pseudonymously just to prevent possible damage to reputation stemming from the allegations at the heart of the lawsuit, allegations that defendants claim are false;
  • when plaintiffs may proceed pseudonymously when they are suing over allegedly false allegations, for instance in a libel lawsuit;
  • whether minors' parents may proceed pseudonymously to protect minors' pseudonymity;
  • whether young adults may proceed pseudonymously on the theory that they are nearly minors;
  • whether adult litigants may proceed pseudonymously when they allege they were assaulted when they were minors.

And many of the distinctions that the cases do appear to implicitly draw are hard to explain. Imagine, for instance, that Arnold is an adult university student accused of sexually assaulting his classmate Veronica:

  • The criminal prosecution would almost certainly be People v. Arnold, not People v. Doe, notwithstanding the harm to Arnold's reputation (a harm that would be present even if he's later acquitted or the charges are dropped).
  • The civil lawsuit would often be Veronica v. Arnold.
  • But some courts would allow it to be Doe v. Arnold, to protect Veronica's privacy.
  • A few courts would allow it to be Doe v. Roe, seemingly on the theory that, just as it can be unjustly humiliating for many sexual assault victims to be publicly identified as such (assuming they are telling the truth that they were indeed victimized), so too it can be unjustly humiliating for many of the accused to be publicly identified as such (assuming they are telling the truth that they were not guilty). But most courts do not accept this theory. (Of course, as a general matter Arnold would need to know Veronica's identity; I focus here on pseudonymity that shields the parties' identity from the general public, and not from other parties, or at least their lawyers, or the court.)
  • If Arnold sues Veronica for libel, claiming Veronica's accusations were lies, most courts would require it to be Arnold v. Veronica or perhaps Arnold v. Roe, but not Doe v. Roe.
  • But many courts routinely allow the pseudonymous Doe v. University of Northern South Dakota, a lawsuit in which Arnold is claiming that the university acted improperly in expelling him for the alleged misconduct—even though there, as in the libel case, Arnold wants pseudonymity to protect his reputation.

It's hard for me to see a sound justification for this pattern.

Finally, let me close with five observations that make pseudonymous cases particularly important, difficult, or both:

[a.] The ubiquity of the desire for pseudonymity: I noted above that many plaintiffs and defendants would prefer to keep their names out of the court record and therefore off Google and out of the newspapers. Courts have observed this and often cite this as a reason to reject pseudonymity—if we let this litigant be pseudonymous, we'd, in fairness, have to let all these other litigants do the same, and then we'd have a very different and much less transparent system of procedure.

[b.] The puzzle of dealing with reputational damage: In particular, a vast range of cases involves material risk of reputational damage to one or both parties—chiefly, damage to the ability to earn a living. Courts often remark that mere risk of reputational damage (including unjust reputational damage, for instance, if the accusations against a defendant ultimately prove to be unfounded) is not enough to justify pseudonymity. But not all cases so hold. This is in part because the reputational concerns can seem so serious and salient. And the cases that allow pseudonymity to protect privacy rather than to protect reputation sometimes boil down to risk of reputational damage as well (for instance, if a plaintiff seeks pseudonymity to conceal information about a mental illness).

[c.] Settlement skew: The settlement value of a case generally turns in large part on the ongoing costs of the lawsuit to the two parties—litigation costs, emotional costs, or reputational costs. All else being equal, if the plaintiff's costs go down, the plaintiff will be emboldened, and the settlement value of the case will likely increase. Likewise, if the defendant's costs go down, the settlement value of the case will likely decrease; most obviously, the settlement value will decrease if the defendant can reduce its litigation costs, perhaps if a defendant gets ideologically minded pro bono counsel.

It follows that, in cases where both sides have reputational or privacy costs stemming from the litigation, giving pseudonymity to one party but not the other would decrease the pseudonymous party's costs and would change the likely settlement value. All else being equal, a Doe v. Smith will tend to yield a larger settlement than Jones v. Smith or Doe v. Roe, which in turn will tend to yield a larger settlement than Jones v. Roe. This can be an argument for rejecting pseudonymity—or for pseudonymizing both parties.

[d.] Pseudonymity creep: Simply pseudonymizing a party seems easy enough, and seems like only a modest restriction on public access. But, of course, other information in the case can lead interested researchers to the party's identity. Even if a minor's name is abbreviated L.V., if the case is Volokh on behalf of L.V. v. Los Angeles Unified School District, it might not be hard for people to identify L.V. based on her representative's (likely her parent's) name. Likewise, if a complaint filed by John Doe in a libel case quotes the alleged libel, a quick Google search for the libel could identify its target. If a woman sues her ex-boyfriend alleging sexual assault, people who know the ex-boyfriend may easily identify the woman.

To make pseudonymity effective, more needs to be done than just pseudonymizing one particular party. This may include sealing important material outright, pseudonymizing the other party as well, or enjoining the other party from revealing the pseudonymous party's name (or other details of the lawsuit) in public comments. But then pseudonymity would also interfere more with public right of access and may further undermine the interests of the opposing parties.

[e.] Sharp variability among cases: As I've noted, cases are sharply split on whether to allow pseudonymity, in nearly every category of cases. And that is unsurprising, given how vague the factors are—factors such as "the bases upon which disclosure is feared or sought to be avoided, and the substantiality of these bases" and "the magnitude of the public interest in maintaining the confidentiality of the litigant's identity."

There are three possible explanations for these different results (all of which may be present in some measure):

  • Differences in circumstances: Perhaps the multi-factor balancing tests that various courts have announced are working well, and judges are carefully drawing distinctions based on real differences between the cases.
  • Differences in litigants: Or perhaps courts sometimes just decide based on sympathies (perhaps subconscious) for certain kinds of litigants—for example, for fellow lawyers, promising young college students, or people who are bereaved (even though such bereavement is generally not seen as a basis for pseudonymity).
  • Differences in judges: Or maybe different judges have different attitudes about pseudonymity generally, with some taking a sharp public-right-to-know attitude and others being much more sympathetic to litigant privacy.

To the extent the explanation is a difference in circumstances, it is a virtue of the vague balancing tests that appellate courts have set forth for pseudonymity decisions. To the extent the explanation is a difference in litigants or judges (or both), it is a vice.