How Civil Subpoenas Are Used To Unmask Online Speakers, And How A Recent Decision Will Help Deter Bogus Ones
from the unexpected-good-news dept
Important cases don’t always happen with a lot of fanfare. It may be easy to follow what the US Supreme Court is up to, with its relatively small docket of high-profile matters, but plenty of other important cases get resolved by state and lower courts around the country with much less attention but just as much import.
This decision by a California appeals court, Roe v. Halbig, is one such example, and happily the impact it stands to have is a good one. It isn’t a showy decision declaring some new principle of liberty. Rather, it stands to quietly help ensure that codified protections for speech, and anonymous speech in particular, work as intended.
We’ve written many times before about how important it is that anonymous speech be protected. Indeed, the US Supreme Court has found that the First Amendment includes the right to speak anonymously, because without that right a lot of important speech could not happen. But it’s one thing to say that anonymous speech must be protected; it’s another to make sure that anonymous online speakers can remain anonymous on a practical level. If it is too easy to unmask speakers, then their right to speak anonymously becomes illusory.
To prevent the right to anonymous speech from becoming meaningless, it’s important that discovery instruments, like subpoenas, intended to unmask speakers, not be vulnerable to being abused, especially by plaintiffs who don’t have a legitimate need to unmask their critics. Because not only is a SLAPP suit chilling to speech, but so is a subpoena arising from a SLAPP suit that strips a speaker of the anonymous protection they counted on having when they spoke.
This decision will help prevent the latter. To understand how, it helps to understand how these subpoenas get used.
What typically happens is that a SLAPP is filed in another state (or country), likely one that does not have a robust anti-SLAPP law, and names a “John Doe” defendant. The plaintiff then issues a subpoena connected to the case targeted at whatever Internet platform (e.g., Twitter, Google, Facebook, Automattic/WordPress, Yelp, Glassdoor, etc.) or platforms may have information that would help identify who the speaker was. Obviously this information would be needed in order to maintain the lawsuit ? you need to know who you are suing in order to actually sue them ? but there is nothing requiring a lawsuit to continue once the identification is made. Sometimes SLAPP plaintiffs file lawsuits only as a vehicle to learn who their critic was because that’s all they need to be able to make their critic regret speaking against them.
If the platform is exposed to the jurisdiction of this other state, and thus subject to the subpoena, then all of that state’s rules about subpoenas will govern what comes next. But if the California-based platform is not subject to the jurisdiction of this other state, then the plaintiff will need to “domesticate” it with the court of the county in California where the platform is located. It is generally easy to domesticate a subpoena; any California-licensed attorney can issue one on a special form provided by the California courts. It contains the same demand to produce information that the out-of-state subpoena had, only now the demand is governed by California law with its various speech-protecting rules.
In general, a platform will try to notify the user that it has received the subpoena to unmask them. (This is an important step, which is why we’ve also been so critical of discovery rules preventing this notice.) Sometimes the platforms might even try to fight the subpoena themselves, which some recent California appellate cases said they have the right to do. It also means that when the courts consider whether to quash a subpoena they will use the Krinsky test, which is a relatively speaker-protective test used by courts to decide whether there is a sufficient basis to warrant a speaker being unmasked. Courts won’t definitively decide the case at this stage, but per the test they will not allow a speaker to be identified if the plaintiff has not made at least a prima facie showing that the claims in the lawsuit may be valid. Speakers shouldn’t lose their anonymity if there’s no chance that the plaintiff might win.
And then that’s where this case comes in. Because if the motion to quash the subpoena is successful, the party who brought the motion gets to recover the fees and costs of doing so.
The rule at issue here is much like the anti-SLAPP statute, which serves to both compensate a wronged speaker who has been forced to defend a lawsuit targeting their protected speech and also to deter plaintiffs from bringing these garbage lawsuits in the first place by making the plaintiffs pay the defendant’s legal fees. But the anti-SLAPP statute only governs actual lawsuits [p. 13]. It doesn’t have any effect on similarly meritless subpoenas arising from out-of-state SLAPPs. In order to prevent litigants from filing their lawsuits in other states (or countries) beyond the reach of the California anti-SLAPP law, and then using those meritless lawsuits as a basis to issue subpoenas to unmask their critics, in 2008 the California legislature inserted some language into its rules of civil procedure to address this situation. Section 1987.2 of the California Code of Civil Procedure reads:
(c)?If a motion is filed under Section 1987.1 for an order to quash or modify a subpoena from a court of this state for personally identifying information, as defined in subdivision (b) of Section 1798.79.8 of the Civil Code , for use in an action pending in another state, territory, or district of the United States, or in a foreign nation, and that subpoena has been served on any Internet service provider, or on the provider of any other interactive computer service, as defined in Section 230(f)(2) of Title 47 of the United States Code , if the moving party prevails, and if the underlying action arises from the moving party’s exercise of free speech rights on the Internet and the respondent has failed to make a prima facie showing of a cause of action, the court shall award the amount of the reasonable expenses incurred in making the motion, including reasonable attorney’s fees.
In other words, if a subpoena targeting protected online speech is successfully quashed, then, like with the anti-SLAPP statute, the court must award the party who expended the resources to quash the subpoena the fees and costs they had to spend to do it. Like the anti-SLAPP statute this language both serves to compensate a wronged-speaker for the defense of their speech rights and is a deterrent to others who otherwise might be inclined to casually issue subpoenas to harass their anonymous critics.
[T]he legislative history of section 1987.2, subdivision (c) highlights the Legislature?s focus on the burden on free speech posed by subpoenas derived from out-of-state cases targeting anonymous speakers on the Internet; the Legislature?s concern with the costs of litigating cases threatening free speech; and the Legislature?s intent to protect the exercise of free speech on the Internet. The anti-SLAPP statute reflects similar considerations. [p. 15]
Given how many Internet platforms there are in California, this is a particularly powerful piece of legal code. It’s a significant reason why we’ve praised decisions requiring subpoenas to be domesticated in California, where so many platform companies are based, so that their users can benefit from the protection that this legal code affords. Similar protection is not generally available in other states (although it really should be), which means that if a platform is forced to respond to a subpoena governed by another state’s law, the users will be on their own to fund their own defense.
But even though this powerful language has been on the books for a decade, there have not been many cases interpreting or affirming it. In fact, the court here suggests it may be the first, which is partly why this case is so important. [p. 28].
For one thing, it helps make it much more usable for Doe defendants. To explain why, we need to return to the story about how these subpoenas play out. After all, the courts don’t automatically run the Krinsky test on every issued subpoena. First the Doe defendant has to find counsel to help challenge it, which is much easier to do when there is the more certain promise of fee recovery. Next, the Doe defendant has to get the subpoena challenge in the courthouse door. Unhelpfully, every California county’s courts handle these petitions to quash a bit differently (which can present challenges for counsel, who needs to figure out what these procedures are ? this is not an area of law well-documented in practice guides).
Also, some counties’ courts require these petitions to quash be adjudicated by a rotating batch of pro-tem judges. These are practicing lawyers trusted by the courts to help clear the thicket of day-to-day discovery disputes that regularly arise from California litigation and would otherwise drown the courts without the extra help. Unfortunately they are not necessarily equipped to adjudicate the significant free speech issues that happen to end up before them because they come wrapped as a discovery matter ? in this case, a motion to quash a subpoena.
(Note also that because these subpoenas are often channeled through the discovery departments of the county courts that is likely why there has been so little written precedent before now, because this sort of adjudication generally leaves no formal written decision, or even much of a record at all.)
So this case allows Doe defendants to more efficiently educate these judges. Not only does it provide a clear judicial precedent interpreting the procedural language and showing that its protection is real and meaningful, but, in rejecting the fee award as being unduly discounted, it tells these judges that the fee awards at stake in these cases are not only mandatory but potentially substantial. It is an important admonition because it is extremely rare that in a normal discovery dispute a judge would award a prevailing party more than a few thousand dollars. Judges hearing discovery matters may therefore be disinclined to ever award more. But it may realistically require an amount deep into five figures in order to properly quash one of these speech-chilling subpoenas, and here the appellate court says that these cost claims should not be presumptively discounted simply because they may be large.
The total bill came to $42,273 for 192 hours of attorneys? work plus $308 for the paralegal?s time. Based on these numbers, the trial court stated, ?[$]42,000 was a lot, extreme in setting in this case. . . . I think a petition to quash a subpoena should not require that amount of time. So I?m going to award $22,000 in fees.? The trial court made clear that, in its view, Roe?s attorneys had spent an unreasonable amount of time on the case. The trial court is entitled to draw that conclusion, for, ?an ?experienced trial judge is the best judge of the value of professional services rendered in [her] court.? ? (Walent v. Commission on Professional Competence etc. (2017) 9 Cal.App.5th 745, 748.) Nevertheless, the starting point for the trial court?s calculation must be all of the hours counsel has spent on the case. Roe suggests that perhaps the trial court awarded an amount equal to the time his attorneys spent on the initial motion to quash. If the trial court?s fee award were based on this metric, it would constitute an abuse of discretion. Halbig filed an opposition to the motion to quash and withdrew the subpoena; both of those events reasonably required further briefing from Roe. An award based solely on the attorney hours spent preparing the initial motion to quash would violate the principle that the starting point of the fee calculation must be all of the attorney hours actually spent on the case. (Ketchum, supra, 24 Cal.4th at pp. 1131-1132.) [p. 27]
This decision also cleared up an area of potential ambiguity in the code’s applicability by discussing what it means to be a “prevailing party” eligible to recover fees and costs. Here’s the problem that this case illustrated: the Doe defendant had to spend a lot of money trying to quash the subpoena, but before the court was able to rule on the original motion to quash, the plaintiff withdrew the subpoena. The question the appeals court had to decide was whether that withdrawal made the Doe defendant ineligible to recover as a “prevailing party” since there was no judicial victory.
The appeals court decided that the Doe defendant could still collect, observing that it was impossible for the defendant to know whether they were out of the woods. The plaintiff, in rescinding the subpoena, did so explicitly “without prejudice,” meaning that it could be reissued at any time, and the court found it contrary to the purpose of 1987.2 to leave the defendant uncompensated for the expenditure they had been forced to make.
Halbig argues that, once he had withdrawn the subpoena, Roe had no reason to fear that Halbig?s further efforts would unmask his identity, and he maintains that Roe?s refusal to withdraw the motion to quash needlessly incurred extra costs [?] Halbig?s argument, however, ignores the possibility that he could have sought another subpoena for Roe?s identity. Halbig asked for and received a ?dismissal? of the Google subpoena ?without prejudice.? The record contains no evidence that at the time of the hearing on the motion to quash Halbig had determined the identity of the ?Doe? defendants or dismissed the underlying lawsuit in Florida. As noted in the analogous anti-SLAPP context, ?[t]he specter of the action being refiled (at least until the statute of limitations had run) would continue to have a significant chilling effect on the defendant?s exercise of its First Amendment rights. At that point, the plaintiff would have accomplished all the wrongdoing that triggers the defendant?s eligibility for attorney?s fees, but the defendant would be cheated of redress.? (Coltrain, supra, 66 Cal.App.4th at pp. 106-107.) [p. 21-22]