Wednesday, July 6, 2016

Of Chimp Sarcasm and IP Addresses.

How far do you need to go to authenticate information obtained off the internet?

A rug cleaning company sued a customer over some defamatory Yelp reviews. Like most reviews on Yelp, however, the comments were essentially anonymous. So plaintiff subpoenaed Yelp for the reviewers’ IP and email addresses. And then she subpoenaed the Internet Service Providers for subscriber info matching the IP addresses at the time of the posts. One ISP gave docs showing the user of the IP addresses as of a date about a month and a half before the posts, but was apparently unable to give IP address info as of the exact posting date. The other gave info on a dead account. Both, however, pointed to the subscriber as Defendant, one at home and one at work.

Defendant moved to exclude the docs in limine, arguing that the records weren’t enough to pin the posts on him. Someone else could have been using his computer, or stealing his internet, or the old IP address could have been reassigned to someone else, who might have been the poster. Maybe it was his mom? Or some dude that looked like him? Or maybe that guy who hacked Anthony Wiener’s phone? Whatever. So the docs weren’t sufficient foundation to authenticate the Yelp posts and tie them to him. 

The trial court actually bought this bs, holding that Plaintiff needed to come forward with expert testimony in order to tie the posts to Defendant through the IP docs. Which effectively resulted in a directed verdict for Defendant on the defamation claim. 

Plaintiff’s appeal raises three grounds to overturn the in limine ruling. The first two are kind of silly. She says that the court’s in limine ruling exceeded the scope of the relief sought in the notice of motion. But all the pertinent issues were fully briefed, including in supplemental briefing ordered by the court, so even if there were some ticky-tack flaw in the notice, Plaintiff got the chance to fully litigate the issues in the trial court. 

She also says that in an earlier ruling a discovery issue—when the case was before a different judge—the court found the IP info germane. She thus says that Code of Civil Procedure § 1008’s limits on reconsideration should have prevented the trial court from granting the motion. But this wasn’t really a reconsideration motion. The discovery ruling was on a limited issue and the reliance on the IP info wasn’t in irreconcilable conflict with the in limine motion. So, even though there might have been some tension or overlap, it wasn’t effectively a reconsideration.

But the trial court did err, big time, in rejecting the evidence based on a failure to authenticate.

The Court of Appeal first addresses the standard of review. It says that when a granted in limine motion precludes evidence such that it effectively results in a nonsuit, a de novo standard applies. And since the directed verdict entered by the court is on the same evidentiary standard as a nonsuit, the court says de novo review applies here too.

I frankly find this a little confusing. Except in very special circumstances (see, e.g., Bose Corp. v. Consumers Union, 466 US 485 (1984)) a standard of review should turn on the nature of the decision made, not its gravity. Evidentiary rulings often entail some combination of law deciding, fact-finding, and reasoned discretionary judgment. In recognition of that, the abuse of discretion standard that usually applies to rulings on evidence bakes in three different tests: A non-deferential de novo test for pure legal questions, such as the interpretation of the Evidence Code. A very deferential substantial evidence test for factual questions, such as the factual foundation for most hearsay exceptions. And a test based on arbitrariness or irrationality for matters of empirically informed, but ultimately value-drive judicial judgment, such as whether prejudice outweighs probative value. If, for instance, the evidence ruling calls for a pure factual finding within the province of the court under Evidence Code § 405—e.g., was there an attorney-client relationship foundational to a claim of privilege—it seems like the deferential substantial test should apply, because second guessing fact findings de novo is not something appellate courts are in a good position to do. Which is true even if the exclusion of the purportedly privileged evidence results in a failure of the claim as a matter of law. 

Interestingly, the California Supreme Court has not resolved the question of whether evidentiary rulings in the summary judgment context get reviewed de novo or for abuse of discretion. See Reid v. Google, Inc., 50 Cal. 4th 512, 535 (2010). But post-Reid cases suggest that the standard turns on the nature of the decision and whether the trial court was in a superior institutional posture position to make it. Because most evidence calls at SJ are decided on a cold paper record, trial courts don’t really have much of an advantage. Thus, cases like Pipitone v. Williams have generally applied the de novo standard. But—contrary to the rationale of the court here—the decision to apply that standard has nothing at all to do with whether the ruling is dispositive of the ultimate result.

And the key case the court cites on this point—Dillingham-Ray Wilson v. City of L.A., 182 Cal. App. 4th1396, 1402 (2010)—is addressing an exclusion of evidence that dooms a claim in a different context. In Dillingham-Ray—and even more clearly in the collected cases cited therein, such as Aas v. Superior Court, 24 Cal. 4th 627, 634–635(2000)—courts ruled that all evidence in support of a particular claim was irrelevant due to an interpretation of a underlying legal issue at stake on a claim. (E.g., your evidence of breach isn’t admissible because the contract didn’t require that kind of performance.) In these cases, the ruling on evidence is basically serving as a shortcut to a “matter of law” motion like a demurrer, summary judgment, nonsuit, jnov, etc., so it makes sense to review it de novo.

But that’s not what’s going on here. The ISP evidence wasn’t rejected because the plaintiff’s claim didn’t encompass a viable theory where it was relevant. It was rejected because the trial judge found that Plaintiff had not established an adequate foundation to show authenticity—i.e., that the ISP documents were sufficient to tie Plaintiff to the posts.

Regardless, the court here was probably right on the standard, but for the wrong reason. The burden to establish the authenticity of a document is addressed by Evidence Code § 403, which requires only a proffer of “evidence sufficient to sustain a finding.” That is, the proponent of the evidence only needs to come forward with enough evidence that would let a reasonable juror find that a document is authentic. While the ultimate fact is a jury question, reviewed for substantial evidence, the sufficiency of the evidence to get there is a legal one. “[I]t is ultimately a question of law whether the evidence, interpreted according to applicable appellate principles, is sufficient to support a ruling.” Oiye v. Fox, 211 Cal. App. 4th 1036, 1067 (2012). And like any question of law, it merits de novo review.

Sorry for the digression. 

On the merits, the trial court applied way too high of a standard. By weighing evidence and demanding expert testimony, it effectively treated authenticity like a § 405 foundation question—where the court decides the foundational issue for itself by preponderance. But authenticity is specifically enumerated as a § 403 foundation question. See Cal. Evid. Code § 403(a)(3). Under § 403, the court is only supposed to ask if there’s enough evidence—crediting reasonable inferences in favor of the offering party—“to sustain a finding.” Cal. Evid. Code § 403(a). Basically, if there is enough to get the question to a jury, § 403 is satisfied.

The trial court, however, drew a slew of inferences the other way. This happens sometimes—judges have a knee-jerk reaction that something that came from the Internet is inherently sketchy. See famously St. Clair v. Johnny’s Oyster & Shrimp, Inc., 76 F. Supp. 2d 773, 774–75 (S.D. Tex. 1999):
“There is no way Plaintiff can overcome the presumption that the information he discovered on the Internet is inherently untrustworthy. Anyone can put anything on the Internet. No web-site is monitored for accuracy and nothing contained therein is under oath or even subject to independent verification absent underlying documentation. Moreover, the Court holds no illusions that hackers can adulterate the content on any web-site from any location at any time. For these reasons, any evidence procured off the Internet is adequate for almost nothing, even under the most liberal interpretation of the hearsay exception rules found in Fed.R.Civ.P. 807.”*

1999 was a long time ago, and the skepticism is fading with time, presumably as more and more jurists become familiar with the online world. Indeed, lots of random internet stuff gets cited in appellate opinions nowadays, albeit usually for background. See, e.g., Browning v. Colvin, 766 F.3d 702, 705 (7th Cir. 2014) (citing internet source for the proposition that “Chimpanzees have been said to be capable of sarcasm[.]”) But the bias still shows up. 

Without that skepticism, if one just looks at what was actually offered to connect the IP addresses to Defendant, it would certainly be sufficient to permit a reasonable fact-finder to infer that Defendant was the poster. So the evidence shouldn’t have been excluded.

And if the evidence were in the record, it would be sufficient to create at least a fact issue that Defendant made a “statement” for the purposes of a defamation claim. Which means that the directed verdict was also improper and the case should have went to the jury.


* Of course, the author of St. Clair (“inherently untrustworthy” internet bio here) turned out to be a bit of a sketchball himself. After a pretty ugly court-employee groping scandal, he got impeached, but quit before his Senate trial. He later pled guilty to obstruction for lying during the investigation and served 2 1/2 years in the federal clink.  

No comments:

Post a Comment