Dictum, Judicial Modesty, and Judicial Efficiency:

Last week, I blogged about the Ninth Circuit's reading-material-as-evidence case. Defendant was accused of attempting to seduce a 14-year-old girl. His defense was that he expected all along that the girl was actually an adult (which was indeed the case, though he didn't expect that the adult was an FBI agent) who was pretending to be 14. He thought, he argued, that their exchange was acting out a fantasy, and he was expecting to meet and have sex with an adult woman. The prosecution introduced evidence that the defendant had read explicit stories involving sex with children, in order to show that he really was intending to seduce a 14-year-old.

The defendant raised several arguments: He argued that the evidence was irrelevant, and thus violated Rule 401. He argued that the evidence was an impermissible attempt to infer conduct from past behavior, which violated Rule 404 (and maybe the First Amendment, given that the past behavior involved reading). And he argued that the potential of at least some of the evidence to create unfair prejudice substantially exceeded its probative value, which violated Rule 403. The en banc court rejected his Rule 401 and 404 arguments, but held that the conviction had to be reversed because the trial court had refused to read all the evidence before admitting it, and thus couldn't perform the proper unfair prejudice / probative value balancing required by Rule 403.

Judge McKeown, joined by Judges Pregerson, Kozinski, Thomas, and Berzon, concurred in the judgment, and wrote:

All fifteen judges on the en banc court agree on one undisputed fact and on the single reason to remand this case: the district judge did not read all of the pornographic stories admitted into evidence to weigh their probative value against the potential for unfair prejudice. Nothing more needs to be said.

Instead of simply stopping at deciding the case, the majority goes on at length about whether to overrule Guam v. Shymanovitz, 157 F.3d 1154 (9th Cir. 1998), which has nothing to do with the failure to review the evidence [but focuses on the Rule 404 question], and speculates about how the Federal Rules of Evidence might play out on remand. The bulk of the majority's discourse is dicta. See United States v. Henderson, 961 F.2d 880, 882 (9th Cir. 1992) (defining dicta as language that is "unnecessary to [the court's] holding"). Once the case goes back to the district court, we don't know how it will be resolved. Maybe there will be a plea. Or, if there is a retrial, we don't know what evidence the prosecutor will offer or how the district court will rule. Indeed, after this appeal, the prosecution may well reevaluate the need or admissibility of these salacious stories. Nonetheless, the majority offers a far-ranging discourse on Rules 401, 403, and 404, among other matters. The case can be resolved on a simple principle.

I see no reason to go further. Not surprisingly, the breadth of the majority's opinion prompted Judge Kleinfeld's thoughtful concurrence, which expresses concerns that I share.

Here's my question: Even if the majority's discussion of Rules 401 and 404 isn't quite dictum (and I think it's not, for complex reasons), the court could well have disposed of the case by just reversing on Rule 403 grounds without reaching Rules 401 and 404. But then there likely would have been a retrial, and the trial court would have had to face the very same Rule 401 and 404 questions. If the trial court decided those questions the same way, there'd be another appeal on those issues. The panel would have been bound by Shymanovitz, and would have reversed — calling for yet a third trial. Then there would have been another en banc call; if that failed, and the panel decision had stood, then there would be a third trial.

The result would be extra cost and delay, for the defendant, for the prosecution, for the trial court, and for the court of appeals, which would have had to conduct another 15-judge en banc hearing. What's more, consider how this would apply in other cases where there are several unrelated evidentiary issues, and not just two (here, the judge-didn't-read-it issue and the substantive Rule 401/403/404 issues).

Say the trial court makes five decisions, A, B, C, D, and E, and finding any of them to be an error would require a reversal; and say the court of appeals is firmly committed to "simply stopping at deciding" the minimum required to reverse. The defendant is convicted; the government appeals; the appellate court reverses on A. Now there's trial 2, at which the trial court makes B, C, D, and E (since it still thinks those decisions are right). The appellate court reverses on B. Now there's trial 3, at which the trial court makes C, D, and E. The appellate court reverses on C. This goes on until trial 6, at which the court finally retries the case without A, B, C, D, and E. Wouldn't it have been better for the appellate court to decide A, B, C, D, and E at once, so that the defendant, the prosecution, the witnesses, the trial court, and the court of appeals had to deal with only two trials, not six?

In fact, my recollection is that courts often do deal with other issues beyond the strict minimum needed to reverse, precisely because such issues are likely to come up again on retrial. I don't recall the cases off the top of my head, but I'm pretty sure that they are not uncommon. Is that approach, which is to say the en banc court's approach, the better one? Is Judge McKeown's approach the better one despite this efficiency concern? Is one better in some situations and the other in other ones, and, if so, where should the line be drawn?

Again, recall that this isn't just about abstractions about the judicial process. Some defendants can handle two trials but would be bankrupted by three. Some who would be bankrupted in any case could at least pay for the lawyer of their choice in the first two trials but have to rely on an overworked appointed public defender in the third trial. In this trial, the witnesses likely wouldn't be much traumatized by having to testify again and again, but in trials with a real child victim they would be traumatized. And of course any money and effort expended by courts, prosecutors, and appointed defense lawyers in extra iterations of this case is money and effort that they'll lack for other cases.