Friday, June 29, 2012

U.S. v. Thoms (9th Cir. - June 29, 2012)

One of the things I do -- or, more accurately, try to do -- when I read opinions is to initially read them without looking first to see who wrote them.  I care about the author, sure.  I definitely look at the end.

But I recognize that I'm subject to bias just like everyone else.  If I know that an opinion is written by a judge with whom I generally agree (and/or respect), I realize that I may potentially view the opinion in a more generous light.  Ditto if the converse is true.

I try to avoid that by skipping over the author at the outset.  Sometimes my eye catches things, of course, and sometimes I can tell the author merely by the style of the opinion.  But I do my best to try to get an "impartial" view of an opinion, at least at the beginning.  Recognizing that it understates things to merely put the word "impartial" in quotes.  But I try.

I mention that because (1) I like to share things, and (2) I had that sense in spades when I read this opinion.  Here are its first four paragraphs:

"In a criminal pre-trial matter, it is well established in our circuit and in most others that, if a district judge is inclined to depart from credibility findings of a magistrate judge that were favorable to the defendant, he may only do so after holding a de novo evidentiary hearing. See United States v. Ridgway, 300 F.3d 1153 (9th Cir. 2002). This right is grounded in the Due Process Clause. Id. at 1155-56.

But what about the reverse? What if a magistrate judge recommends denying a motion to suppress, but the district judge wishes to reverse that determination because he is inclined to disbelieve the testimony of government officials? In that case, must he hold a de novo hearing for the benefit of the government, which is the plaintiff in a criminal case?

This appeal considers such a case, and our circuit has yet to address this question. The government primarily urges us to adopt a broad rule: the government’s right to a de novo hearing before a district judge is identical to that of defendants, even though the government cannot directly invoke the protections of the Due Process Clause. The defendants counter that, without the Due Process Clause to ground this right, the decision to hold a de novo hearing in such cases is entirely within the district judge’s discretion.

We disagree with both of those categorical options, and today we adopt a middle ground, though our rule counsels strongly in favor of holding a de novo hearing. We agree with the defendants that the government does not have an unqualified right to a de novo evidentiary hearing whenever a district
judge reverses a magistrate judge’s credibility determinations in a way adverse to the government. But we also agree with the government that its interest in the integrity and accuracy of judicial proceedings—which, after all, similarly underlie a defendant’s due process rights to such a de novo hearing— will often counsel in favor of such a hearing. Thus, we hold that a district court abuses its discretion when it reverses a magistrate judge’s credibility determinations, made after receiving live testimony and favorable to the government, without viewing key demeanor evidence, with one exception: where the district judge finds that the magistrate judge’s credibility determinations had no legally sufficient evidentiary basis, so that, were they jury determinations, judgment as a matter of law would issue for the defendant."

I had chills when I read that introduction, saying to myself:  "Holy crap.  This is good.  That's seems exactly right to me, and articulated with really good language.  Simply awesome."  A reaction on my part that was immediately followed by:  "Oh no.  I hope this isn't written by a judge I think is terrible!"

I read the rest of the opinion, and then went back and looked to see who wrote it.  Judge Bea.  Joined by Judges Gould and Bybee.  So a really good start by Judge Bea.  Well done.

That's not to say that I agree with every single thing in the opinion.  I'm probably a little less inclined than the panel to say that the district court has to always hear the witness in person if it's not a JML-like situation.  In part because I'm probably a little less confident than the panel that judges can actually make accurate calls on credibility (and I think I'm backed up by the research on this one).  And also in part because I think there's a process value -- somewhat distinct from (albeit intertwined with) the Due Process Clause -- that distinguishes when the government as opposed to a defendant is found uncredible.  There's a value in saying that you can't put someone in jail on the basis of finding him uncredible without hearing him out first.  It'd just seem wrong to do so.  For many of the same reasons we allow allocution before you're sentenced.  It's not just that there is an accuracy concern.  It's also one of process.  But that doesn't apply when you're finding a witness for the government uncredible.  You're not sending the witness (or anyone else) to jail; you're just disbelieving their testimony.  Not as big of a deal.

Plus, sometimes it's hard to find a witness uncredible when they're right in front of you.  And I say this from experience, having been on the bench numerous times in small claims court in which I've had to decide cases on the basis of pure credibility calls.  Usually you can come up with a "creative" way to sell the decision to the parties -- while simultaneously doing the right thing -- without expressly calling a witness or party a liar.  But sometimes you can't.  The pressure not to call someone a liar to their face (even when you think they are) is presumably even higher when the witness is a police officer or other government figure.

So sometimes -- and I admit it's rare -- I might allow a judge to not hold an actual hearing if the judge is convinced that no matter how credible the government witness seemed, there's no way the judge would change his or her mind.  This might even be one of those cases.  Look at what the district judge said for why he disbelieved the government witness:

"To conclude that Investigator Young did smell marijuana from the road, while in his vehicle would require the court to assume that Thoms’ filtration system was either saturated or not functional; that the odor of marijuana left the outbuilding unfiltered and remained warm long enough to stay above the vegetation behind the Thomses’ house; that it either traveled around the Thomses’ two-story residence or stayed warm long enough to traverse above it then suddenly dropped in the area Young claimed to smell marijuana; and that it followed the described 450 foot course without dispersing beyond perceptible levels. Those assumptions are contrary to a preponderance of the evidence presented at the Franks hearing."

That sounds like pretty decent reasoning to me.  In short:  "The story you're telling is just silly.  You thought there was marijuana there for different reasons but those weren't enough for a warrant, so you did what we know officers do all the time:  Said you 'smelled' it.  Sorry.  No dice.  I don't care how sincere or official you looked at a hearing or how smooth your uniform was.  I'm not buying it."

Now, maybe the panel here would say that if that was indeed the case, then the JML standards would apply (since the evidence is "implausible;" i.e., no reasonable jury could buy it).  But I'm not sure that this standard -- which is designed for the Seventh Amendment, an entirely dissimilar context -- is exactly right.  Instead, a district court might say:  "Look, it's not that a reasonable jury might not buy your argument.  It's just that I don't.  No matter how credible you look."  That's a different standard than the Ninth Circuit articulates, and I'm not sure it wouldn't make sense.  Why hold an evidentiary hearing to call someone a liar to their face when you're certain it won't make a difference?  And why require one if you fear (as I do) that having to call someone a liar to their face may in some cases make you hesitant to call them a liar even if you think they are.  That gets the wrong result, and at real cost.

Mind you, I think there's a different reason why I'd normally require district judges to hold evidentiary hearings in cases like this one. Again, a process reason. Sometimes you get more information when you hold a hearing not about pure credibility, but about actual facts. When the witness personally appears, the judge can ask questions. Those questions can either (1) provide additional evidence that the witness is lying (indeed, that's what they'll usually do), as the witness provides increasingly incredible responses to a judge's targeted inquiries, or (2) sometimes (albeit perhaps rarely), convince the district court that what might at first glance appear implausible is in fact plausible. Not because we're making a pure demeanor call or judging credibility. But because the interactive process of questioning reveals additional facts that change one's assessment (e.g., you learn that there was a really strange wind that night, that studies show that one can detect a marijuana smell from hundreds of yards away, etc.). [Parenthetically, since I live in a hippie-like area across the street from a beach, I can testify that the latter is, in fact, more plausible than you might initially think.]

I still end up thinking that the introductory paragraphs of this opinion are really good, and that the result is largely right. But I might nonetheless tinker with both the reasoning and the resulting rule to make it a little better. Because it seems nearly right to me, but the few cases in which it's a near-miss might be important ones.