State v. Ellis

Justice Himonas, concurring:

¶ 51 I concur in every aspect of Associate Chief Justice Lee's opinion. As expressed by the Associate Chief Justice, it was obvious error for the district court to find Ms. Thomas unavailable. And that error wasn't harmless-had Ms. Thomas's preliminary hearing testimony been excluded, or had the trial been postponed until Ms. Thomas could appear and be subject to cross examination, there's a reasonable likelihood that the jury's verdict would have been different.

¶ 52 I write separately to highlight an additional consideration militating against finding the district court's error harmless. In the course of asking the district court to admit Ms. Thomas's preliminary hearing testimony, the line prosecutor assigned to this case stated that, in his view, Ms. Thomas was "a key witness." He called her an "essential witness[ ]" and averred that she was so critical to the State's case that trial could "not proceed without [Ms. Thomas's testimony because] ... the case is just not viable without [that testimony]."

¶ 53 On appeal, however, the State urges us to find that admitting Ms. Thomas's preliminary hearing testimony was harmless error. That is, the State urges us to conclude that the prosecutor below-an experienced practitioner who had lived closely with this case in the run up to trial-was mistaken when he told the district court that the State could not, in good faith, proceed to trial without Ms. Thomas's testimony.

¶ 54 There is, of course, nothing improper about how the State has proceeded here. Far from it-the State ably presented its argument for harmlessness on appeal, and there's absolutely no indication in the record that the trial prosecutor misrepresented his view of the importance of Ms. Thomas's testimony. Additionally, there are surely many circumstances in which we should not hold the State to the line prosecutor's pretrial assessment of a witness's importance. The evidence after trial almost always looks different than it looked before-sometimes radically so, sometimes (like here) only in the details of what was elicited at trial-so it's possible that evidence a prosecutor might think critically important before trial has started appears insignificant after all the evidence is in.

¶ 55 But when a prosecutor has stated their belief that evidence is important, I'd tread carefully before finding any error in admitting it to be harmless. I'm concerned about the effect a finding of harmlessness under these circumstances would have on both the public's perception of fairness and the incentives governing prosecutorial choices.

¶ 56 As I've said, the State's conduct in this case is beyond reproach. But what message would it send the public (including those caught up in the criminal justice system) if our trial courts were to regularly admit testimony based, in part, on the State's representation that it's crucial, only to have our appellate courts affirm the resulting conviction because the error in admitting the testimony was harmless? "Yes, the evidence was absolutely crucial before your conviction. And, yes, it was error to admit it. But the difference now is that you've been convicted. Now it's clear the evidence was unimportant. Now the error is of no consequence at all." This smacks of Kafka.

¶ 57 And what about future prosecutors? Every time we declare an error to be harmless, the potential effect is to "encourage the State to [solicit] it with impunity." Harris v. State , 790 S.W.2d 568 , 587 (Tex. Crim. App. 1989) (en banc), overruled by Snowden v. State , 353 S.W.3d 815 (Tex. Crim. App. 2011) ; see also Goodwin v. State , 751 So.2d 537 , 546 (Fla. 1999) (harmless error analysis should proceed with an eye toward "providing an incentive on the part of the State, as beneficiary of the error, to refrain from causing error to occur in the trial of a case"); cf. In re Janet S. , 305 Ill.App.3d 318 , 238 Ill.Dec. 700 , 712 N.E.2d 422 , 424 (1999) ("To conclude that [the State's failure to make an effort to locate exculpatory evidence] is harmless error could encourage the State to avoid looking for such documents and then argue after the fact that such a search would have been a waste of time."). Ex ante, the more willing an appellate court is to find an error harmless despite a prosecutor's contrary representations to the trial court, the greater the incentive for prosecutors to make such representations-representations that cannot help but have an effect on the likelihood that a court will erroneously allow the evidence in. We should take care to ensure that the doctrines we apply on appeal don't create an incentive structure that's reasonably likely to increase the overall incidence of error below.