Members of the public–specifically, Kentel Weaver’s family–were excluded from the overcrowded courtroom during jury selection for his trial. Violations of the Sixth Amendment right to public trial have been called structural errors not susceptible to harmless error analysis. But Weaver’s lawyer didn’t object, so this is an ineffective assistance claim, which of course requires him to show prejudice. But how do you show that you were prejudiced by a structural error–after all, the term refers to an error whose “effect … cannot be ascertained”? United States v. Gonzalez-Lopez, 548 U.S. 140, 149 n.4 (2006).
To the majority here, that rhetorical question is a bit misleading. The opinion looks back over the Court’s structural error precedent and discerns that there are really three sorts of structural errors, or perhaps three different features that, alone or in concert, can render an error structural. There are those errors, as noted above, whose effect is too difficult to ascertain–for these errors, like denial of a non-indigent defendant’s right to choice of counsel, it makes little sense to ask the government to show harmlessness beyond a reasonable doubt. Then there are errors–like the wholesale denial of counsel–that categorically render a proceeding fundamentally unfair. Finally, there are those errors that violate some right which serves not to protect the defendant from unjust conviction, but to protect some other interest. The court gives the right to self-representation as an example of this category. Slip op. at 6-7.
This trisection of structural error is an innovation, and this segment of the opinion will be required reading for anyone wrestling with the doctrine. Things get a little arbitrary from here though. Weaver argued that the denial of a public trial is inherently fundamentally unfair. He further submitted that “fundamental unfairness” is an alternative way to satisfy the Strickland prejudice prong (that is, if an error renders the proceedings fundamentally unfair, the defendant need not show a reasonable probability of a different outcome). The majority assumes that Weaver’s reading of Strickland is correct (and could, perhaps, be read to endorse it). It determines, however, that violations of the public trial right do not automatically lead to fundamental unfairness, instead fitting into one or both of the other two structural error categories. And since Weaver hasn’t attempted to show his trial might have gone differently had his family been allowed into voir dire, the Court holds that the procedural posture of this case–it’s an IAC collateral attack–means that interests of finality must triumph. So, if Weaver’s trial counsel had objected to the closure, this would be a reversal, but since he didn’t, it’s an affirmance.
Justice Alito, joined by Gorsuch, concurs, arguing that the whole concept of structural error is irrelevant in an ineffective assistance claim. In his view, absent Strickland prejudice, there is no “error” at all, because the defendant has not been denied the Sixth Amendment counsel right. It thus makes no sense as whether this non-error was “structural.” He distinguishes structural error from the conceptually similar category of ineffective assistance claims amounting to a “complete denial” of the right to counsel. He also disdains the majority’s flirtation with a “fundamental unfairness” route to showing Strickland prejudice.
Justice Breyer dissents, joined by Kagan. In his view, all structural errors share the characteristic of being difficult or impossible to assess for harm. This includes the right to a public trial, the benefits of which the Court has called “frequently intangible, difficult to prove, or a matter of chance.” Thus, he says,
I do not see how we can read Strickland as requiring defendants to prove what this Court has held cannot be proved. If courts do not presume prejudice when counsel’s deficient performance leads to a structural error, then defendants may well be unable to obtain relief for incompetence that deprived them “of basic protections without which a criminal trial cannot reliably serve its function as a vehicle for determination of guilt or innocence.” Neder, supra, at 8–9 (internal quotation marks omitted). This would be precisely the sort of “mechanical” application that Strickland tells us to avoid.