Possible Revisions to Georgia’s “Super-Plain Error” Rule, OCGA § 17-8-57


Update March 2, 2015: Two bills have been introduced in the General Assembly to amend OCGA § 17-8-57. My post about them is here: Georgia General Assembly Considering New Rules for Judicial Expressions of Opinion. Below is my original post from December 5, 2014.


On November 17, in Rouse v. State, the Georgia Supreme Court reversed a murder conviction because the trial court told the venire “you will be hearing about a case, which is a murder case, that happened in Muscogee County.” These words expressed an “opinion as to what has or has not been proved” in the case, requiring reversal under OCGA § 17-8-57. Justice Nahmias, joined by Justices Hines and Blackwell, dissented. The statute did not require reversal under these facts, they argued; if it did, “the time has come for the General Assembly to seriously consider amending” it.

Code Section 17-8-57 states in full:

It is error for any judge in any criminal case, during its progress or in his charge to the jury, to express or intimate his opinion as to what has or has not been proved or as to the guilt of the accused. Should any judge violate this Code section, the violation shall be held by the Supreme Court or Court of Appeals to be error and the decision in the case reversed, and a new trial granted in the court below with such directions as the Supreme Court or Court of Appeals may lawfully give.

This is an unusual statute. I’ve written a little about it before, here and here. The Court has described Section 17-8-57 as requiring “super-plain error” review: A prohibited judicial expression of opinion leads to automatic reversal of a criminal conviction, even if the expression was harmless and even if the defendant failed to preserve the error by objecting. Strict application of the statute, as in Rouse, requires trial judges to be extremely careful in their choice of words when explaining a case to a jury, and results in reversals for what may seem like insignificant judicial comments.

The Rouse dissenters “urge the General Assembly to consider repealing the second sentence of OCGA § 17-8-57.” That is certainly one way to address the dissenters’ concerns. However, I believe there are other ways of amending Section 17-8-57 that the General Assembly should consider. This blog posts addresses some of the possibilities.

Require harm, but not preservation. Trial court errors are ordinarily reversible only if they prejudiced the appellant’s case. This is the “harmless error” doctrine. The General Assembly could amend Section 17-8-57 to remove the exception it provides to the harmless error doctrine, while retaining its exception to the preservation of error rule. Effectively, this would shift judicial expressions of opinion from “super-plain error” down to “plain error” review. The Rouse dissent highlights several examples of existing plain error statutes in Georgia.

There are good reasons for retaining plain error review for all judicial comments on the evidence (in other words, for allowing reversal based on such comments even if the error was not preserved). Generally, plain error review is appropriate as a policy matter when errors “seriously affect the fairness, integrity or public reputation of judicial proceedings.” Smith v. State, 288 Ga. 348, 358 (703 SE2d 629) (2010) (Hunstein, C.J., dissenting) (quoting Silber v. United States, 370 U.S. 717, 718 (1962)). Improper judicial commentary on the evidence will often satisfy this condition. Such error harms not the appellant alone, but the public interest in the perceived fairness of the judicial system. Thus, it may be appropriate to allow appellate correction of such errors (categorically) even if trial counsel did not object. Indeed, the Rouse dissenters (at slip op. 35) agree that plain error review under Section 17-8-57 may be appropriate as a matter of policy.

Deleting the second sentence of Section 17-8-57 would not necessarily remove judicial expressions of opinion from plain error review. But legislative incorporation of “plain error” into the statute’s text would remove any doubt.

Require preservation, but not harm. Another possibility is amending Section 17-8-57 to retain its exception to the harmless error doctrine, while removing its exception to the preservation of error rule. In other words, a timely objection to the improper judicial expression would be required for appellate review, while prejudice to the appellant would not be.

This would be an unusual, perhaps unprecedented, approach, but it may be helpful to think about why. If an error is of a type that an appellate court will reverse even without harm, the law typically does not require preservation of error either. Most exceptions to the harmless error doctrine arise when the trial court’s order is void ab initio, such as when the trial court lacked jurisdiction. Courts often don’t even speak of review of such errors in terms of “harmless error” or “plain error.”

There may be good reasons (discussed below) for retaining the harmless error exception in Section 17-8-57. But retaining the harmless error exception while removing the preservation of error exception is an unlikely solution.

Keep the “super-plain error” rule, with clarifications or exceptions. The General Assembly could also retain both the harmless error and preservation of error exceptions, while making smaller changes to address the concerns raised in the Rouse dissent. I discussed above some reasons for keeping at least “plain error” treatment. Why might we want to also keep the harmless error exception?

The harmless error doctrine usually makes policy sense. I discuss some of the reasons why, and also some of the criticisms, here. It conserves judicial resources and prevents judicial resolution of theoretical questions, but insulates some errors from appellate review.

The Rouse dissent asks, “Why should we have an automatic-reversal rule when a trial court improperly expresses its opinion on whether the evidence has or has not proved a fact, when we do not have such a rule when the court improperly admits or excludes the evidence that is actually needed to prove that fact?” But I think there is a meaningful difference between these scenarios. When admitting or excluding evidence, a trial court is required to act, one way or the other. The trial court must exercise legal reasoning, and sometimes its decisions will be wrong. In contrast, a trial court is never presented with a situation where it must express an opinion on the evidence. When tempted to express such an opinion, trial judges can (actually must) say nothing. That is not to imply that obeying the rule is always easy, but it seems easier for a judge to stay silent about the proof than to get every evidentiary ruling right over the course of a trial.

And there is another difference between judicial resolution of evidentiary disputes and judicial commentary on the evidence. Incorrect evidentiary rulings may reflect on the competency of the judiciary, but they do not generally reflect on its integrity. They do not implicate societal interests in the dignity of the courtroom or respect for the judicial process. In contrast, judicial comments on the evidence or on the guilt of the accused are more likely to create an appearance of partiality, or otherwise call into question the judiciary’s esteem. This is precisely because such judicial comments are unnecessary to the conduct of the trial. Reasonable observers understand that judges must make tough evidentiary decisions and that some of them will be wrong. But judges should be able to entirely avoid improper comments by choosing their words carefully.

Of course, there are also many policy arguments against the automatic-reversal rule. It is far from clear that the trial court’s comments in Rouse, just to use the most recent example, would cause any reasonable person to doubt the impartiality of the judicial system. I don’t intend to endorse any particular solution. But I do think there are reasonable policy arguments in support of the current statute as interpreted and applied by the Rouse majority.

Clarify the exception for uncontested matters. Assuming the “super-plain error” part of Section 17-8-57 remains in place, one small clarification the General Assembly could make is that the statute does not apply to expressions of opinion on uncontested matters. The Rouse dissenters believed this was already the law; the majority found that the matter was in dispute and did not directly address the issue. Judicial expressions of opinion as to uncontested matters should not result in reversal. This is a change that is probably unnecessary but might be helpful for clarity.

Make an exception for venue comments. Judicial comments on venue seem to generate a disproportionate number of reversals under Section 17-8-57. Rouse is one example. Likely this is because venue often may seem as a matter of common sense to be undisputed (or indisputable), when as a matter of Georgia law it must be proved beyond a reasonable doubt by the prosecution if a defendant pleads not guilty. Because venue is so perfunctory in many cases, trial judges make the mistake of commenting as if it were undisputed. The General Assembly could exclude comments about venue from the “super-plain error” review standard of Section 17-8-57. (Venue comments would still be reversible error, but only if the error was harmful and the appellant preserved it.)

Make an exception for preliminary comments. The statement that resulted in reversal in Rouse was made “about two minutes after [the prospective jurors] first entered the courtroom.” Such preliminary comments seem to be a danger zone for Section 17-8-57 violations. The trial court must give jurors background, without suggesting that any disputed facts have been proven. Inserting “allegedly” into every sentence is awkward. It might make sense for the General Assembly to provide a grace period, perhaps up until the first witness has been sworn, when the automatic-reversal provision of Section 17-8-57 would not apply.

Delete the entire rule. If Section 17-8-57 comes up for legislative review, it is worth considering whether the entire code section, not just the second sentence, should be repealed. Do we need a statutory prohibition against judicial commentary on the evidence? As the Rouse dissent points out, the prohibition is not constitutionally required. Repealing the entire code section would not immunize all judicial commentary from appellate review; a comment so egregious that it violated due process could still be reversible error. Even if we think judicial commentary on the evidence is bad, there might be better remedies for it than making it grounds for reversal. This is a broader issue than I want to address in this blog post, but it is not obvious that we need a prohibition on judicial expressions of opinion.

For better or worse, Section 17-8-57 is an anomaly. I’m glad the Rouse dissenters have brought it to the attention of the General Assembly. I hope to have more to say on this topic soon.