Socrates once said, “Four things belong to a judge: to hear courteously, to listen wisely, to consider soberly, and to decide impartially.”


The federal appellate courts have long recognized that a judge is more than “mere moderator” in federal jury trials. Mindful of the continuing duty to be impartial, a federal trial judge may examine a witness when he/she deems it necessary to clarify the trial issues. However, this can prove to be a tricky course to navigate because the judge must scrupulously avoid creating even the impression that he believes the defendant is guilty.


On August 4, the First Circuit Court of Appeals reversed a criminal conviction in which the trial judge improperly intervened through the questioning of witnesses and during closing arguments in a manner that bolstered the prosecution’s case. The case, United States v. Rivera-Rodriquez, was a massive drug conspiracy case involving 64 defendants. The thrust of the prosecution’s case against Carlos Rivera-Rodriquez came from two cooperating defendants who had already pled guilty pursuant to plea agreements. The prosecutor questioned each witness about his plea agreement during which the trial judge intervened by telling the witnesses that if they did not testify truthfully, they could be charged with perjury, making false statements, or obstructing judges. The judge also told the witnesses they could receive a sentence beyond the range they had agreed to in their plea agreements, if they committed perjury.


The appeals court found that the cumulative effect of this intervention, as well as the judge’s intervention during closing arguments, created more than an appearance of bias. This finding is exceptional because federal judges enjoy such broad discretion in their decisions to intervene in the trial process. To qualify for relief, a defendant must make a showing of what the First Circuit called “serious prejudice” in order to secure a reversal of conviction. Prejudice is difficult to show. For example, in ineffective assistance of counsel cases, the defendant must also establish prejudice, i.e. that the guilty verdict would have been different had it not been for counsel’s errors. In effect, the trial results must be reliable.


Unlike ineffective assistance claims where the defendant bears the burden of showing prejudice, a showing of ordinary errors, preserved by contemporaneous ojections, such as evidentiary ruling, shifts the burden to the prosecution to show that this trial error was not prejudicial. But there is a strong caveat to this rule: when this trial error—the appearance of judicial bias—involves “interventions” by the trial judge, as in the Rivera-Rodriquez case, the burden to demonstrate “serious prejudice” remains with the defendant.


So what constitutes “serious prejudice”? Prior to the Rivera-Rodriquez case, the First Circuit had no occasion to explicitly define the term in relation to an improper intervention claim. In defining the term for the first time, the appeals court elected to follow the prejudice standard adopted by the U.S. Supreme Court in ineffective assistance claims. Namely, “improper judicial intervention ‘seriously prejudiced’ a defendant’s case when we find that there is a reasonable probability that, but for the error, the verdict would have been different. Moreover, in cases with multiple judicial interventions [as in this case], determining the appearance of bias and the prejudicial effect of that bias generally involves a cumulative effect inquiry.”


This inquiry requires the appeals court to examine whether the intervention through the questioning of witness by the judge “exposes bad facts, inconsistencies, or weaknesses in the case itself.” The First Circuit said the prejudice flowing from this kind of intervention is not “worrisome prejudice.”


The court said it is concerned only when the prejudice is so “problematic” that it gives the “jurors the impression that [the judge] has an opinion on the correct or desirable outcome of the case.”


For the most part, the rule of prejudice did not exist in American jurisprudence until sometime around 1920. State courts began to apply the rule in order to end the practice of criminal convictions being reversed for minor, often trivial reasons. In 1946, the Supreme Court in Kotteakos v. United States pretty much adopted the same prejudice rule, expressing concerns that some criminal trials had been reduced to “games” and that his “gamesmanship” created “widespread and deep” concern about the integrity of the trial process.


State and federal courts, however, did not apply the prejudice rule to constitutional errors—such errors always required reversal until 1967 when the Supreme Court in Chapman v. Ohio held that even certain constitutional errors could be considered “harmless” thereby opening the door to the rule of prejudice as the court did with ineffective assistance claims in 1984. Today, Federal Rule 52(a), Rules of Criminal Procedure, codified the “harmless error” rule by providing: “Any error, defect, irregularity, or variance which does not affect substantial rights shall be disregarded.”


What the rule of prejudice and the harmless error doctrine means is this: a person accused of a crime enjoys a constitutional right to a fair trial, but a fair trial does not mean an “error free” trial. Thus, a federal judge can engage in some forms of prejudicial intervention, but a reversal is required only when the judge creates an impression that he believes the defendant is guilty.