To read on the Washington Post website click here

By Nancy Gertner

It is not unusual for judges to intervene in court proceedings from time to time — to direct the lawyers to move the case along or to admonish them that evidence is repetitive. The judge's role is to act not as a "mere moderator," as the Supreme Court noted in Herron v. Southern Pacific in 1931, but as the "governor of the trial" responsible for assuring the proper conduct of all participants.

The performance of U.S. District Court Judge T.S. Ellis III in the trial of Paul Manafort on bank fraud and tax evasion charges has been decidedly unusual.

During the trial, Ellis intervened regularly, and mainly against one side: the prosecution. The judge's interruptions occurred in the presence of the jury and on matters of substance, not courtroom conduct. He disparaged the prosecution's evidence, misstated its legal theories, even implied that prosecutors had disobeyed his orders when they had not.

Under the Code of Conduct for U.S. judges, a judge is supposed to be fair and impartial, as well as "patient, dignified, respectful and courteous" to those in his courtroom. The rule's concern is as much about the appearance of justice as its reality. If the judge violates that rule and a defendant is convicted, there may be a trial remedy — an appeal.

But there will be no appeal available to address Ellis's anti-prosecution bias if Manafort is acquitted by the jurors, who began deliberating on Thursday. The prohibition against double jeopardy precludes it. And if President Trump's former campaign chairman is convicted despite Ellis's interventions, the judge's hostility toward the prosecution will have been irrelevant.

For now, we have only the extraordinary evidence of Ellis's conduct during the 12-day trial. The judge continually interrupted the prosecution's questioning of witnesses, prompting lead prosecutor Greg Andres to pointedly note: "Your honor stops us and asks us to move on." Ellis pressed the prosecution to rush through testimony about important financial documents. He made critical comments about prosecution evidence and strategy — all in front the jury.

Ellis also questioned the relevance of Manafort's work as a political consultant for Russian-backed politicians in Ukraine, for which he was paid tens of millions of dollars from 2010 to 2014.  But if Manafort didn't disclose some payments because he was not registered in the United States as a foreign agent, it would provide a motive to hide the amounts from the U.S. government — just what the trial was about. Ellis chided prosecutors for eliciting testimony about Manafort's lavish lifestyle, but that kind of testimony is also a classic element in a tax evasion case. That your cars, boats, condos and clothing suggest you made much more income than you reported would surely be relevant.

After prosecutor Uzo Asonye questioned a bank employee about Manafort's failed attempt to obtain a $5.5 million construction loan on a Brooklyn brownstone, the judge — unprompted by a defense objection — declared: "You might want to spend time on a loan that was granted." The comment strongly implied to jurors that the prosecution was wasting their time. But an attempt to defraud was part of the conspiracy count in the indictment; false representation to secure a loan, successful or not, is itself a crime.

Despite appearances, Ellis doesn't think he is infallible. He admitted on Aug. 9, "I may have made a mistake" the day before when he berated prosecutors for allowing the government's tax-expert witness to sit in the courtroom during other testimony. Earlier in the trial, prosecutors had explicitly sought and obtained Ellis's permission for the tax expert to be present while other witnesses testified.

The Manafort trial is the most prominent case to emerge so far from the investigation by special counsel Robert S. Mueller III into Russia's interference in the 2016 presidential election. Clearly worried about its outcome, prosecutors went so far as to urge Ellis to tell the jury, in his instructions before they began deliberating, not to let his commentary affect their decision-making. Ellis essentially did just that on Wednesday.

It may well have been too little, too late. The potential for judges to influence juries is so great that courts often caution jurors against reading too much into a judge's subtle nonverbal behaviors, a nod of the head, a smile or a frown. Jurors do it anyway. Legal scholars study whether judges' conduct on the bench influences juries — and reliably find that it does. The research typically concerns a judge's anti-defendant bias. Defendants have all the protections of a trial — the high standard of proof, the presumption of innocence, the evidentiary rules designed to protect them. What will be the impact of Ellis's anti-government bias, which was not nonverbal and hardly subtle? Hard to tell. We know next to nothing about this jury. Seeking to expedite matters, Ellis impaneled the Manafort jury in a single day.