In 2005, when he's a magistrate judge, Judge Larson reviews Dawson's habeas petition and recommends its denial. Dawson then objects and seeks review (as is his right) by an Article III district court judge. In 2006, who in the Central District is assigned to review the objections to Judge Larson's recommendation? Yes, Virginia. None other than Judge Larson, who's now been elevated to the district court. Who, shockingly, agrees with his own recommendation.
Judge O'Scannlain says that's not reversible error. I understand his reasoning, but don't think I'm persuaded. Sure, Dawson got an Article III judge to take a look and make findings. But it seems to me there's an implicit temporal element there that's not satisfied here.
Imagine, for example, that after Dawson filed his objections, the Clerk for the Central District said "You already had someone who's now an Article III judge look at this. Denied." Surely that would be reversible error, right? You're entitled to have someone look at it who's an Article III judge at the time, it seems to me. To take another (I hope) uncontroversial example, if Judge Larson had been elevated to the Ninth Circuit or Supreme Court, surely it wouldn't be okay to have Judge Larson merely say "Yes, I agree with what I said as a magistrate judge" and thereby satisfy the requirement of Article III review. Even though, yes, he was an Article III judge at the time of the review.
If I'm correct about those two examples, then it seems to me that the same may well be true here. Sure, Judge Larson arguably "re-reviewed" his own findings as a magistrate judge. But even if reviewing your own findings technically satisfies the requirement of de novo review in some formalist doctrinal sense (i.e., "I didn't give my prior factual findings any deference"), it hardly accomplishes what we're looking for here, which is review by a qualified Article III judge who lacks investment in the findings.
Surely it wouldn't be okay if Judge Larson had, say, written a brief opposing Dawson's habeas petition while he was a private lawyer and then adjudicated the habeas petition himself once he was appointed to the district court, right? Even if Judge Larson had written (as here) a final report that was somewhat different in form than what he had previously authored, we wouldn't find that permissible. Ditto if Judge Larson's decision was reviewed by the Ninth Circuit and the panel included Judge Larson, either sitting by designation or after being elevated. No good, even though you Dawson would technically have gotten a fully qualified Article III appellate judge.
So I think that Judge O'Scannlain views this problem too narrowly. It seems to me that there's more at stake here than simply formal review by somewhat who now has Article III capacity.
Separately, even if I'm wrong, I think that Judge O'Scannlain is far too tentative when he says that "in our supervisory capacity over the district courts of this Circuit, we suggest that district courts avoid assigning new district judges to cases they handled as magistrates." I appreciate (and obviously agree) with that sentiment, but why merely "suggest" it?! At a bare minimum, I'd change the word "suggest" to "direct." The district judge and the magistrate being reviewed should not be the same person. Period.