Pardon the ambiguity in the headline, there–the CLE seminar was what was “interesting,” not necessarily my thoughts about the seminar, but this CLE seminar was quite thought-provoking. Today was the LSBA’s annual Advanced Appellate Practice Seminar; as always, seminar chairs Thomas Flanagan and Rick Stanley did an excellent job putting together dynamic panels.
And, as always, Judge Stephen Higginson of the U.S. Fifth Circuit Court of Appeals was generous and forthcoming with his thoughts on the methodologies employed by the Court and ways he believes it could be better; for example, less use of the “unpublished” opinion and the “per curiam” designation. The thought that provoked for me was the daily mining of statistics I do here on exactly those issues, and how it plays out in the monthly and annual statistics recaps. So, just to give an idea of the heft of the changes Judge Higginson is referring to, for the most recent annual period I aggregated statistics (October 2024-September 2025), of the 2,424 total opinions in that period, 82.67% were unpublished decisions (2,004 unpublished decisions) and 83% were designated as per curiam (2,012 decisions). And, relatedly, though I’m unsure what to make of this statistic, the affirmance rate in decisions designated for publication (i.e., counting opinions where the district court decision was affirmed in full, the appeal was dismissed, or a petition for BIA or other agency review was denied in full) was only 51.43%, compared to an affirmance rate for unpublished decisions of 87.07%.
Judge Higginson also spoke at some length about the effectiveness of oral argument, which provoked for me the thought about how much more even the playing field is in the Fifth Circuit, statistically speaking, just to have oral argument granted–even if oral argument is subsequently withdrawn by the merits panel; in other words, even before getting to oral argument, the improvement of an appeal’s chances just by making it “successfully” through the screening panel. I don’t have a full year of statistics on that, because I’ve only begun mining for those statistics beginning with this year’s October decisions, but that gives me two months to look at (October and November 2025). For those two months, the overall affirmance rate (affirmed in full, appeal dismissed, or petition for review denied in full) for all decisions at the Fifth Circuit is 89.6% (336 out of 375 total opinions in those two months). Where no oral argument was ordered at all, that affirmance rate jumps up to 95.62% (284 of 297 NOA opinions). Where oral argument was initially ordered but then withdrawn, the affirmance rate drops a statistically significant margin to 72.73% (16 of 22 OA-withdrawn opinions). But where cases keep oral argument, the affirmance rate is much closer to an even chance of modifying the original decision: 56.45% (35 of 62 OA opinions)! Likely, this has more to do with the briefing and the types of cases that a screening panel believes should get oral argument than it does with the actual skill and process of oral argument itself, but still an important factoid–and another thought provoked by today’s excellent CLE seminar.
Take the Fifth 