The Sixth Circuit has had lots of badpress over the past few years for its long string of reversals by the Supreme Court, mostly in habeas cases. Over the past term, the Supreme Court has granted certiorari in 11 cases from the Sixth Circuit and reversed in all but 2.  While that sounds bad, it is only slightly worse than the average reversal rate for all circuits, which was 73% this year.  And four circuits fared worse than the Sixth by that metric. Over the past few years the Supreme Court has taken significantly more cases from Sixth and Ninth Circuits. This year fully one-third of the total cases from all circuits came from those two. But this means little by itself.  As we have previously noted, the statistics must account for the fact that the Supreme Court’s decisions on circuit splits effectively overturn the decisions of many circuits.

SCOTUSblog has pointed to three recent papers (here, here, and here) that calculate reversal rates that account for all the circuits on each side of the splits resolved by the Supreme Court.  Each study uses its own timeframe and methodology, but all conclude the true differences in circuit reversal rates are much less important than would appear from the normal “win-lose” statistics cited in legal publications. Surprisingly, the results of each study for individual circuits varied widely.  In one study, the D.C. Circuit had the worst record in one and nearly the best in another; the First Circuit had similar swings. The Sixth and Ninth Circuit also varied, though neither ever rose above seventh place in the standings.

Interestingly, the studies came to wildly different conclusions on whether the Supreme Court usually sides with the majority in circuit splits. One found that the majority approach was affirmed “90 percent of the time” and another found that the Court was just as likely to affirm the minority as the majority. As shown by these three studies, attempting to determine which courts are more likely to be reversed is not an easy task—even when looking after the fact.