# Ed Morrissey: Does the 9th Circuit really have an 80% reversal rate?

Will Truman

Will Truman is the Editor-in-Chief of Ordinary Times. He is also on Twitter.

### 9 Responses

1. Burt Likko says:

So given that 15.3% of the Ninth’s rulings are taken up for review by SCOTUS at all, an 80% reversal rate means that the real reversal rate is 12.3%.

Note that the median review rate is 10.8%, and the media reversal rate after review is granted is 62.3%. So the Ninth is over the median on both counts.

Of some interest might be cross-indexing the margins of the Ninth’s reversals — we may find that a disproportionate number of them are split decisions as opposed to unanimous or per curiam rulings. I further suspect that for most of last Term and this Term, we’ll find the Ninth’s reversal rate declining significantly (b/c a 4-4 tie vote results in affirming the decision below), at least until such time as Judge Gorsuch becomes Justice Gorsuch.Report

• Nevermoor in reply to Burt Likko says:

You missed a few decimal points there, and I have no idea what you’re trying to calculate. From the article:

The Ninth Circuit accounted for 175 of the cases reviewed, or about 26.5%, but the same circuit handled 114,199 of all appellate cases — 18.9% of the total.

Assuming that’s true, the real reversal rate is 140 out of 114,199 cases, or 0.125%.Report

• Will H. in reply to Nevermoor says:

I’m wondering why the cases where no petition is filed are even being counted toward the reversal rate anyway.Report

• Nevermoor in reply to Will H. says:

It all depends on what question you’re asking.

If the question is “how often is the Ninth Circuit the final word on an issue” the answer seems to be 97.5%.

If the question is “how often does the Supreme Court disagree with the Ninth Circuit” the answer would be (175*0.8)/[# of 9th circuit cases petitioning for cert] (which is going to be lower than 97.5%, but not by all that much).

If the question is “of the 9th Circuit decisions that trouble the Supreme Court, how many are affirmed” then the answer is 20%.

All of which matters because the idea that the 9th Circuit is wrong 80% of the time is silly, but a non-lawyer could easily conclude it from coverage of the last statistic.Report

2. Nevermoor says:

Interesting analysis, though I don’t think “rates of review” correlates with badness, so much as with importance of decisions (which he acknowledges at least to a degree). That’s why you see the DC Circuit (which largely handles major governmental litigation) off-the-charts on review rate, but below median on reversal. What that tells me is that the Ninth Circuit has the lions share of the important non-specialized (DC or Fed Circuit) cases, and thinks about things somewhat differently from the Supreme Court.

Another problem with the analysis that goes unaddressed is that “the 9th Circuit” isn’t a thing the same way other circuits are. Not even en banc decisions involve the whole court, so its entirely possible for the outcome of the Ninth Circuit’s analysis to depend completely upon which judges you draw. Especially on controversial/important issues (political or legal-philosophical).Report

• PD Shaw in reply to Nevermoor says:

I vaguely recall a defense of the 9th circuit at Volokh from about 10 years ago that argued that the 9th Circuit reversal rate was higher than other circuits because the other circuits were better at predicting and anticipating the SCOTUS while the 9th Circuit was better at following past rulings. There was a larger philosophical point about the role of appellate jurisdiction, but my take-away was the court had moved to the right in the early 90s, signaled its willingness to take up certain issues in the future, and the 9th circuit stayed put.Report

3. Oscar Gordon says:

NoOoOoOoOo! Statistics, how could you lie to me again?! I thought we had an understanding?Report

4. Will H. says:

I think this would take a breakdown of what kind of cases are being reviewed & which ones are being reversed to be of much use.
Criminal cases tend to have a much lower reversal rate, iirc.Report