Some Thoughts on the Supreme Court’s Reversal Rate

supremecourt6.jpgEvery term, commentators attempt to predict the outcomes of the cases in the Supreme Court docket. The statistics, however, suggest that the betting person’s answer should be reversal. According to a recent article in Slate, and based on SCOTUSBlog data (2004 term, 2005 term, 2006 term):

Overall, this past term the Supreme Court reversed 75.3 percent of the cases they considered on their merits. The pattern holds true for the 2004 and 2005 terms as well, when the Supremes had overall reversal rates of 76.8 percent and 75.6 percent, respectively.

It is interesting how remarkably constant the reversal percentage is — 75%. It suggests that the Supreme Court primarily takes cases it wants to reverse, with only a few exceptions. Assuming the Court takes about 70 cases a term, it will only affirm in about 17 of them. So perhaps the new game for commentators should be listing those 17 lucky cases that will get affirmed.

Much ado has been made of the 9th Circuit’s dubious honor of having the most cases reversed or vacated — 19 out of 22 this term. But the 9th Circuit decides an incredibly large number of cases, and it typically has the most cases of any circuit before the Court. So it is likely to be victorious every year in terms of numbers of cases reversed. One must look at its reversal percentage to get a better picture, and it typically exceeds the average of about 75% by being between 80 and 90%. However, as the Slate article points out, in 2004 and 2005, “the 9th was reversed 84 percent and 88.9 percent of the time, or about a case or two more each year than it would have been if it had conformed to the reversal rate of the other circuits.” In other words, the 9th Circuit’s bad reversal reputation is largely earned because it’s big.

I wonder whether the Supreme Court’s reversal rate of 75% is a recent phenomenon or existed throughout its history. Does anybody know of where to find stats on reversal rates beyond recent times?

Another interesting statistic, but surely not one easily compiled, would be to examine the percentage of cases contrary to Supreme Court precedent that lead to a grant of cert. Judging from the 9th Circuit statistics, it decides about 6000 cases per year, and the Court takes about 20. That’s 0.3%. If that rate holds true generally — that the Court is taking just a fraction of one percent of cases, it can mean one of two things: (1) the Court is ignoring many cases in which lower courts depart rather overtly from Supreme Court precedent; or (2) lower courts are doing a remarkable job of staying in line with the Court.

It is doubtful that a statistic exists for cases that depart overtly from Supreme Court precedent and their likelihood to result in a grant of cert, and such a tally would necessarily involve some subjectivity, but it would be interesting to find out. Assuming that the Court takes about 70 cases per year, and reverses in 75% of them, that’s only about 53 cases. Many of those cases are reversed not because lower courts clearly contravened Supreme Court precedent, but for other reasons such as circuit splits or issues of first impression. Only a handful, then, are reversed for involving significant departures from Supreme Court precedent. Can it be that there are only a handful of such cases? Or is the Court just ignoring the others?

You may also like...

7 Responses

  1. Mike Dimino says:

    Lee Epstein, et al., The Supreme Court Compendium 244-45 (4th ed. 2007) has the data for the 1946-2004 Terms. There have been reversal rates over 70% in the 1962-65, 1968, 1983, 2001 and 2003 Terms, with many Terms in the high-60s.

    It stands to reason that there would be more reversals than affirmances if, as I believe to be the case, the Justices take cases either (1) to reverse or (2) to resolve a conflict. If they take equal numbers of those cases, and if the conflicts cases yield an equal number of reversals and affirmances, one would expect a 75% reversal rate.

  2. These are all interesting trends and questions that political scientists have researched in useful ways. Although it is not my personal cup of tea, the names that came to my mind – from the grad school days – were Tannenhaus, Ulmer and Songer. Their statistical research focused on predicting whether or not the Supreme Court would grant cert. Looing over 40-50 years of data, they found the Court was more likely to grant cert when a lower court issued a ruling in conflict with a Supreme Court decision or if there was conflict between circuits. Cert is also much more likely to be granted when the federal government is party to a suit. I think more recent studies note the importance of the presence of a civil rights or liberties claim.

    But you are right with respect to subjectivity. Ulmer tried to distinguish between direct and partial conflict and made some other distinction between conflict that was “weak” or “strong.”

    Still, the working generalizations in political science remain (and have long been): Most federal cases begin and end in the district courts. Most appeals result in the decision of the lower court being uhpeld. If the Court decides to give a case further consideration, the odds of reversal increase significantly.

  3. barkleyg says:

    So, if Sonia Sotomayor had 6 cases, out of over 300 decisions, heard before the Supreme Court, and 3 were over tuned, that would mean that her 50% reversal rate would be way higher than the Supremes usual 75% reversal rate.
    And the REPUGS have the nerve to question her brains and if she is competent to be a Supreme Court!


  4. medlaw says:

    “In other words, the 9th Circuit’s bad reversal reputation is largely earned because it’s big.”

    Sorry but that’s not what the numbers say. The percentages say that the 9th circuit is getting reversed at a significantly higher rate than the other circuits taken as a whole. I don’t have a full data set but let’s work off the numbers cited in the article above.

    70 cases in an average term
    reverse rate of 75%
    70 x .75 = 52.5 (I know we can’t have a 1/2 reverse but let’s just go with it for simplicity sake)
    19 out of 22 9th cir. reversed
    which means,
    33.5 out of 48 for the rest of circuits.

    Assuming the above numbers, the reversal rate is 86.3% for the 9th circuit and 69.7% for the other circuits. That’s a statistically significant reversal rate … and one that is not due solely to the size of the circuit (i.e., the number of 9th circuit cases going up on appeal).

  5. wayne says:

    The sentence “In other words, the 9th Circuit’s bad reversal reputation is largely earned because it’s big.”
    seems to be ambiguous since it is not clear whether “it” refers to the court or to its reversal rate.