Wednesday, June 20, 2007

The Home Court Advantage?

Lee Epstein

In an article on the Uttecht death penalty case, Linda Greenhouse highlighted Justice Stevens' dissent, which made an "unmistakable reference" to the "special relationship" between Justice Kennedy and Judge Kozinski. Now a judge on the Ninth Circuit, Kozinski served as a law clerk to Kennedy when the justice was on the Ninth. Because Kennedy, writing for the Court, reversed Kozinski, it seems possible that Stevens was attempting to persuade Kennedy to vote the other way. But Ms. Greenhouse suggests that the reference was less about persuasion than about frustration on the part of the Court's left wing.

Then again, perhaps Justice Stevens was on to something.

A preliminary statistical analysis I conducted with Jeff Segal shows that, overall, former federal appellate court judges on the U.S. Supreme Court are more likely to affirm decisions from their own circuit than from all others. Take Antonin Scalia. Looking across all 1,587 decisions coming out of the U.S. courts of appeals since he joined the Court in 1986, Justice Scalia affirmed about 40 percent and reversed 60 percent. The figures for the D.C. Circuit, where he served between 1982-1986? Nearly the mirror image: 56 percent affirmed and 44 percent reversed. Likewise, over the course of his career Thurgood Marshall reversed federal appellate court decisions about as often as he affirmed them—except when it came to his former court, the Second Circuit. He affirmed better than 6 out of every 10 cases from the Second (162/254, or 63.8 percent).

The relationship holds even after taking into account the ideological direction of the appellate court's decision. If the First Circuit were to reach a decision favoring, say, the defendant in a criminal case or a plaintiff in an employment discrimination suit, the odds are quite high that Justice Breyer—former member of that court—would vote to affirm (about .77). If a similar case were to come from another circuit, odds are that he'd reverse (about .58). Clarence Thomas evinces the same behavioral pattern. He is far more likely to affirm the D.C. Circuit, compared to other circuits, even after controlling for whether it issues a conservative or liberal decision.

More generally, over half the former circuit judges serving since 1953 show a statistically significant tendency to favor their home court. They range from the Eisenhower appointees Harlan and Whittaker to Clinton's two picks, Breyer and Ginsburg. And they run the ideological gamut from Marshall to Thomas.

The list does exclude Justice Alito and Chief Justice Roberts. They have not served long enough for meaningful analysis. It also excludes Chief Justice Burger, who was significantly more likely to reverse the D.C. Court of Appeals, where he had a famously poor relationship with Judge Bazelon.

Finally and interestingly enough, it excludes Justice Kennedy, who—like Burger—turns out to be rather anomalous. Even after taking into account whether the Ninth Circuit reached a liberal or conservative decision, Kennedy tends to reverse his former court. Were the defendant to prevail in a Ninth Circuit decision, as in Uttecht, Kennedy would be more likely to reverse than if the decision came from any other circuit! (For the Ninth, a .78 probability of reversal; for all others, .66.)

So Ms. Greenhouse probably got it right after all. If the past is any indication, Justice Kennedy shows little favoritism toward his former court. Based on our preliminary analysis, however, we cannot at this point say the same of some of his colleagues, past and present.


This is fascinating. What does it say about rationality? For it seems to suggest that there are (irrational?) emotional attachments to former colleagues--or, I suppose in the case of Berger, detachment. This is particularly the case if it overrides ideology (I presume ideological controls were done for all justices [you are explicit on ths only for Thomas]).

Of course, it may have nothing to do with bias or emotion and all to do with the extent the individual influenced, or was influenced by, the ideology/philosophy of particular Court of Appeals in question (assuming you can point to a "philosophy" of a particular court) . . .


But I thought Scalia just interpreted the law and the Constitution as the original authors intended - you mean to suggest that his subjectivity matters?

This is terribly interesting, but Thomas only served on the DC Cir for about a year, how was the sample size big enough? Likewise, Roberts was only on the DC Cir for two years; is that enough time to know that the that results (when he's spent enough time on SCOTUS) will be stastically significant. Can you link to your paper on SSRN?

form over substance.

the bane of justice.

This is terribly interesting, but Thomas only served on the DC Cir for about a year, how was the sample size big enough?

The question is how many cases from the DC Circuit he affirms; thus, the sample size is determined by the length of his Supreme Court service, not the length of his DC Circuit service.

You could, of course, argue that it's implausible that a relatively short tenure on a given court could make much a difference; but that's a point to be made after all the evidence is assembled.

I wonder if the study accounts for the types of cases that tend to come through each circuit? For instance, the DC Circuit's focus on administrative law presumably imparted upon Scalia and Breyer expertise on the subject which could be gained only by their service on the DC Circuit. That would explain a greater rate of affirmance for the home circuit.

Second, I don't buy the liberal/conservative outcome measurement. It's too simplistic, even if being used in the most generalized way, particularly where there are other dichotomies which the justices themselves publicize. With so many fragmented opinions in which the justices write separately to explain how they reached their result, there should be plenty of evidence for a more refined study.

It's a good idea worth the inquiry, but such a stunning result should not be obtained with such blunt instruments.

Steve, you made the point much more eloquently than I did . . . it seems like a certain amount of time would have to pass before the "home circuit" phenomenon holds in their cases; that is, a big enough sample size (e.g., tenure) on the home circuit. Anyhow, thanks for making my fuzzy post clearer.

Well, I feel reasonably comfortable in asserting that home-field be damned, I don't see them overturning much in (what HAS to be) the most Draconian Circuit, the one where I now reside...let's hear it for the 8th!

Pssst, Wollman, Hansen, etc...there's a reason the nation doesn't think the "Heartland" is very bright.

The topic of the thread is likely to be wonderful for an array of dissertations which would have time to groom the raw data further. Some aspects that occur to me would be: how about paybacks, settling old arguments once the judge has acceded to a bench which can overrule the lower bench on which he or she once held a post. Another dynamic that is hidden in the bulk first-pass data would be how the rulings of a judge on a more authoritative bench than in the local region from which he or she came, relate to shifts in the political fulcrum of those regional bodies once the promotion has taken place; for example, two years before nomination to the supreme court, in now-associate-justice A.Kennedy's home state there were two hyperbolic issues on statewide ballots, one pertaining to the bench locally, the other about capital punishment's reinstitution in his home state. Campaigners used rhetoric conjuring a linkage between these disparate matters, whereby these activists tarred three state supreme court judges by critiquing their anti death penalty views; the outcome was the local state supreme court saw three judges recalled in that election, and the capital punishment law was approved by the voters. Those were the pushback years of the Reagan effort nationally to control social movements' progress, so there was a historical congruency to these events. However, looking at A.Kennedy's opinions published from Scotus would be viewed insightfully if taking into consideration these seachanges back home. There is a tempering effect upon justices once appointed to the highest bench in the land, as well. Yet another context with potential for templating against the raw data of decisions of several current Scotus justices would be examination of international law impacts, though various justices are more willing than others to discuss this facet of their perceptions of the ambit of their rulings' impact.

Post a Comment

Older Posts
Newer Posts