Irons v. Carey:
Two years ago, the Ninth Circuit panel of Judges Reinhardt, Noonan, and Fernandez created quite a stir when they strongly hinted that they were going to strike down 28 U.S.C. 2254(d)(1) of AEDPA, the 1996 statute that limits the power of federal courts to provide habeas relied for state prisoners. Under this statute, federal courts can grant habeas relief only when state courts really messed up; if the legality of the state court processes was a close call, the federal courts cannot intervene.

  The Ninth Circuit and the Supreme Court have been playing a bit of a cat and mouse game over the last few years with this particular statute. A number of Ninth Circuit judges (Reinhardt included, of course) opted to ignore it or minimize it to see if the Supreme Court would notice; the Supreme Court noticed, and reversed the Ninth Circuit (and Reinhardt in particular) in a string of cases.

  The interesting question was whether Reinhardt and Noonan were going to take things up a notch and strike down the statute outright, effectively forcing the Supreme Court to add a case to its docket and respond to their view that AEDPA interferes too much with the judicial function.

  Today the panel finally handed down its opinion, Irons v. Carey. Somewhat to my surprise, the panel opted against that course. The judges filed a short unanimous opinion resolving the case and noting that Ninth Circuit precedent upheld the constitutionality of the provision years ago. (The judges were therefore bound by the earlier panel, although in the Ninth Circuit such rules are followed somewhat sporadically.) Next, Judges Noonan and Reinhardt made their arguments that the statute should be deemed unconstitutional in non-controlling concurrences rather than in the majority opinion. Judge Noonan wrote the main concurrence; Judge Reinhardt wrote a short concurrence agreeing with Noonan's concurrence; and Judge Fernandez wrote a short concurrence disagreeing with the other concurrences and noting that none of the concurrences were binding.

  Hat tip: How Appealing.
I appreciate the judges' policy concerns, but Judge Noonan's concurrence was pretty awful on the separation of powers point. I see no reason why Congress cannot limit the scope of a court's review. AEDPA doesn't deny the Ninth Circuit the use of its own precedents. What it does is limit the scope of the habeas right to a small class of errors. Congress does this sort of thing all of the time in other contexts.

Judge Reinhardt's concurrence made a few good policy points, and Judge Fernandez properly preempted any mischief that might later result from Judge Reinhardt's expression of agreement with Judge Noonan.

I'm pleasantly surprised by the result in this case.
3.6.2007 3:09pm
Hans Bader (mail):
The Reinhardt and Noonan concurrences suggesting AEDPA is unconstitutional are mistaken.

AEDPA's requirement that conduct violate clearly-established law to justify relief is no different from the doctrine of qualified immunity, which denies damages to people injured by constitutional violations in civil cases unless the constitutional violation is clearly established.

In many, many cases, the effect of qualified immunity is to deny ALL relief to the injured litigant, since alternative (injunctive) relief is often unavailable due to mootness or other jurisdictional limitations on relief.

By contrast, AEDPA does not limit relief on direct review in state courts at all, but rather focuses on limiting duplicative challenges brought in federal court after the same challenges have already been rejected by state courts.

(Note that the Supreme Court has said that qualified immunity is rooted not in the Constitution itself, but rather is an implied judge-made exception to liability fashioned for policy reasons, so it stands on no stronger a footing than the AEDPA, which is based on an express federal enactment (federal laws override judge-made rules such as rules of civil procedure)).
3.6.2007 3:13pm
Doogie H:
Wasn't much of the "clearly established rule" doctrine for habeas already there because of the Teague case even before AEDPA put a similar requirement into statute?

While I think AEDPA and the anti-habeas provisions are abominable, its difficult to claim that a statute controlling the jurisdiction over lower courts is unconstitutional when the jurisdictional standard provided parallels Supreme Court precedent. I hope I'm wrong on this, because I'd really like Reinhardt to be right.

I guess we'll have to first prove that the executive can't unilaterally revoke habeas at the drop of the hat before we even worry about the AEDPA standard, though.
3.6.2007 3:26pm
Zoe1 (mail):

I think Judge Noonan's very lame response would be that qualified immunity is different because it is judge-created, and it's part of the judicial function to decide cases. Lame, but that seems to be his argument.
3.6.2007 3:27pm
Kovarsky (mail):
reinhardt concurred with his own opinion to avoid the perception that the 2-1 agreement on noonan's opinion was binding.

basically, the argument in this case is about whether 2254(d) is really a meta-rule of interpretation and, if so, whether it's constitutional.

reinhardt decides it on the narrowest possible grounds. the parole due process standard is that the parole decision be supported by "some evidence in the record." but while the rule is federal, it's content is determined by state - i.e. the definition of "some evidence in the record" is determined by reference to state law. obviously the state cannot define "some evidence" so narrowly as to destroy the right, and that is what the two cases the opinion relies on are about. the state case says what the term "some evidence" means for california prisoners (a definition which, in turn, the federal law incorporates). and the federal case (sass) says that this definition does not fall below the constitutional minimum.

noonan writes separately to talk about why 2254(d)(1), which requires courts sitting in habeas jurisdiction to abide by the state decision if it is not "contrary to or an unreasonable application of" clearly established federal law (as defined by the supreme court). noonan's concurrence is a bit oddly placed, because it appears that all three judges agreed that, even if 2254(d)(1) prescribed a de novo standard of legal review, the parole determination would still be constitutional. maybe i am misinterpreting the court on this point, but i don't think so. noonan seems to think 2254(d)(1) is a meta-rule (like a Teague bar), and that congress can't do those.

noonan's concurrence is odd because not only is she bound by prior precedent, as she says, but it seems this doesn't present the sort of unconstitutional-but-reasonable scenario that would occasion consideration of 2254(d)'s constitutionality.

Fernandez acknowledges that the other two have aligned on the opinions strangely to avoid setting a conflicting precedent, although he crypically states that "at least for now" they are upholding the constitutionality of 2254(d). it is, however, worth noting how many judges - both conservative and liberal - express their contempt for certain precedent as dicta or in concurrences. i've seen it on a number of cases in abortion and atkins decisions, for example.
3.6.2007 3:31pm
Kovarsky (mail):
Doogie H,

Wasn't much of the "clearly established rule" doctrine for habeas already there because of the Teague case even before AEDPA put a similar requirement into statute?

A few differences. (1) Teague is thought of as an affirmative defense; 2254(d) is a substantive limit on relief. (2) Teague applies to all prisoners; 2254(d) just to state prisoners; (3) the Teague standard for retroactivity is "dictated by precedent;" 2254(d) requires that a law be clearly established; (4) under Teague a federal court can consider circuit precedent; 2254(d) limits a court sitting in habeas jurisdiction involving a state decision to supreme court precedent.

While I think AEDPA and the anti-habeas provisions are abominable,

quite so.

its difficult to claim that a statute controlling the jurisdiction over lower courts is unconstitutional when the jurisdictional standard provided parallels Supreme Court precedent.

Noonan's point is that Congress can't supply an interpretive rule, a "meta-rule" if you will, that directs courts how to evaluate precedent, etc. I know the habeas statute very well, but I'm not as familiar with the law involving interpretive as opposed to substantive directives.
3.6.2007 3:37pm
Kovarsky (mail):
obviously the reference to Noonan should be "he."
3.6.2007 3:39pm
Trevor Morrison (mail):
"although in the Ninth Circuit such rules are followed somewhat sporadically"

Not sure that's a fair characterization. Maybe "such rules aren't always followed with perfect uniformity" (a statement that would be true of all courts of appeals), but "somewhat sporadically"? Hmmm.

[OK Comments: Trevor, I thought "somewhat sporadically" was about right.
"Not with perfect uniformity" would be technically accurate, but might leave the false impression that the problem is human imperfection rather than a rather loose sense of precedent out there in the Pacific time zone. I thought "somewhat sporadically" made the point more accurately.]
3.6.2007 4:02pm
Justin (mail):
As Kovarsky so ably put it, there's a difference between a jurisdictional act (i.e., what the courts can and cannot look at), and an act of interpretation (i.e., telling the court how to interpret cases, or what standards to use). The Supreme Court has previously struck down at least one statute (can't remember what the case was called or what was about) where Congress basically tried to overrule a previous Supreme Court case by statute, where the question was one about standards of review rather than substance.

I tend to agree with Noonan and Reinhardt, at least based on the substantial difference between wthe rights implicated and the procedure required here and under Saucier v. Katz, and subject to Teague (or to Teague being overruled in favor of a better limit on habeus, as I briefly discussed in a prior post on that issue). Congress can, subject to the substantive Constitutional right, shut down Habeus for all cases, or for certain substantive lines of cases, but it cannot simply ask the Court to "be more conservative" about the issue. And since the right under HC is not whether the state recklessly or intentionally took away a person's freedom, but simply whether they did so unlawfully, the distinction cannot be deemed substantive. By Congress imposing that standard, I am concerned they are not just limiting, but also REDEFINING, a constitutional right - for this they lack the authority.
3.6.2007 4:05pm
Justin (mail):
Also, "somewhat sporadically" happens to be an incorrect usage. Sporadically includes a definition of "how much" or "how often." If somewhat was meant to have a similar emphasis, it is redundant - if it was meant to limit the word, it is inherently contradictory.

[Ok Comments: Justin, I don't follow you. "Sporadic" means "appearing or happening at irregular intervals in time." Thus, "somewhat sporadic" would seem to mean "appearing or happening at somewhat irregular intervals in time." Why is that inherently contradictory? What am I missing?]
3.6.2007 4:06pm
Kovarsky (mail):
without getting involved in the debate over the meaning of "sporadically," suffice it to say that the 9th was going to adhere to precedent on this issue, because the constitutionality of 2254(d) is something that is an EXTRAORDINARILY high profile issue and there is a ninth circuit case precisely on point. there is no wiggle room to distinguish the issue as presented in '06 from when it was considered in '03.
3.6.2007 4:22pm
James Grimmelmann (mail) (www):
The opinions in this case raises all sorts of jurisprudential and metaphysical points about the nature of a holding and the nature of precedent. There are some mighty fine hairs being split here as to when judges are "bound" and when their statements "bind" other judges.
3.6.2007 4:25pm
Justin (mail):
[Deleted by OK. Justin, if you have a personal problem with me or my posts, please e-mail them to me directly rather than occupy the blog with them.]
3.6.2007 4:39pm
Justin (mail):
[Deleted by OK. Justin, seriously, I meant it: If you want to discuss things with me one-on-one, just send me an e-mail. I cannot reach you one-on-one because you are anonymous; I have no idea who you are. Please e-mail me if you want to comment on my blogging rather than use a thread for that.]
3.6.2007 4:55pm
Kent Scheidegger (mail) (www):
For a minute there, Kovarsky, I thought you had put Peggy Noonan on the Ninth Circuit.

For anyone interested in the long version of this debate, see Liebman's article at 98 Colum. L. Rev. 696 and the response by yours truly at 888 of the same issue.
3.6.2007 5:47pm
rothmatisseko (mail) (www):
Perhaps the two concurred, while deciding the case based on earlier circuit precedent, as a polite way of lobbying for en banc review.
3.7.2007 2:24am
Mark P. (mail):
Leaving the judicial politicking out of it, the concurrences in this decision raise a very interesting point. What power does Congress have to establish legal standards and procedures for deciding cases, particularly in the area of habeas corpus? Some of the AEDPA is procedural, and some of it is substantive. Are the substantive standards infringing upon the Judicial Power? Similarly, how about the Harmless Error Rule and the Plain Error Rule in the Federal Rules of Criminal Procedure. Are those substantive? If the concurrence is correct, do these rules similarly infringe upon the Judicial Power? My gut says no, but I'm curious what others think.
3.7.2007 10:51am