Most of what Judge VanDyke (who authors the opinion) and Judge Miller (who authors the concurrence) seems right to me. The Ninth Circuit previously relied on California state precedent to say that first-party attorney's fees are recoverable as damages when third-party attorney's fees are recoverable. But thereafter came a ton of California appellate cases that held to the contrary, saying that prior California (and federal) precedent was wrong.
Normally, a Ninth Circuit panel is bound by circuit precedent unless the intervening authorities from higher courts make it clear that circuit precedent is wrong. But the standard's different when it's state law, since each panel is "predicting" what the highest tribunal in that state would do. There's more flexibility in that context. So, here, when it's (now) pretty clear that California courts would hold X, a Ninth Circuit panel applying state law can hold X even though prior circuit precedent has held Y.
So far, so good. I agree.
But it seems to me that
the panel's opinion, while correct in the
current case, is perhaps a bit too broad as a general matter.
Here, the intervening change in California precedent was fairly clear. California precedent was a bit split even at the time of the initial panel's decision, but thereafter, the Court of Appeal's holdings seem almost uniformly tilted the other way. So I agree that California law has changed.
But both Judge Van Dyke and Judge Miller take quite the "permissive" (to use Judge Miller's words) view of the ability of one Ninth Circuit panel to depart from another circuit panel when issues of state law are in play. To quote Judge Miller (with emphases in original, but citations omitted):
"On the other hand, we have stated, albeit without
explanation, that circuit precedent interpreting state law “is
only binding in the absence of any subsequent indication
from the [state] courts that our interpretation was incorrect.' That statement suggests a
more permissive view—that a decision of an intermediate
state appellate court could be a sufficient basis on which to
depart from circuit precedent on a state-law question. . . . Today the court correctly adopts the more permissive
view."
Isn't that way too broad?
Imagine, for example, that there are 20 Court of Appeal decisions that squarely hold X, and that none go the other way. On that basis, a Ninth Circuit panel holds that state law is X, and applies and follows it in a given case. In the next two decades, there are 50 Court of Appeal decisions that similarly hold X, all in reasoned opinions, but since there's no horizontal precedent in California, there's also a single outlier opinion by a crazy panel that holds Y. No other Court of Appeal opinion holds Y, and the other 50 Court of Appeal opinions critique the one outlier and say it's wacko.
According to AGK Sierra, a new Ninth Circuit panel isn't bound by the earlier circuit precedent, and can hold that state law is in fact Y. Because there is indeed some intervening precedent that says Y. (Remember: There just needs to be "any" indication from "any" state court that Y is true, and even a single opinion can suffice.)
That seems to me to grant insufficient consideration to the value of horizontal precedent in the federal system. Particularly in the present politically polarized era. A conservative panel that likes Y (or go the other way -- a liberal panel that likes Y) might well be inclined to go with Y in such a setting despite the existence of circuit precedent and the weight of authority in the state. Then you'd need an en banc court to fix the problem: the very thing that Judges Van Dyke and Miller (rightly) want to avoid.
I would think that the better standard would say something like: "Circuit precedent on state law is not binding when there is substantial intervening authority in the state court suggesting that state law on the issue has changed." That's not as tough as the standard for intervening federal authority, which requires "clearly irreconcilable" precedent, but still tougher than the very "permissive" state law standard set forth in this opinion.
A somewhat minor change, perhaps. But I think a valuable one regardless.
Because circuit precedent matters. A fair piece.
Even in state law cases. Predictability, safeguard against political manipulation, etc. All that jazz.