There is a very surprising feature of the travel ban litigation on the west coast of the United States that seems to have escaped the attention of the mainstream British media.
In an earlier blog I set out the reasons why a three-judge panel of the US Ninth Circuit Court of Appeals refused to disturb a ruling made by a single judge in the State of Washington which had frozen the effect of President Trump’s first executive order. Since then, the President has issued a new executive order, replacing the first, and a federal judge in Hawaii has made a similar order in relation to this new ban.
Once the President had decided on this change of tactics, his lawyers requested the Court of Appeals to dismiss their appeal, a request which was granted without opposition on 8 March 2017. They did not seek a ruling from a larger “en banc” panel of the court.
This, one would have thought, was the end of those proceedings.
Not a bit of it.
Under the court’s procedural rules, not only may the losing party request a rehearing before an “en banc court” (consisting of the chief judge and ten non-recused members of the court, selected at random) but any judge of the court may also make the same request, without any need to disclose his/her identity.
This is what happened in this case. There were at the time 25 members of the court, so that 13 of them would have to agree before an en banc hearing could take place.
On 15 March 2017 the court published an order to the effect that such a request had been received, but that the court had rejected it. However, five members of the court had dissented, and one of them, Judge Bybee, filed a lengthy judgment giving his reasons for the dissent. There was also filed a very short concurring judgment, filed by Judge Reinhardt, one of the majority of the members of the court.
Two days later an amending order was filed. On this occasion there were two concurring judgments (the first a slightly longer version of the original) and two further dissenting judgments.
They can all be found on the court’s excellent website, but for present purposes I am content to quote only two of them: the concurring judgment of Judge Reinhardt (before and after it was amended) and the new dissenting judgment of Judge Kozinski, the former chief judge of the court (with which the other four dissenting judges concurred).
While retaining the text in full, I have reformatted the judgments, so that they will be easier for non-US readers (and particularly non-lawyers) to follow. The dissenting judgment (including some of the footnotes) is written in a swashbuckling style that seems to be more at home in California than it would be in Parliament Square.
It would be unthinkable for anything like this to happen in an English court, especially when the court was likely to receive a fresh appeal relating to the new order. The majority of the court took the view that the issue was now academic. The appeal was withdrawn, and there was nothing for the court to decide. And they would not wish to be seen to prejudge any fresh appeal relating to similar subject-matter.
The minority thought differently.
Here is Judge Reinhardt’s concurring judgment:
“I concur in our court’s decision regarding President Trump’s first Executive Order – the ban on immigrants and visitors from seven Muslim countries. I also concur in our court’s determination to stand by that decision, despite the effort of a small number of our members to overturn or vacate it. Finally, I am proud to be a part of this court and a judicial system that is independent and courageous, and that vigorously protects the constitutional rights of all, regardless of the source of any efforts to weaken or diminish them.”
“Judge Kozinski’s diatribe, filed today, confirms that a small group of judges, having failed in their effort to undo this court’s decision with respect to President Trump’s first Executive Order, now seek on their own, under the guise of a dissent from the denial of en banc rehearing of an order of voluntary dismissal, to decide the constitutionality of a second Executive Order that is not before this court. That is hardly the way the judiciary functions. Peculiar indeed!”
And here is Judge Kozinski’s dissent:
“I write separately to highlight two peculiar features of the panel’s opinion. First, the panel’s reasoning rests solely on Due Process. But the vast majority of foreigners covered by the executive order have no Due Process rights. Nevertheless, the district court enjoined the order’s travel provisions in their entirety, even as applied to the millions of aliens who have no constitutional rights whatsoever because they have never set foot on American soil.
In short, the panel approves the district court’s nationwide injunction using a rationale that applies to a small percentage of those covered by the President’s order. The panel itself seems to acknowledge this strange state of affairs when it notes that there “might be persons covered” by the district court’s restraining order who have no Due Process claims. “Might” indeed! The overwhelming majority of the hundreds of millions of people covered by the order lack Due Process claims; only a tiny proportion have been accorded lawful status. Yet the panel offers no explanation for allowing the district court’s extraordinarily broad restraining order to stand in full. This St. Bernard is being wagged by a flea on its tail.
Because we have an obligation to maintain as much of the order as is legal, we normally ask: Can we keep it operational in a way that avoids constitutional conflict? The law of our circuit is that we consider the severability of an executive order just as we would consider the severability of a statute. If we applied this framework to the executive order, we would “refrain from invalidating more of the [order] than is necessary” and “maintain the [order] in so far as it is valid. This would have been easy: We could have approved the injunction as to the relatively few who have lawful status in the United States and allowed the executive order to cover everyone else. This workable solution would have respected the President’s prerogative to regulate immigration as delegated to him by 8 U.S.C. § 1182(f), a provision about which the panel says nothing.
Which brings me to the second peculiar feature of the opinion, a topic about which the panel says all too much: the Establishment Clause. While its opinion does not come to rest on this issue, the panel still sows chaos by holding “that evidence of purpose beyond the face of the challenged law may be considered in evaluating Establishment and Equal Protection Clause claims.” This matters because one Establishment Clause test requires a showing of secular purpose, and the panel gives its imprimatur to considering the “numerous statements by the President” about Muslims, most of them made before he was elected or took office.
This holding has continued vitality: It was relied on only days ago by a district judge in Hawaii who, in the ongoing contretemps between our circuit and the executive, enjoined the President’s new executive order nationwide. Indeed, this holding is spreading like kudzu through the federal courts. Taking a cue from the panel’s opinion and citing a trove of informal and unofficial statements from the President and his advisers, the district judge found that plaintiffs had shown “a strong likelihood of succeeding on their claim” that the new order violates the Establishment Clause.
And why shouldn’t he? After all, the panel made this evidentiary snark hunt the law of the Ninth Circuit; the district judge was (in his own word) “commanded” to follow it. This is folly. Candidates say many things on the campaign trail; they are often contradictory or inflammatory. No shortage of dark purpose can be found by sifting through the daily promises of a drowning candidate, when in truth the poor shlub’s only intention is to get elected. No Supreme Court case—indeed no case anywhere that I am aware of—sweeps so widely in probing politicians for unconstitutional motives.
And why stop with the campaign? Personal histories, public and private, can become a scavenger hunt for statements that a clever lawyer can characterize as proof of a -phobia or an -ism, with the prefix depending on the constitutional challenge of the day. This path is strewn with danger. It will chill campaign speech, despite the fact that our most basic free speech principles have their “fullest and most urgent application precisely to the conduct of campaigns for political office.”
And it will mire us in a swamp of unworkable litigation. Eager research assistants can discover much in the archives, and those findings will be dumped on us with no sense of how to weigh them. Does a Meet the Press interview cancel out an appearance on Face the Nation? Does a year-old presidential proclamation equal three recent statements from the cabinet? What is the appropriate place of an overzealous senior thesis or a poorly selected yearbook quote? Weighing these imponderables is precisely the kind of “judicial psychoanalysis” that the Supreme Court has told us to avoid.
The hopelessness of this weighing exercise is why the Supreme Court has never “deferred to comments made by [government] officials to the media.” And it’s why the panel’s case citations for the supposedly “well established” proposition that the President’s informal statements are admissible, upon closer inspection, turn out to refer to a much more limited universe: the text of city council resolutions, early drafts of legislation, transcripts of legislative discussions and contemporaneous statements by legislative members.
Limiting the evidentiary universe to activities undertaken while crafting an official policy makes for a manageable, sensible inquiry. But the panel has approved open season on anything a politician or his staff may have said, so long as a lawyer can argue with a straight face that it signals an unsavory motive. Even if a politician’s past statements were utterly clear and consistent, using them to yield a specific constitutional violation would suggest an absurd result—namely, that the policies of an elected official can be forever held hostage by the unguarded declarations of a candidate.
If a court were to find that campaign skeletons prevented an official from pursuing otherwise constitutional policies, what could he do to cure the defect? Could he stand up and recant it all (“just kidding!”) and try again? Or would we also need a court to police the sincerity of that mea culpa—piercing into the public official’s “heart of hearts” to divine whether he really changed his mind, just as the Supreme Court has warned us not to?
This is yet another reason my colleagues err by failing to vacate this hasty opinion. The panel’s unnecessary statements on this subject will shape litigation near and far. We’ll quest aimlessly for true intentions across a sea of insults and hyperbole. It will be (as it were) a huge, total disaster.”
As I have said, it would be unthinkable for anything like this to happen in an English court. The growing politicization of judicial process in the United States (where the majority party in the Senate first refuses even to consider the merits of an Obama nominee for the vacancy in the US Supreme Court, and now seems poised to change its rules in order to permit a Trump nominee to be approved by a simple majority which does not pass the requisite 60% threshold) should give pause to those (like the Editor of the Daily Mail and Mr Iain Duncan-Smith MP) who seem keen on a similar politicization of the judiciary over here.
And the language of Judge Kozinski’s dissent may be a harbinger of what might happen in this continuing litigious struggle if and when it reaches a fully constituted nine-judge US Supreme Court.
 See Zadvydas v. Davis, 533 U.S. 678, 693 (2001); United States v. Verdugo-Urquidez, 494 U.S. 259, 269 (1990).
 Panel Order at 23.
 See Matter of Reyes, 910 F.2d 611, 613 (9th Cir. 1990); see also Minnesota v. Mille Lacs Band of Chippewa Indians, 526 U.S. 172, 191 (1999) (assuming without deciding that the same severability analysis applies to executive orders as to statutes). Indeed, we know that this executive order can be severed because the district court did precisely that: It enjoined the five subsections of the executive order relating to travel and left the other eleven intact. Washington v. Trump, No. C17-0141JLR, 2017 WL 462040, at *2 (W.D. Wash. Feb. 3, 2017) (order granting temporary restraining order).
 Regan v. Time, Inc., 468 U.S. 641, 652 (1984).
 Panel Order at 25.
 I don’t endorse Lemon v. Kurtzman, 403 U.S. 602 (1971), as the appropriate test in this context. Like Judge Bybee, I am puzzled why Lemon should be plucked from domestic contexts and applied to laws affecting immigration. See Bybee Dissental at 8 n.6. If we apply this test so casually to immigration policy, I see no reason it should not apply to every foreign policy decision made by the political branches, including our dealings with various theocracies across the globe. I see many reasons to resist this gross intrusion of the judicial power into foreign affairs.
 See Hawaii v. Trump, No. 17-00050 DKW-KSC (D. Haw. Mar. 15, 2017) (order granting temporary restraining order).
 See Int’l Refugee Assistance 2 page 4 Project v. Trump, No. 17-00361-TDC, at 5, 29 (D. Md. Mar. 16, 2017).
 See Hawaii at 33–37.
 Id. at 41.
 Id. at 32
 There is an anecdote, doubtless apocryphal, about Franklin Roosevelt during a whistlestop tour. He had two speeches that took opposite positions on a hot-button issue of the day. When the train stopped at a town that favored the issue, he would give his “pro” speech. And in towns that opposed the issue he’d give his “con” speech. One day he approached a town that his advisors told him was divided evenly between the pros and cons. FDR’s advisers worried about how he’d handle the situation, but FDR was undaunted. He gave a speech and when he was done the pros in the audience believed he was in their corner and the cons were convinced he agreed with them. And that, friends, is the nature of electoral politics.
 Respect for a coordinate branch should also counsel against focusing on campaign statements. Candidate Trump, unlike President Trump, had not taken an oath to “preserve, protect and defend the Constitution,” U.S. Const. art. II, § 1, cl. 8, and was not bound to “take Care that the Laws be faithfully executed,” id. art. II, § 3.
 McCutcheon v. Fed. Election Comm’n, 134 S. Ct. 1434, 1441 (2014) (citation and internal quotation marks omitted).
 McCreary County v. ACLU of Ky., 545 U.S. 844, 862 (2005).
 Hamdan v. Rumsfeld, 548 U.S. 557, 623–24 n.52 (2006).
 See Church of Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 534–35 (1993); Larson v. Valente, 456 U.S. 228, 254 (1982); Vill. of Arlington Heights v. Metro Housing Dev. Corp., 429 U.S. 252, 268 (1977).
 See McCreary, 545 U.S. at 862.
 Contrary to the claims of Judges Reinhardt and Berzon, the substance of the panel’s opinion continues to be highly relevant. Because the panel has refused to vacate it, the opinion continues to be the law of the circuit and is being followed by courts in the circuit and elsewhere. My criticism bears directly on the mistake our court has made in failing to vacate the opinion, and will hopefully warn other courts away from similar errors. My colleagues’ effort to muzzle criticism of an egregiously wrong panel opinion betrays their insecurity about the opinion’s legal analysis.