Editor’s Note: This post was originally published on December 10, 2015 at 12:50am.

On Tuesday, I wrote a short piece for CNN.com explaining why Donald Trump’s proposed ban on all Muslim immigration, separate from its moral depravity, would almost certainly be struck down by the courts if, somehow, it ever came to pass. Then came blog posts from, among others, Eric Posner, Eugene Volokh, and Pratheepan Gulasekaram, and an op-ed from Peter Spiro, all of which argue that Trump’s proposal, while heinous, would not actually be unconstitutional. Although the arguments differ in some key respects, they have at their core one or both of two basic constitutional conclusions: (1) that the “plenary power” doctrine allows the government to do things in the context of immigration regulation that it couldn’t do otherwise; and (2) that, insofar as the “victims” of such a proposal would be non-citizens outside the United States, they wouldn’t have constitutional rights that such a ban could violate. In the post that follows, I want to amplify a point that my CNN piece made implicitly: Although both of these statements have some merit, they don’t actually support the conclusion that a categorical ban on Muslim immigration would be upheld by courts today. Simply put, the argument that such a ban would be constitutional neglects both (1) the decline of the plenary power doctrine; and (2) the very different issues that the ban would raise under the Free Exercise and Establishment Clauses — provisions that have never been enforced with lesser scrutiny at the border.

I.  The Posner/Spiro Argument

Here’s Eric Posner, writing Tuesday at his eponymous blog:

[A]ny honest answer to a journalist’s question about whether Trump’s plan to ban Muslim immigration is unconstitutional should start with the plenary powers doctrine, and observe that it would be an uphill battle to persuade the Supreme Court to abandon a century of precedent.

And here’s Peter Spiro in yesterday’s Times:

The court has given the political branches the judicial equivalent of a blank check to regulate immigration as they see fit. This posture of extreme deference is known as the “plenary power” doctrine. It dates back to the 1889 decision in the Chinese Exclusion case, in which the court upheld the exclusion of Chinese laborers based on their nationality.

Unlike other bygone constitutional curiosities that offend our contemporary sensibilities, the Chinese Exclusion case has never been overturned. More recent decisions have upheld discrimination against immigrants based on gender and illegitimacy that would never have survived equal protection scrutiny in the domestic context. Likewise, courts have rejected the assertion of First Amendment free speech protections by noncitizens.

In other words, courts have historically given Congress wide latitude when it comes to immigration law, and that latitude is enhanced by the minimal extent to which the Constitution applies to non-citizens without substantial voluntary connections to the United States. Applying these two principles to a ban on all Muslim immigration, then, they conclude that the ban may well be upheld.

II.  The Decline of the Plenary Power Doctrine

It is certainly true that the Supreme Court for nearly a century embraced the “plenary power” doctrine, and the idea that the Constitution had little (if anything) to say about how the government decides which non-citizens will be allowed to enter the United States. But before turning to the unique challenges posed by the religion clauses, it’s worth stressing a point largely (albeit not completely) neglected by the authors of these posts — that the “plenary power” doctrine is on the decline, and has been for some time.

Much of the plenary power doctrine dates to an era in which virtually all federal legislation was only reviewed under the deferential “rational basis” standard. The Supreme Court did not apply heightened scrutiny to any alienage-based discrimination until 1971 — two decades after the decisions typically cited as the high-water mark of plenary power. It’s true that, even in the years after 1971, the Court gave Congress more deference in the immigration context than it would receive elsewhere — including in cases like Kleindienst v. Mandel (which Posner cites), and Matthews v. Diaz (which I cited in my CNN piece). But all of those cases are older than I am, and there really haven’t been similarly deferential decisions in the four decades since.

Instead, more recently, the Court has repeatedly applied ordinary constitutional principles to a wide array of immigration matters, even in the face of the plenary power doctrine. As Justice Breyer wrote for the majority in Zadvydas v. Davis, “[Congress’s plenary] power is subject to important constitutional limitations. In these cases, we focus upon those limitations.” To that end, Zadvydas imposed due process constraints on the government’s power to detain non-citizens who had already been ordered removed indefinitely pending their removal.

This is not to say that the plenary power doctrine is dead, or that decisions like the Chinese Exclusion Case are no longer good law (whether or not they ought to be); rather, it’s to suggest that contemporary courts will be unlikely to embrace the plenary power doctrine with the same gusto as their predecessors — especially when the question is how it applies in a new context, as opposed to one (like national origin discrimination) in which there is established (if controversial) precedent.

III.  The Unique Role of the Religion Clauses

My skepticism about the plenary power doctrine is only buttressed by the very different constitutional objections that a ban on Muslim immigration would raise as compared to even the more pernicious immigration regulations courts have previously upheld. After all, although national origin discrimination is (today, at least) subject to the same heightened scrutiny as religious discrimination for equal protection purposes, religious discrimination also implicates the Free Exercise and Establishment Clauses — two provisions that have seldom, if ever, shown up in the Supreme Court’s plenary power jurisprudence.

For obvious (and similar) reasons, a ban on Muslim immigration implicates both provisions — it discriminates expressly against all practitioners (or, at least, all non-citizen practitioners) of a particular religious belief. As Justice Black wrote for the Supreme Court in 1947, “The ‘establishment of religion’ clause of the First Amendment means at least this: … No person can be punished for entertaining or professing religious beliefs or disbeliefs … .” Indeed, it is difficult to imagine a more patent violation of both the Free Exercise and Establishment Clauses than an Act of Congress that singles out Muslims as such — a violation of free exercise insofar as it inhibits those Muslims’ religious freedom in a manner that is grossly overbroad (and underinclusive); and an impermissible establishment of religion insofar as Congress would be disfavoring a specific religious affiliation (and thereby favoring all others). And, as Professor Ira Lupu has pointed out, administering such a ban would raise its own grave Establishment Clause questions, since it would require the relevant government officers to ask (and answer) ecclesiastical questions — exactly what the Establishment Clause prevents the government from doing. We don’t look for Shibboleths.

Tellingly, the defenses (such as they are) of the constitutionality of a ban on Muslim immigration don’t grapple head-on with how courts would apply the plenary power doctrine in the face of the religion clauses (other than simply asserting that plenary power would prevail). Instead, they punt by arguing that non-citizens outside the United States aren’t protected by the religion clauses insofar as they lack constitutional protections more generally.

As I wrote in my CNN piece, even assuming that the ban could be applied in a way that didn’t sweep in non-citizens with substantial connection to the United States (a rather difficult assumption, methinks), these analyses fail to grapple with the very real possibility that, like the plenary power doctrine, categorical understandings of the Constitution’s extraterritorial applicability are also on the decline. But even if they’re not, defenses of the ban’s constitutionality also fail to grapple with an argument unique to the Establishment Clause — that it is, in fact, a structural limit on the federal government’s power (just look at its distinct wording), and, as such, constrains the federal government no matter where (or against whom) it is acting. This is the exact view of the Establishment Clause that Justice Thomas has articulated — first in the pledge of allegiance case in 2004, and then again last year in the Town of Greece case. To be sure, Justice Thomas’s argument was offered as a reason for why the Establishment Clause shouldn’t have been incorporated against the states (on the theory that the purpose of the Clause was to keep the federal government from interfering with state establishments), but the structural view of the clause ought to run in both directions. If he’s right, and means what he says, then the Establishment Clause would apply even to an Act of Congress that only affected non-citizens lacking substantial voluntary connections to the United States (and would be violated by a statute that categorically singled out one religious group for discriminatory treatment).

More to the point, a view of the Establishment Clause as a structural limit on the government would also only further undermine any argument for relying upon the plenary power doctrine; no Supreme Court case has ever suggested that Congress can violate structural constitutional principles simply because it’s an immigration case, and plenty of decisions (e.g., INS v. Chadha) stand rather decisively to the contrary.

I don’t mean to oversell the structural view of the Establishment Clause — all the more so since it only appears to have one vote among the current Justices. Rather, it just underscores my broader view, which is that there are surely at least five Justices (and probably more) who would support the proposition that the Establishment Clause limits Congress’s power even as applied non-citizens seeking to enter the United States — whether because it is structural (the Thomas view), because the “entry fiction” is an historical relic, or because it is not rigidly territorial in its application. Whatever the theory, the result would be the same — that a ban on immigration by adherents to a specific religious faith would be invalidated.

*          *          *

Professor Posner concluded his post with the following lamentation:

any honest answer to a journalist’s question about whether Trump’s plan to ban Muslim immigration is unconstitutional should start with the plenary powers doctrine, and observe that it would be an uphill battle to persuade the Supreme Court to abandon a century of precedent. Unfortunately, that is not what scholars–who certainly know better–are telling journalists. They are likely being abetted by journalists and headline writers who don’t like the idea that Trump’s ban would be lawful. Not everything that is stupid or offensive is unconstitutional.

I certainly agree with the last sentence. But it doesn’t prove the point. Indeed, some things that are stupid or offensive are unconstitutional, and this strikes me as one such example. It would be a much closer question, methinks, if Mr. Trump had proposed a ban on Syrian immigration — which would raise the continuing scope of the plenary power doctrine (and of the Supreme Court’s decision in the Chinese Exclusion Case) without the baggage of the religion clauses. But regardless of why a Muslim immigration ban would be struck down, the fact that it would, to me, seems clear, and not just because I’ve been “abetted by journalists and headline writers who don’t like the idea that Trump’s ban would be lawful.”