The Establishment Clause took center stage in Monday’s 10-3 Fourth Circuit decision in International Refugee Assistance Project (IRAP) v. Trump upholding a Maryland district court’s preliminary injunction against President Trump’s revised Refugee EO. In holding that the revised EO violated the Establishment Clause, Chief Judge Gregory’s majority opinion was not satisfied with applying one legal test. The court needed two tests to explain its reasoning. However, while the court’s explanation of its holding was complicated, the driving force was obvious: cherry-picking the campaign, post-election, and post-inaugural utterances of Donald Trump to make out a case for anti-Muslim animus. Unfortunately, that approach will not produce clear guidance on the contours of the Establishment Clause in foreign policy.
Application of the Establishment Clause to national security and foreign policy is a risky undertaking, because of the multitude of variables that drive government action in these dynamic domains. (See Josh Blackman’s post here.) The Establishment Clause prohibits government actions intended to and likely to produce injury or aid to religion. Sorting out religious from neutral factors is a difficult task. Consider a salient event in post-World War II U.S. foreign policy: President Truman’s recognition of Israel. Truman expressly recognized the creation of a “Jewish state” within Israel’s borders. Clark Clifford, Truman’s close advisor and future Secretary of Defense under LBJ, recalled that Truman was fond of quoting Deuteronomy 1:8, which quotes God as commanding the Israelites to “go in and possess the land [of] … Abraham, Isaac, and Jacob.” Since one of Truman’s demonstrable motives was a desire to help Judaism survive and prosper, was Truman’s decision infected with an improper religious purpose?
Even (or perhaps especially) if one grants that President Truman’s handle on the Bible was surer than President Trump’s, this question should answer itself. A complex matter like recognition of a state entails a range of motivations. Isolating the religious motivation and invalidating the action taken would be a sure path to paralysis in foreign policy. Madison, who in Federalist No. 41 counseled vigilance regarding the hostile “exertions” of foreign powers, would be aghast. (This is just an illustration; the President may have plenary Article II power over recognition decisions, although the Establishment Clause might restrict this authority, just as the Fourth Circuit found in IRAP that this clause restricted the political branches’ authority over noncitizens with no ties to the United States.)
The Supreme Court took a sensible approach to these fraught issues in Salazar v. Buono (2010). The Court found that a blatantly religious symbol—a cross—was almost certainly intended by Congress not as a promotion of religion but as a commemoration of Americans killed in World War I. The Court viewed the complex matters of war and peace addressed by the political branches of the federal government as eons removed from the state and local religious displays that dominate the Establishment Clause docket.
The Fourth Circuit’s approach in IRAP v. Trump rejects this common-sense approach. Instead, the court aggregates two tests. One test is the “facially legitimate and bona fide” immigration test established by the Supreme Court in Kleindienst v. Mandel (1972) and recently applied by Justice Kennedy in his concurrence in 2015’s Kerry v. Din (2015). The other test is from the Establishment Clause chestnut, Lemon v. Kurtzman (1971). Under that approach, for the state to prevail, a court must find a “secular purpose,” a “primary effect [of the challenged action] that neither advances nor inhibits religion” and the absence of a tendency to “foster excessive state entanglement with religion.” The Fourth Circuit’s approach to the revised EO becomes slippery in the awkward transition from Din to Lemon.
The Fourth Circuit majority construes the Mandel/Din test as requiring proof that a challenged government action is both, 1) “facially legitimate,” and, 2) “bona fide” (i.e., as reflecting good faith throughout, not merely on the face of a document or action). The good faith vel non of a government action is clearly crucial. If a plaintiff makes what Justice Kennedy in Din called an “affirmative showing of bad faith,” the court can conduct a “more searching judicial review.” For the Fourth Circuit majority, that more searching review is teed up when the government’s good faith is “seriously called into question.”
Asking whether the government’s good faith has been “seriously called into question” sounds like a high threshold for more intrusive judicial review—until it isn’t. Consider Congress’s exaltation of crosses in Salazar v. Buono (2010), or President Truman’s fervent invocation of Deuteronomy as he recognized Israel. Surely, each prompts “serious questions” about the role of official intent to aid religion. And that’s the rub. Cherry-pick Congress’s favorable comments about crosses or Truman’s citations to Deuteronomy and—presto!—“serious questions” proliferate. At that point, any decision by the political branches on matters of foreign policy is subject to the vagaries of the Lemon test, which the Supreme Court often ignores, even in more routine Establishment Clause cases. As Fourth Circuit judges Niemeyer, Shedd, and Agee noted in dissent, the majority’s approach is “unworkable and inappropriate.” Justice Jackson, who championed a “workable government” in his canonical Youngstown concurrence, would share Madison’s puzzlement at the peculiar turn the Fourth Circuit has taken.
However, the Fourth Circuit majority is right and the dissenters are wrong about one key issue: purpose should matter. Indeed, Jeffrey Wall, who as Acting Solicitor General argued the government’s case, appeared to concede at one point that a sliding scale might best regulate the relationship between a measure’s putative neutral objectives and the statements that illuminate the motives behind the measure. A palpably impermissible purpose, measured by the intensity, uniformity, and duration of a politician’s statements (even campaign utterances), should trigger more searching review. Suppose that President Truman, the second Baptist president, had taken to convening a daily Orthodox minyan in the Oval Office, pressing prayer shawls and yarmulkes on Clark Clifford, Secretary of State George C. Marshall, and the President’s many other advisors. General Marshall may not have been amused (he opposed recognition of Israel in the first place) and Truman’s hypothetical display of Jewish devotion may have triggered more robust investigation of recognition’s secular aims (having an ally in the Middle East, etc.).
Unfortunately, the Fourth Circuit’s “serious question” test, as applied by Judge Gregory, finds an impermissible purpose based on far more ambiguous evidence. Take the explanation of the first refugee EO by Trump advisor Rudy Giuliani. According to Giuliani, Trump tasked him to “[p]ut a commission together … [and] show me the right way to do it legally” (emphasis added). Judge Gregory, at page 51 in his opinion, describes Giuliani as recalling that President Trump wanted “to find a way to ban Muslims in a legal way,” proving that Trump’s request was merely a legal dodge to achieve his desired result of a Muslim ban. Judge Gregory’s paraphrase of Giuliani’s account assumes the very fact in contention: what “it” means.
According to Judge Gregory, the answer is obvious. Indeed, the meaning of “it” is so clear that Judge Gregory simply excises that inoffensive pronoun entirely and conveniently substitutes “ban Muslims.” Substituting assumptions for answers is the hallmark of what cognitive psychologists call “confirmation bias”: viewing new facts as fitting into the observer’s preexisting narrative. Most human beings are prone to this error, including President Trump. Judges aren’t immune, either, but they should acknowledge this frailty, not succumb to it. Instead, the Fourth Circuit’s “serious question” test invites judges to double down on their own predilections.
Judge Gregory’s leap to the conclusion that “it” means a “Muslim ban” is surely one inference, but it’s not the only one. The “it” in Giuliani’s explanation could also refer to pausing certain immigration temporarily, in order to assess whether the United States’ current criteria for screening immigrants actually work. Pending an answer to this key question, it’s not inherently invidious to temporarily suspend immigration from countries riven by terrorism and armed conflict (like five countries on the EO list: Libya, Somalia, Sudan, Syria, and Yemen) or a state that sponsors terrorism (Iran, the sixth country on the list). The revised EO’s method is unnecessary and unwise, in my view. However, courts have never viewed a foreign policy measure’s lack of wisdom as dispositive evidence of its illegality. For the sake of both the political branches and the judiciary, the statements of President Trump shouldn’t tempt judges to go down this road.
Confirmation bias is also at work in the Fourth Circuit’s citation to President Trump’s description of the revised EO as a “watered-down version” of the original order. As I noted in an earlier post, the Supreme Court has consistently interpreted the phrase “watered-down” as connoting a material change. The Fourth Circuit majority could have followed the Supreme Court’s guidance. That might have led them to acknowledge that the revised EO is substantially different from the original order and “tailored” (to use President Trump’s term) to the Ninth Circuit’s decision on the original EO. After all, the revised EO exempts lawful permanent residents and current visa-holders, who have the strongest claims among noncitizens for protection from broader restrictions on entry into the United States. Unfortunately, the Fourth Circuit opted to substitute confirmation bias for reasoned inquiry. Viewed through the prism of a narrative that assumes invidious bias, the phrase “watered down” no longer means “material alteration.” Instead, it magically assumes the attributes of a “superficial tweak,” reinforcing the Fourth Circuit’s holding that the revised EO inherited the flaws of the original.
Viewed in this light, the Fourth Circuit’s cherry-picking of Trump’s utterances is not salutary pushback against his administration’s excesses, but yet another example of collateral damage. George Orwell noted more than seventy years ago, in his classic essay Politics and the English Language, that politicians view language as infinitely malleable to suit their own agendas. President Trump lacks clean hands in this pursuit. But judges who reciprocate do not improve the situation.
The difference between the aims of the revised EO asserted by the Trump administration and the assumptions made by the Fourth Circuit highlight a flaw with a statutory argument made in a concurrence by Judge Wynn. This argument (also made by Neal Katyal for Hawaii in the Ninth Circuit case argued last week and by veteran Supreme Court advocate Andrew Pincus in an amicus brief for technology firms) goes as follows: the Immigration and Nationality Act (INA) includes a long list of grounds for inadmissibility based on terrorism, enumerated at 8 U.S.C. 1182(a)(3)(B). Judge Wynn asserts in his concurrence that the logic of these enumerated grounds precludes presidential action beyond the ordinary realm of visa processing.
This argument from statutory structure is misconceived, because it fails to acknowledge that the revised EO does not seek to supplant the grounds for lawful status and inadmissibility set out in the INA. Rather, the revised EO only seeks to ensure that the U.S. is collecting and analyzing data correctly to determine whether noncitizens abroad meet statutory criteria. Viewed in this light, the revised EO is an interstitial, gap-filling measure, not a substantive change in the statute. A longer pause, or a series of consecutive pauses, would raise red flags about material changes in the INA. Only Congress can make such changes. However, the temporary pause outlined in the revised EO does not rise to that level.
In sum, the Fourth Circuit’s decision doubles down on the dubious practice of cherry-picking Donald Trump’s campaign utterances and subsequent statements. My hope is that the Ninth Circuit panel that heard a related appeal last week (see my post here) will offer an alternative approach that allows some room to assess purpose, but also safeguards against the blandishments of confirmation bias. In any case, the Supreme Court is waiting in the wings.