The Long-Term Stakes in the Remain in Mexico Case

 by Michael C. Dorf

On Friday, the Supreme Court granted certiorari in Biden v. Texas, setting an expedited briefing schedule that will result in an oral argument this April. A federal district court forbade the Biden administration from terminating the Trump administration's "Migrant Protection Protocols" (commonly called "Remain in Mexico" but which I'll abbreviate as MPP) under which undocumented immigrants presenting themselves at the southern border for admission as refugees or otherwise are temporarily removed to Mexico to await a hearing to determine their eligibility to enter the United States. The US Court of Appeals for the Fifth Circuit denied relief from the district court order.

The case presents two basic questions: (1) Whether MPP is required by statute?; and (2) whether the Fifth Circuit rightly rejected detailed memoranda memorializing the work done by the Biden administration's Department of Homeland Security to strengthen the case for MPP rescission in response to the initial district court determination that rescission was arbitrary and capricious? Here I'll consider each of them with an eye towards showing how--if one sets aside the particular context and supposes, perhaps naively, that SCOTUS would apply the same rules and standards to Republican and Democratic administrations alike--the ideological stakes are not entirely clear.

(1) The Biden administration's fundamental argument is quite straightforward: the key statutory provision in question, 8 U.S.C. § 1225(b)(2)(C) states that the government "may return" an undocumented immigrant pending their hearing, and "may" is permissive language. Thus, while the Trump administration was permitted to adopt MPP, likewise the Biden administration was permitted to rescind it.

Two main arguments have been offered by the plaintiffs and the Fifth Circuit in response. First, they argue that the foregoing statutory provision works in tandem with another provision of the same statute, 8 U.S.C. § 1225(b)(2)(A), which states that an undocumented immigrant "shall be detained" pending their hearing. According to the plaintiffs and the Fifth Circuit, the choice afforded by (C) is really only a choice if the alternative is detention in the United States. However, the government lacks the capacity to detain everyone who enters, so the only way it can satisfy its "shall be detained" obligation under (A) is to utilize what the Fifth Circuit termed the safety valve of (C). According to the plaintiffs, then, yes, the government "may" choose not to send undocumented immigrants to wait in Mexico, but only if it detains them here; and because it isn't detaining all of them, it must stick with MPP.

The administration counters that under the plaintiffs' reading of the statute, every administration--including the Trump administration--was apparently in violation of the statute by paroling anyone. Moreover, the government notes that utilizing the "may return" authority of (C) requires an agreement from a foreign sovereign (either Mexico or Canada, as the provision is limited to persons arriving from "contiguous territory"), which will not necessarily be forthcoming. The government gets around "shall" by pointing to other contexts in which the courts have said that while "shall" is generally obligatory, it isn't necessarily so. It makes good use of Justice Scalia's majority opinion in Castle Rock v. Gonzales, which held that the use of "shall" did not suffice to displace "[t]he deep-rooted nature of law-enforcement discretion."

Needless to say, that's not the end of the back and forth. However, I want to bracket the question of who's right to look at the broader issue of agency discretion not to act in various ways. In both Biden v. Texas and the litigation over DACA and related programs, Democrats have argued for the power to exercise prosecutorial discretion not to enforce immigration law to the full extent possible, while Republicans have taken the opposite position. The then-4-4 Court didn't produce a full opinion when Texas challenged related programs, but it is not difficult to imagine that since the seating of the three Trump appointees, the current Supreme Court--if faced with that issue again--would have invalidated the programs that expanded DACA.

Yet while Democratic administrations have made liberal/progressive use of enforcement discretion in immigration cases, it is hardly clear that over the long run a broad view of agency non-enforcement discretion is good for liberal/progressive causes. It is easy to imagine a future Republican administration invoking prosecutorial discretion to under-enforce or not enforce laws regulating firearms, environmental pollution, abusive business practices, and any number of matters that Congress tasked the executive with addressing.

Although there is much of concern in the Fifth Circuit opinion in the MPP case, Judge Oldham's statement that "Congress can rebut the common-law presumption that nonenforcement discretion is unreviewable" is both correct as a matter of law and ideologically even-handed. Indeed, there's at least a plausible argument that over the long run constraints on executive non-enforcement discretion will be better for progressives than for conservatives. After all, as a general matter, progressives want the government to regulate, whereas conservatives tend to be skeptical of regulation.

(2) The second issue is more technical and thus I won't attempt to address it in detail. I'll simply observe that it too presents mixed ideological stakes. At bottom, the question is how much an agency must do to validate a policy that was adopted in a way that was initially flawed. Stated at that level of generality, that question was the subject of three high-profile cases during the Trump administration, two of which were defeats for Trump.

The Court held that Trump's DACA rescission was arbitrary and capricious, notwithstanding the post hoc creation of a memo that purported to give better reasons than those originally offered. It also invalidated the Trump Commerce Dept's effort to add a citizenship question to the census, given the evidence that Commerce Secretary Wilbur Ross tried to manufacture a pretext of Voting Rights Act enforcement in defense of the action. However, the Court didn't block the third version of the Travel Ban based on the efforts of the Departments of State and Homeland Security to sanitize what began in Trump's anti-Muslim animus.

Thus, here too the long-term ideological stakes are unclear. Close review of procedural regularity in agencies will tend to frustrate those agencies--which one will favor when the agencies are up to no good. There doesn't seem to be a systematic bias here, however, because even if we begin with the view I advanced above--that other things being equal, progressives want government to act and conservatives don't--the "no good" in which agencies might be engaging can be accomplished either through action or inaction.

* * *

To reiterate a point I made above in an aside, it's probably naive to expect the courts to apply principles of administrative law or anything else in a way that is blind to the substantive policy stakes. Thus, if you are a betting person, you would not bet that any of the Democratic appointees will rule against the Biden administration's rescission of MPP or that any of the Republican appointees (with the exception of Chief Justice Roberts) will rule in its favor. Put differently, legal realism tells us that the Justices will see this case chiefly as about immigration policy, not administrative law or statutory interpretation.