Criminal Justice & the Rule of Law

Judicial Imperialism and the “Remain in Mexico” Ruling

David A. Martin
Tuesday, February 22, 2022, 9:01 AM

Late last year, the Court of Appeals for the Fifth Circuit affirmed a sweeping injunction against President Biden’s termination of the Migration Protection Protocols by grossly distorting statutory provisions enacted in 1996.

U.S. - Mexico border wall (Official photo by Robert Bushell, U.S. Customs and Border Protection)

Published by The Lawfare Institute
in Cooperation With

In December 2021 the Court of Appeals for the Fifth Circuit affirmed a sweeping injunction against President Biden’s termination of the “remain in Mexico” program (also known as the Migrant Protection Protocols or MPP). The central part of the 117-page opinioncaptioned Texas v. Biden, presents itself as a straightforward correction of an egregious misreading by the agency of certain statutory provisions enacted in 1996. Under the court’s interpretation, Congress has denied the Department of Homeland Security (DHS) any discretion to terminate the program;  MPP is mandatory unless the agency detains virtually all arriving migrants whose admissibility is in question until their admission cases are resolved.

The court’s opinion carries the reader along on what purports to be textual analysis and implacable logic. On closer examination, however, it is a startling exercise in judicial imperialism. The opinion seizes on fragments of statutory text, taken out of context, to construct a presumed congressional intent that would be more to the judges’ liking. It ignores contrary indicators in the wording and the historical development of the key provisions. It makes no attempt to reconcile the supposed strict mandate with the historical fact that Congress went 20 years without really noticing—much less objecting to—the absence of implementation. The court also shows an arrogant disregard for the operational realities of border enforcement, including the sensitivity of diplomatic relations with Mexico that sustain cross-border cooperation—on migration issues as well as other policy priorities.

I can bring some special perspective in analyzing the appeals court’s decision.  I have been a scholar and teacher of immigration law for 40 years, and I also was fortunate to hold policy-level positions dealing with immigration in three different departments, under three different Presidents. My years in government gave me close exposure to the operational realities at a level most law professors—and judges—don’t experience. One of those stints consisted of 30 months during the mid-nineties as General Counsel of the Immigration and Naturalization Service (INS) the period when the key reform bills on which the Fifth Circuit relies were introduced, debated, amended, enacted and implemented.

MPP and the “Contiguous Territory” Provision: Background and Structure

The Trump administration launched the MPP program in January 2019 to help deter a surge of migrants—overwhelmingly composed of Central American asylum seekers— at our southwest border. It called for U.S. border officials to do preliminary processing, including the issuance of a charging document, of undocumented non-Mexicans arriving by land before sending most of them promptly back to Mexico to await the next steps. The individuals would be scheduled, months later, for a brief return to a U.S. immigration court hearing and would eventually receive a decision to admit or reject. 

The Trump administration found authority for its actions in an obscure provision, section 235(b)(2)(C) of the Immigration and Nationality Act (INA), enacted in 1996 (8 U.S.C. § 1225(b)(2)(C)). It states that, for people arriving by land from “contiguous territory,” the Secretary of Homeland Security “may return the alien to that territory pending” an immigration court proceeding. For these purposes, only Mexico and Canada count as contiguous territory.

Over 68,000 persons were returned to Mexico under the MPP. This practice did help advance President Trump’s objective of deterring undocumented migration. After it was fully implemented throughout the spring of 2019, border officer’s encounters with such migrants at the southwest border decreased from 145,000 in May 2019 to fewer than 35,000 the following March. Critics charged, however, that the MPP, as implemented by Trump’s DHS, hampered legal representation, made grossly inadequate provision for housing during the Mexico waiting period and exposed the migrants to rampant crime in the border region. President Biden fulfilled a campaign promise when, on inauguration day, he directed DHS to suspend the MPP program and move toward termination. DHS issued a seven-page termination memo in June 2021. 

Not surprisingly, the number of arriving undocumented migrants rose steeply. In response, Texas and Missouri swiftly filed suit in federal district court in Texas. The district court enjoined the MPP termination in August. Both the Fifth Circuit and the Supreme Court left the injunction in place while further appeals proceeded. The appeals court then issued its full opinion in December 2021. On February 18, 2022, the Supreme Court granted the Biden administration’s petition for review and expedited the briefing schedule so that the case can be heard and decided before the summer recess.

The Picture Painted by the Fifth Circuit

The appeals court decision tells the following story about the legal framework and the sequence of actions taken by the Trump and Biden administrations:

In the mid-1990s, Congress was exasperated with the executive branch’s regular failure to detain applicants for admission to the U.S., despite an INA provision saying that “[e]very alien [seeking admission] . . . who may not appear to the examining immigration officer to be clearly and beyond a doubt entitled to land shall be detained” for an immigration hearing. The court acknowledges that the agency historically has lacked the resources to detain every such person. For decades, that mismatch of apparent obligation and limited detention capacity has been accommodated (or rationalized) by invoking the immigration parole authority, a longstanding agency practice given explicit statutory authorization when the INA was enacted in 1952. An arriving migrant granted parole can be allowed temporary physical presence in the U.S., free of detention, but is not considered “admitted.”

Parole pending an admission hearing has been a routine practice for the 70 years since the INA was adopted. It is granted to tens of thousands each year, who are then released with agency documentation showing parole status—and often with other conditions attached. When the hearing date arrives, a share of these parolees win their cases and gain admission, but a sizable number fail to show up at stages of the proceedings or after losing their immigration court case.  Most of those absconders thereby achieve an illegal stay.

Congress’s displeasure with these releases, according to the court of appeals, resulted in the inclusion of three key provisions in the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 (IIRIRA): 

  •  The “shall be detained” mandate was reaffirmed in INA § 235(b)(2)(A), with slightly revised wording to accommodate technical changes elsewhere in the statute. 
  •   Under INA § 212(d)(5), parole now can be granted “only on a case-by-case basis for urgent humanitarian reasons or significant public benefit.” This replaced the 1952 formulation, which had allowed parole “for emergent reasons or for reasons deemed strictly in the public interest.”
  • And, in the court’s words, “Congress expressly authorized the MPP program by statute,” citing § 235(b)(2)(C).

That third section of the statute does not read as though it creates a “program,” much less a mandatory one. It states that DHS “may return” a person “who arrives by land from a foreign territory contiguous to the United States” to that territory pending an immigration hearing. But in the court’s apparent view, the MPP provision has absorbed the mandatory character of its nearby subparagraph (b)(2)(A). Detention, the court reasons, is the default position, set forth there in a “general, plainly obligatory rule.” [Emphasis added.] The statute then “authorizes contiguous-territory return as an alternative.” The court sees § 235(b)(2)(C) “as a statutory safety valve to address [the] problem” of inadequate detention space. DHS’s “refusal to use the statute’s safety valve,” combined with its “lack of detention capacity, means DHS is not detaining the aliens that Congress required it to detain.”   

But what about parole? The court refuses to see that as a true alternative, nor as a statutory safety valve, because “parole can be exercised only within narrow parameters (case-by-case and with a public-interest justification).”  The opinion elaborates: “simply parol[ing] every alien [whom DHS] lacks the capacity to detain . . . solves nothing. . . . Deciding to parole aliens en masse is the opposite of case-by-case decisionmaking.” Termination of the MPP is therefore enjoined “until such a time as the federal government has sufficient detention capacity to detain all aliens subject to mandatory detention under [INA § 235] without releasing any aliens because of a lack of detention resources” —a standard almost surely unattainable. Limited parole releases based on individual circumstances are still permitted by the court, however, because they are case-by-case.

What Congress and the Agencies Actually Did

The picture the court paints differs drastically from what really went on in 1995-96 as the IIRIRA was being debated and enacted. First, a word about the political setting. The 1994 midterm elections had brought Republican control of both the House and the Senate for the first time in 40 years. Proposition 187 had passed in California with 59 percent of the vote, enacting state-law measures meant to crack down on unauthorized migration. Both parties wanted to head into the 1996 presidential election with a record of passing strong immigration enforcement measures, though there were substantial differences on specifics. Building on committee hearings held early in the decade, both the Clinton administration and the Republican congressional leaders developed their own lengthy, detailed enforcement bills.

A mandatory detention program that nobody thought to implement for 20 years?

If the MPP provision was a crucial innovation—indeed a mandatory program—meant to overcome the resistance of the agencies to detaining as they should, one would expect it to have claimed prominent attention from the drafters and the judiciary committees, as well as lobbyists, advocates, and the media. No such attention was accorded. The conference report provides only a one-sentence reference to § 235(b)(2)—never even mentioning the “contiguous territory” portion.    

This silence is particularly telling because another section of the bill—an unmistakable detention mandate—did provoke robust debate and negotiation between Congress and the President’s high-level representatives. That provision, a Republican proposal, required INS to detain “criminal aliens” when released from their criminal confinement, to have them securely in custody for removal. The provision gave the agency extremely limited authority to release (only in connection with the witness protection program).  

The Commissioner of INS listed this costly and novel detention mandate for criminal aliens as one of the top 11 administration concerns she presented in individual meetings with the Republican conference committee members in the final days before passage. (As General Counsel, I accompanied her in those meetings.) The Republican leadership refused to relax the substantive criteria mandating detention, but they did eventually offer amelioration. They allowed for a delay of the effective date by up to two years, so that the agency could adjust its budgets and acquire the necessary staff and detention capacity. No one mentioned any comparable need to budget and hire for the launching of a “remain in Mexico” system.

More than 20 years elapsed after enactment before this provision drew serious executive branch attention. Even then, the operational steps taken by the Trump administration do not match the court’s imagined model.  Their public MPP documents never suggest any legal obligation to proceed. Instead, the implementers wanted to take credit for dusting off and deploying a neglected tool.

Who came up with this idea?

Another part of the drafting history is revealing. The Fifth Circuit portrays § 235(b)(2)(C) as part of an overall effort by a Republican Congress to expose Democratic dereliction at the border. In this account, the new mandatory tool was intended, from the beginning, to shame or coerce the executive branch into ending the release of inadmissible noncitizens into the U.S.

But that framing doesn’t fit the chronology. The “contiguous territory” provision was not part of the early Republican drafts of the reform legislation. Instead, it makes its appearance in a 1995 draft bill (H.R.1929, § 106) prepared by the Clinton Administration. That provision was apparently seen as a minor technical addition to permit localized return practices that were being questioned by immigration judges. It was finally incorporated into a significant Republican measure when the Senate Judiciary Committee crafted a comprehensive bill (S. 1664, §141(a)) and reported it to the floor in April 1996.

Parole: Taking Plain Meaning Seriously

Parole is a more challenging conceptual and practical issue—both for the courts and for the political branches. Since President Eisenhower first used statutory parole to take in Hungarian refugees in 1956, members of Congress have complained about overuse of the parole power—even while they mostly supported or acquiesced in specific parole programs. The congressional critics generally charge that parole has been used to evade the requirements and limitations Congress has placed on regular admission categories. The real challenge for legislative response, however, has been how to draft language that identifies workable parole limits while still permitting the timeliness and flexibility that are the whole point of the parole power.

In 1996, Congress went beyond generalized grumbling in committee reports and actually amended the parole statute. The Fifth Circuit panel relies heavily on those changes in its argument that use of the contiguous-territory provision is mandatory. But the argument, coming from judges who present themselves as rigorous textualists, fails to attend seriously to the amendment’s actual wording.

To develop this point, let’s examine closely the key passage from the opinion:

As the district court found, DHS lacks the resources to detain every alien seeking admission to the United States. . . . So it's left with a class of people: aliens it apprehended at the border but whom it lacks the capacity to detain. By terminating MPP, DHS has refused to return that class to contiguous territories, as permitted by [INA § 235(b)(2)(C)]. The Government's position thus boils down to this: We can't do one thing Congress commanded (detain [arriving aliens]), and we don't want to do one thing Congress allowed (return [to Mexico]).

Parole does not provide a way out of the box created by DHS's can'ts-and-don't-wants. . .

[INA § 212(d)(5)] gives DHS the power to parole certain aliens “only on a case-by-case basis for urgent humanitarian reasons or significant public benefit.”  DHS cannot use that power to parole aliens en masse; that was the whole point of the “case-by-case” requirement that Congress added in IIRIRA. . . .  So the Government's proposal to parole every alien it cannot detain is the opposite of the “case-by-case basis” determinations required by law. 

The court’s phrasing is both deceptive and sharply out of keeping with the operational realities, for a few key reasons. First, the government has never “propos[ed] to parole every alien it cannot detain.” That is not how the issue arises. At any given time, each field office has some space for detention and a (usually) larger pool of apprehended violators. None of those suspected violators comes into custody with a label reading “cannot detain.” Every field office makes choices based on individual characteristics—in order to use wisely the detention space it does have. The immigration agencies have protocols and standard forms to guide the required interview and to capture case-specific information useful in deciding whom to release. The information can also help locate those who do not appear for the hearing (or for removal, if denied admission).  

Second, the Fifth Circuit opinion—on three occasions—describes Congress’s amendment of the parole provision as intended to end the use of paroles en masse. But when has the executive branch ever paroled en masse? In my 43 years of involvement with immigration law and administration, I have never seen such a parole. The court offers no examples, because they don’t exist. Even in a situation like the Mariel boatlift of 1980, with thousands of Cubans coming daily to south Florida in private boats (totaling 125,000 over five months), no groups were paroled en masse. The same is true of the roughly 70,000 Afghan evacuees who arrived last summer and fall after the chaotic troop pullout from Kabul.  

Arrivals, in these circumstances, are first taken to an established separate area for individual immigration processing. It might be a Border Patrol station, a secondary inspection area within a port of entry or, in mass-influx situations, a school gymnasium, a tent, a military base, or part of the locality’s emergency response facilities. The new arrivals are interviewed individually (or as a family), and their fingerprints are taken. Law enforcement and terrorism databases and watchlists are checked. Various digital and paper records are created. After this processing, some of the individuals are retained in custody and others are released.

Every parole, large- or small-scale, has conformed in this fashion to the statutory requirement of action “on a case-by-case basis.” Are there also other policy factors at work in specific parole decisions? Of course. But the plain terms of the parole statute allow–and indeed require–the introduction of other policy judgments. The statute gives authority to parole “for urgent humanitarian reasons or significant public benefit,” as determined by the Secretary of Homeland Security. Decisions about significant public benefit clearly call for value judgments that transcend individual interviewing.

There can and should be vigorous debate over whether a particular parole practice or program is wise or justified—whether it really provides a “significant public benefit.” But that is for the political and policy arena. A standard this vague provides no solid footing for judicial second-guessing—at least absent utterly irrational action by administrators. Releasing selected arrivals on parole when detention space is tight, after systematic interviews and security checks, is by no means irrational, even if a large-scale influx, coupled with inadequate resources, ultimately forces expansive use of releases.  

It is quite true that Congress often has manifested a general urge that use of the parole power be reduced; that urge lay behind the 1996 addition of the “case-by-case” specification. But as Justice Antonin Scalia has insisted, in his largely successful campaign to proselytize plain-meaning textualism, courts don’t enforce general urges of Congress, even if manifested in stray parts of the legislative history: “We are governed by laws, not by the intentions of legislators.” 

The actual words Congress chose for the 1996 law amending the parole standard do not require that all paroles be small-scale, nor do they render parole illegitimate when used to release a portion of arriving migrants awaiting an admission decision—no matter how hard the panel judges wish that were the case.


The remaining administrative procedure questions in the lawsuit pose complex issues, in light of the Supreme Court’s 2020 DACA decision. Those controversies are irrelevant in this case, however, if the MPP cannot be terminated, no matter how good the agency’s process or its prose.

The Supreme Court should vacate the statutory construction perpetrated by the Fifth Circuit panel and reaffirm the legitimacy of parole pending a hearing. Use of the “remain in Mexico” provision is clearly discretionary, as its plain language indicates.  

Or to summarize using the opinion’s terminology: parole, unaided by the MPP, does provide a way out of the “shall detain” box. It has been doing so for at least 70 years.  

David A. Martin is Warner-Booker Distinguished Professor Emeritus at the University of Virginia School of Law and also a fellow of the Miller Center for Public Affairs and of the Migration Policy Institute. He served in the State Department’s Human Rights Bureau 1978 to 1980, as general counsel to the Immigration and Naturalization Service, 1995 through 1997, and as principal deputy general counsel to the Department of Homeland Security, 2009 through 2010.

Subscribe to Lawfare