President Biden came to office promising to undo scores of new immigration policies instituted by the Trump administration. On inauguration day, Biden suspended one of the most controversial policies of his predecessor: the Remain in Mexico policy, also known as the Migrant Protection Protocols (MPP). That policy had radically altered the screening of asylum seekers along the southern border, requiring asylum seekers to wait in camps on the Mexican side for their claims to be heard. On June 1, 2021, Biden’s Department of Homeland Security terminated the program for good. Or so the administration thought.

A district court in Texas blocked the move and ordered the administration to reinstitute the program, citing procedural and substantive defects in DHS’s new policy. Under the new policy – which is mostly a return to the status quo that existed prior to the creation of the Remain in Mexico policy – DHS would have detained some arriving asylum seekers and released others on “parole” pending adjudication of their claims, rather than forcing them all to wait in dangerous and unsanitary makeshift camps in Mexico, in the midst of the COVID-19 pandemic.

After trying a second time, this past October, to terminate the program and address the deficiencies cited by the district court, the administration has again been blocked, this time by the Fifth Circuit. In an opinion issued last Thursday, the Fifth Circuit rebuffed the Biden administration’s second rescission decision, effectively forcing a reinstatement of MPP. 

Judicial oversight of presidential immigration law is crucial (as we ourselves have argued at length). The problem with the Fifth Circuit’s decision, however, is that it delays policy change with pointless procedural hoops while simultaneously making change of any kind more difficult by ratcheting up the status quo bias of administrative law. All of this threatens policy development and democratically-driven change not just in immigration policy but across the entire administrative state – and in ways that do little or nothing to promote the values of judicial oversight.

The opinion also happens to upend and misunderstand decades of immigration law and policy making. The court effectively reads the Immigration and Nationality Act (INA) to require the government to adopt something like MPP, unless it can otherwise detain each and every asylum seeker as they pursue the claims for protection that the law entitles them to make – a logistical impossibility in times of crisis, and a sharp break from what courts and policymakers alike have understood the law to require. 

Public Law and Policy Change

At least three aspects of the court’s conception of administrative law and the role of the courts in reviewing agency action reflect a strong and, in our view, inappropriate and damaging status quo bias. Each of these twists is likely to interfere with the ability of new administrations to undo or rethink the decisions of their predecessors, including those the new administration believes to be harmful, counterproductive, and inconsistent with the policy agenda the new president was elected to pursue. And none of these aspects of the Fifth Circuit’s opinion ultimately serves the rule of law values that judicial review in this context is designed to protect. 

First and foremost, the Fifth Circuit’s opinion reflects a deeply misguided view about the role of reason-giving and justification in agency decisionmaking. 

A core aspect of administrative law is that agencies, unlike legislatures, are required by law to give reasons for their decisions. But an equally foundational principle is that agencies are permitted to rethink their policies and positions, including by developing new justifications for an action initially found wanting by a court. This bedrock principle, established at least as early as 1946 in SEC v. Chenery, leaves policy development in the hands of the agencies and executive branch officials, who have the statutory authority, institutional tools, and political legitimacy that courts lack. 

In its MPP decision, the Fifth Circuit casts aside the black-letter rule that agencies are free to support with new reasons a policy whose initial justification was found wanting. Instead, the Fifth Circuit treats it as deeply troubling for an agency to re-paper a decision by revising its explanation of what justifies a policy. Rather than understand the agency’s efforts to re-do its decisionmaking process as part of a good-faith dialogue between agencies and courts, the Fifth Circuit treats the administrative effort as some sort of gamesmanship – a reason to doubt the validity of the policy even more deeply.

This view is reflected in the court’s bizarre assumption that the policy decision it is evaluating somehow exists independently of (and hence cannot be supported by) the actual reasons given by the government for that decision: 

[T]he Government misunderstands the States’ challenge. The States are challenging DHS’s Termination Decision – not any particular memo that DHS might have written in the past or might write in the future. [emphasis in original]

In other words, the reasons the government gives to justify a policy change are beside the point; the administration’s efforts to cure the defects identified by the first district court decision enjoining the rescission of MPP have no legal significance. 

But this flies in the face of long standing administrative law doctrine, under which the reasons the government gives for its actions help determine their legality. If the government cannot go back to the drawing board to rethink its policy decisions once faced with an adverse court ruling and then reach the same conclusion, only for different and better reasons, then the government could be forever stuck with a policy not required by law because it failed to satisfy a district court in the first instance. That approach empowers courts and thwarts change without serving any demonstrable legal values. 

Second, the Fifth Circuit appears to be ratcheting up arbitrary and capricious review into a form of interventionist administrative law that empowers courts to second guess agency policy judgments and block policy development. Part of the fault for this lies with the Supreme Court’s decision in Regents v. DHS, the decision that rejected the Trump administration’s efforts to unwind DACA. The Fifth Circuit parrots the two flaws the Supreme Court identified in the DACA case – that the government failed to take account of reliance interests and did not adequately consider alternatives – and purports to find these same flaws in DHS’s termination of MPP, holding that DHS did not consider states’ reliance interests in MPP or alternatives to ending the program. 

As we have written before, the real problem with the DACA rescission was that the reasons actually offered by the agency for its rescission were not legally sound. Because the Trump administration preferred to avoid political accountability for its decision to terminate the program, the administration tried to hide behind the erroneous conclusion that DACA was unlawful and therefore had to be rescinded, rather than acknowledging that, as a matter of discretion, it preferred to end the policy. While this problem with the initial rescission was reflected in many of the decisions in the sprawling litigation spawned by the rescission attempt, Chief Justice John Roberts was ultimately unwilling to make this point plain in his opinion for the Court. As a result, Regents can be read to increase status quo bias, empowering courts to stall policy change by inventing policy alternatives that a judge believes the agency should have considered. 

The Fifth Circuit relies heavily on this aspect of Regents to conclude that the MPP decision was arbitrary. (slip op. pp. 95-98). But while the court concludes that the Biden administration “insufficiently addressed alternatives to terminating MPP,” it doesn’t even bother to suggest a single policy alternative that it believes the agency should have considered – let alone point to any alternative that was actually presented to the agency during its decisionmaking process (or even during litigation!). 

This is a far cry from the canonical failure-to-consider-policy-alternatives holding in Motor Vehicle Manufacturers’ Association v. State Farm on which Regents relied. In State Farm, the agency completely failed to consider a well-developed policy alternative that was under active consideration by the agency. In sharp contrast, the Fifth Circuit seems to authorize courts to set aside any policy when the court itself thinks that the agency should have spent more time thinking about (some entirely unspecified set of) alternative policies. This part of the Fifth Circuit’s opinion amounts to a demand without end, one that a court can invoke no matter what the agency takes into consideration. 

In this way, the Fifth Circuit distorts the Supreme Court’s most recent arbitrary and capricious decisions – not only Regents, but also its conclusion in New York v. Department of Commerce that the Trump administration’s efforts to add a citizenship question to the Census were illegal. Both of those decisions made clear that the government must provide its real reasons for acting, in order to ensure that it can be held accountable. In both Regents and in New York, the basic problem was that the government’s purported reason for acting was pretextual. But neither of these holdings authorized the sort of deep inquiry into policy reasoning that the Fifth Circuit seems to advance in this most recent decision, even if such an outcome might have been a predictable consequence of the way Regents, in particular, was written.

Third, the Fifth Circuit has forced not just this administration but the government generally into a litigation vise that will have the effect of stymying change. Anytime a new policy is enjoined by a trial court, the Fifth Circuit’s opinion puts the government to a difficult choice: the government can appeal the district court injunction and wait for the eventual resolution of the validity of the policy at issue in the litigation; OR, the government can accept a district court injunction and attempt to re-do its policymaking process in order to satisfy that trial court. But it cannot do both simultaneously. 

It’s not hard to see why the government would want to pursue both strategies at once: the government both wants to defend the legality of its positions while also pushing forward with important policy changes that are central to the administration’s objectives. This was as true for the Trump administration in the DACA litigation as it is for the Biden administration here. And this dilemma will only become more prevalent given the state of today’s politics, marked by (1) high levels of polarization (and thus sharp policy differences across administrations); (2) executive-driven policy, not just in immigration but across policy domains, because of congressional weakness or incapacity; and (3) increasingly aggressive judicial oversight of administrative decisions by the lower courts in particular. 

Basic strategies of administrative governance would have the government return to the drawing board to correct the legal deficiencies identified by the court and then to defend the new policy through litigation, or to respond to subsequent rulings through policy development. But the appellate lawyer is likely to want to persuade the court of appeals that the trial court got it wrong, and the government as a whole has a deep interest in the appeal, too, in order to defend both its specific policies and its options for administrative decisionmaking. 

This two-pronged strategy enables the government to re-do its initial policy without having to admit that its initial decision was defective, which is important politically and may in fact be correct as a matter of law, the judgment of a single district court notwithstanding. Indeed, in the MPP case, had the government declined to appeal the original district court decision, the district court’s conclusion about the scope of the parole power could have become the law of the case and led to the invalidation of the second attempt at ending MPP through the use of parole. As we explain below, we happen to think that the district court and the Fifth Circuit got the legal question wrong. But even if that’s up for debate, it’s not a debate the government should be forced to give up once a trial court has weighed in. 

But again, the Fifth Circuit basically prohibited the government from pursuing this two-pronged approach, further empowering lower courts. At the very least, the ability to pursue an appeal at the same time that the government attempts to adjust the underlying policy in order to satisfy the trial court seems like one way to mitigate the concerns, voiced across the ideological spectrum, about trial-court-issued nationwide injunctions. 

Instead, the court’s approach exacerbates the incentives for an administration’s adversaries to use litigation to delay policy change. This strategic use of litigation will certainly appeal to activists who see stopping the depredations of their opponents as worth the cost of hamstringing policy change more generally. But we think this is a distortion of judicial review of agency action and destructive of the administrative state (which in the end may be the point of some of this litigation and the interventionist administrative law that follows). Of course there can be good reasons for wanting policy change to take time, such as to ensure deliberation or wide ranging public input and to guard against arbitrary government action. But the brakes applied by this aspect of the Fifth Circuit’s holding promote none of those values. 

What Presidential Immigration Law?

Beyond general concerns with the function of the administrative state raised by its decision, the Fifth Circuit comes to a stark conclusion about the Immigration and Nationality Act in validating the claims presented by the state of Texas a conclusion unsupported by the text and history of the INA. The court concludes that the government has a binary choice: either it must detain all arriving asylum seekers at the southern border or force them to wait in Mexico for the adjudication of their claims. 

We’ve said it before and we will say it again: the INA is full of incredibly detailed rules and procedural minutiae that might lead one to believe that it constitutes a comprehensive set of instructions dictating exactly how the executive branch should screen immigrants and enforce immigration law. But that view of the law is mistaken. 

It was mistaken when Hawaii and other states sued to invalidate President Trump’s ban on noncitizens from a number of majority-Muslim nations. That policy was unlawful in our view, despite the Supreme Court’s conclusion, because it unconstitutionally discriminated against people on the basis of their religious beliefs; but the ban did not violate the immigration statute. And it is mistaken again now, when Texas and other states claim that the immigration code mandates that every single noncitizen who shows up at the southern border without documents, seeking asylum, be either detained or returned to Mexico. 

The Fifth Circuit’s interpretation of section 235(b)(1) of the INA makes a hash out of the statute and belies the way its interlocking parts have been administered for decades, in two ways. First, the court begins from the premise that the statute mandates that all persons seeking admission but lacking documentation must be detained (8 U.S.C. s. 1225(b)(2)(A)) until their asylum claims are fully resolved by DHS. From there, it reasons that the statute provides only one exception to this mandatory detention rule – if the government sends these asylum seekers to await their hearings outside the United States (8 U.S.C. 1225(b)(2)(C)). 

Reading the statute to restrict the government’s choices to these two options ignores the fact that the immigration code also expressly empowers the executive to “parole” into the United States noncitizens who are otherwise inadmissible. This power, created by Congress in 1952, was used by successive Presidents in the second half of the twentieth century to invent American refugee policy. And for decades since, executive branch officials have used parole to permit many arriving asylum seekers to enter the United States pending the processing of their asylum claims. 

The Fifth Circuit acknowledges the availability of parole under the statute. But its cramped reading of that power ignores the long history of how the executive has used the parole power, even in the face of periodic congressional efforts to cabin that power. The court implicitly suggests that those asylum seekers the government would allow into the United States (rather than detain or hold in Mexico) pending their claims would not meet the urgent humanitarian need established by the statute. But parole in that context has been, we underscore, a common practice for decades, as has judicial deference to executive branch determinations about what constitutes a humanitarian need. 

The court also concludes that parole cannot be used for arriving asylum seekers because the parole provision requires that it be awarded on an “individualized” basis. Because there are potentially a large number of arriving asylum seekers who would receive parole if the Remain in Mexico program were terminated, the court reasons that their release would amount to a categorical, rather than individualized, exercise of the parole power. (This reasoning parallels the claim, made by Texas and other opponents of President Obama’s deferred action policies, that DACA is unlawful because it leads to the award of deferred action on a categorical, rather than an individualized basis.) This reasoning is wrong. Even when parole decisions are made according to an announced policy that ensures more uniform application of parole, and even if large numbers of applicants ultimately receive parole, parole judgments are nonetheless made on a case-by-case basis.

If the Fifth Circuit’s reading of the statute were correct, then the U.S. government has been violating the INA in plain view of Congress since 1996 when the provision regarding undocumented arrivals was enacted. Throughout that period, administrations have used parole policies to both adhere to the immigration laws and enable applications for asylum on humane terms under those very laws. The Fifth Circuit would conclude that all of those actions were unlawful and that undocumented arriving asylum seekers can never be released under any circumstances.

The Fifth Circuit’s erroneous understanding of the parole power would seriously interfere with the President’s power to manage asylum screening at the southern border, especially in moments of mass influx or during crises like that created by COVID-19. What is more, reading section 1225(b)(2)(C) as the sole exception to mandatory detention for arriving asylum seekers elides the fact that return to Mexico is an option expressly made available by the statute (and actually physically possible) only for those arriving by land from a contiguous territory. This provision cannot and does not apply to the millions of asylum seekers who arrive at the nation’s airports. On the Fifth Circuit’s reading, the executive would have literally no discretion to manage the asylum screening of those arriving at airports.


The Fifth Circuit’s decision has made it increasingly likely that it will be many more months, if not years, before the Biden administration is able to rescind a program that the Trump administration created as an exercise of executive discretion. This litigation, and the courts’ resistance to changes in administrative policy, in some sense mirrors the Trump administration’s failed efforts to unwind DACA, another policy that has proven to have more staying power than would be expected, given the supposedly defeasible discretion that supported it. 

Though there are reasons to distinguish between the government’s decisionmaking in each of these cases, the more important point to underscore is the courts’ seemingly growing skepticism of dramatic policy change and their assertion of authority to rigorously test the government’s reasons for change. Such testing might seem to promote accountability, but in the case of MPP it has prevented the government from bringing an end to dangerous and unjust treatment of asylum seekers, and forced upon the new administration a discretionary policy inconsistent with its professed objectives and with political commitments seemingly ratified by the voters in the last election. This judicial skepticism of change parallels the wariness with which the court approaches the government’s turn to the parole power, rejecting a long history of its use to manage refugee crises and inject humanitarian flexibility into the immigration law. Both aspects of the court’s decision misinterpret precedent to unjustifiably constrain government policymaking discretion in an area of law where courts have, until recently, deferred to the executive in recognition of both democratic legitimacy and agency expertise.        

Image: Aerial view of an improvised camp of asylum seekers and refugees at El Chaparral border crossing in Tijuana, Baja California state, Mexico, on December 6, 2021. – The United States reimplemented the Migrant Protection Protocol (MPP) program, also known as Remain in Mexico, on December 6 after a court order. (Photo by GUILLERMO ARIAS/AFP via Getty Images)