Wrong in 1993, Still Wrong in 2026: Al Otro Lado and Incompetence in the US Supreme Court

[Guy S. Goodwin-Gill is Emeritus Fellow of All Souls College, Oxford, Emeritus Professor of International Refugee Law, University of Oxford]

In 1993, the U.S. Supreme Court allowed the Coast Guard to stop Haitians on the high seas and return them to the country they had fled, despite international law prohibiting such conduct. In 2026, the Court allowed U.S. officials to stop asylum seekers at the southern border, forcing them to remain in Mexico, and to wait indefinitely for any claim to protection to be assessed.

Article 33 of the 1951 Refugee Convention uses the French word non-refoulement when it prohibits return “in any manner whatsoever” of any person who may face persecution or other serious harm. In 1980, Congress passed the Refugee Act with the express purpose of conforming U.S. refugee law and practice with international law. Specifically, it included a procedure for a noncitizen “physically present in the United States or at a land border or port of entry,” irrespective of their status, “to apply for asylum,” thus ensuring that any claim would be assessed prior to any action on removal or deportation.

Although the decision of the U.S. Supreme Court is necessarily about an international matter – the protection of those seeking refuge and asylum – the majority does not deal in any meaningful way with the law or the facts. “Metering,” as the government euphemistically though wrongly calls it, follows from an initial refusal by a U.S. official to hear and assess, plus a turnback. Without an inquiry and a formal assessment of a fear of persecution or the risk of danger in Mexico, the principle of non-refoulement is presumptively violated.

No doubt Justices Alito, Roberts, Gorsuch, Thomas, Kavanaugh and Barrett found it too hard to deal with international law, the weight of opinions from the U.N. Refugee Agency in its supervisory role, the plethora of “foreign” judgments confirming the meaning of refoulement, or the consistency of academic opinion presented in an amicus brief led by the Global Strategic Litigation Council and co-signed by 140 leading academics, experts, and NGOs from across the world. All pointedly supported the applicability of non-refoulement at the border.

Nowhere in the majority judgment are the international obligations of the United States examined. In just a single, short paragraph, Justice Alito invokes the “precedent” of the 1993 judgment in Sale v. Haitian Centers Council Inc. to argue that somehow the principle against the extraterritoriality of legislation limits the protection of refugees. But as Justice Blackmun said in Sale, the Court was dealing:

“with a statute that regulates a distinctively international subject matter: immigration, nationalities, and refugees. Whatever force the presumption may have with regard to a primarily domestic statute evaporates in this context.”

That is part of why Sale was wrong in 1993 and why Al Otro Lado is also wrong in 2026. The Court quotes Sale again as saying that Article 33 of the 1951 Convention “cannot reasonably be read to say anything at all about a nation’s actions” towards noncitizens “outside its own territory.” But if they had only looked over the wall, they would have seen what has been the case for decades now, evidenced in the practice of other States and countless judicial decisions that, as Justice Scalia once notably said, ought to be considered as valuable guides to the meaning of words.

Instead, the Court reduced its analysis to one that is both facile and simplistic: From the perspective of domestic law, was the Government obliged to examine and determine asylum applications from those who arrived at a port of entry but who “have not yet stepped across the border.” Much of the analysis for both the majority and the dissent turned on statutory interpretation – whether asylum seekers have “arrived” in the United States when they are “at” the port, irrespective of where exactly their foot may be.

But international law sees things differently. As soon as they are subject to the authority and jurisdiction of State officials, asylum seekers are entitled to be heard before they are turned away or required to remain in a particular country. And the interpretation of Article 33 requires a proper consideration of its terms in good faith, according to their ordinary meaning in context and in light of the object and purpose of the 1951 Convention and the 1967 Protocol.

The Court, true to form, put on its blinkers, concluding that Article 33, “surely allows” border officials to stop those without valid travel documents from entering, and merely obliges States not to refoule those “within their borders.” (I learned in elementary school that the surest answer to any argument that begins “Surely…” is one that begins “Surely not…”)

“Metering,” claimed the Court, does not operate as a permanent bar, but this misses the point of its immediacy, the consequences of which are caught succinctly by Justice Sotomayor in her dissent. She notes how a Mexican may be forced to remain in the very country from which they were trying to flee owing to a well-founded fear of persecution, to which one can add that anyone of any nationality arriving and then “metered out” may have a fear of violence or death if required to remain beyond the border. The Court, she says, now allows everyone to be turned back, “even if the asylum seeker is certain to be persecuted or killed” – precisely the result which nonrefoulement and the object and purpose of refugee protection are intended to prevent.

The majority’s interpretation of “arrives” makes no sense, any more than its spurious attempt to invoke its 1993 decision in Sale. Justice Blackmun then noted in dissent that the terms of Article 33 are unambiguous, and that the majority in Sale had difficulty with the use of the words “return” and “refouler.” Return, it said, did not actually mean return, and it relied not on international practice for evidence of actual meaning, but on the peculiar American law distinction between deportation and exclusion, claiming that Article 33 somehow only applies to the exclusion of those on the threshold of initial entry. As Justice Blackmun clearly stated:

“the Convention does not ban the ‘exclusion’ of aliens who have reached some indeterminate ‘threshold’; it bans their ‘return.’”

The Court in Sale had in fact accepted that the “ordinary meaning” of refouler is to repulse, to drive back, to repel, and that is exactly what was happening to Haitian refugees in 1993 and has happened with the practice of metering at the southern border of the United States. To deny that this is refoulement in either case, is to spin words into meanings that have no reference to the goal of protection – the object and purpose of the 1951 Convention; and neither are they rooted in the treaty which U.S. domestic law is supposedly implementing in good faith.

There are other worrying dimensions to the decision in Al Otro Lado, not the least being that it gives people smugglers an added bonus – now they must actually get refugees into the U.S. and so will profit additionally from the vulnerabilities of the displaced (just the sort of business on which the present regime appears to thrive…).

A further point of concern had also been noted by Justice Blackmun in Sale, and this is the argument that in immigration and refugee matters, “the President and not Congress has sole constitutional authority.” Justice Thomas, concurring, hinted at such, while Justice Sotomayor observed that the majority’s interpretation “vests limitless discretion in the executive to abandon its inspection and processing duties,” which will make it all the harder to call the U.S. to account for the blatant violation of clear international obligations. “More people will die,” she wrote:

“More people will turn back and be subjected to violence because of something they cannot or should not have to change about themselves, such as their race, religion, nationality, or political opinion.”

In failing to uphold the principle of non-refoulement against the interpretation hitherto accepted by States at large, the Court has become once again a party to the breach of international law. It has shown that the system of refugee protection formally adopted in the Refugee Act of 1980 has become a sham, as a result of the anti-migrant and anti-refugee policy and rhetoric of the President. The result, all too clearly, is that notwithstanding the continuing efforts of many lower federal courts, no noncitizen in need of protection can ever be sure of being heard, let alone of enjoying asylum.

In the years after Sale, no other country followed U.S. practice or otherwise endorsed it. On the contrary, State practice and jurisprudence in both domestic and international courts developed to support a more protective scheme. The climate is starting to change, however, and many governments are looking for alternatives to the goal of protection and durable solutions; it may be that they, too, will think about re-interpreting their obligations.

Is Al Otro Lado yet another example of American exceptionalism? From time to time, this emerges not so much out of a sense of global responsibility and commitment to a rules-based order that includes human rights and refugee protection, but rather as an isolating doctrine that rejects universal values, such as freedom and equality. No doubt Justices Roberts, Alito, Gorsuch, Thomas, Kavanaugh and Barrett were exceptionally self-isolating in their analysis; they were also exceptionally persistent in ignoring the relevant international law that tells them how to interpret treaties in good faith, confirms that their application takes place between States that are also parties to those treaties, that the meaning of words comes not only from the text itself, but from practice developed over time, and that binding interpretations can be the product both of consistent decisions by the courts of fellow State parties.

Though they will not appeal to the present regime, there are alternatives to mass deportations and wholesale turnbacks, among them, international cooperation in protection and solutions. Together with renewed funding, increased staffing and technological resources, and the active involvement of civil society, non-governmental organizations, and agencies such as the United Nations High Commissioner for Refugees and the International Organization for Migration, effective responses can be developed and implemented, which are in tune with human dignity and protection needs.

We all know the suffering that drives the refugee, and can now better understand the positive contribution that refugees are able and willing to make, if allowed to do so, to the communities that receive them. Their lives, their protection and the future of their children, depend on the good faith incorporation of obligations such as non-refoulement, and on a commitment to vigilant oversight – clearly, there is much still to be done