I. Introduction

In late October 2019 it was reported that more unaccompanied minors had been detained trying to cross the southwest border of the US over the previous 12 months than in any previous year on record. This was 76,020 children and adolescents travelling without their parents, mostly from Central America: 52% more than during the previous fiscal year. During that same time Mexico, under pressure from the US, detained around 40,500 underage Central Americans, bringing the total to 115,000. Many families headed North as well (^{Villegas, 2019}). And they were part of a much larger global phenomenon.

The world now has more forcibly displaced persons (refugees and internally displaced persons) than any time since World War II. The European Commission in July 2022 put the total number at 89.3 million, of whom 27.1 million were refugees, 53.2 million internally displaced persons (who have fled their homes but not left their countries), and about 4.5 million each of asylum seekers and Venezuelan refugees and forced migrants.² That figure amounts to more than one percent of the entire world population. 85% of those designated as refugees under current international law are hosted in “neighboring and developing countries, often for many years”, or even decades.³ The good news is that nearly 99% of the people in the world are not displaced from or within their home countries. The bad news is that rich nations’ responses have been, in general, hard-hearted. Indeed, refugee intake and support has gotten worse in the US and many other Western nations. After leading the world in refugee resettlement for decades, the Trump administration slashed the numbers by more than three-quarters. While the Trump administration is harshly criticized for its migration policies, the Central American migration crisis began under President Obama, dubbed “the deporter in chief”. President Biden is also being condemned in progressive quarters for not doing more to reverse Trump policies. Meanwhile, the political discourse around refugees and asylum seekers has grown harsher across the West: in the UK, Hungary, Austria, Germany, and France, among other places, as far-right leaders enjoy increased support and influence thanks in part to popular hostility to immigration and refugee admissions.⁴

There may be no area of public policy in advanced Western states more fraught with deep moral and practical dilemmas than immigration. The United States has general humanitarian duties to people in need, and it also has more specific obligations under international law to those who make credible claims for asylum to the United States. The scope of both our moral duties and legal obligations are at stake in the ongoing debates, surveyed here, on the question: “who is a refugee?”.

This essay is an attempt to think about this question and assess the literature that addresses it, especially from the standpoint of ethics and political theory, and a grounding in real-world problems. To do so, we need to consider the various moral grounds or categories that must be considered by policymakers or citizens who aspire to support morally defensible refugee policies. Those who are well off, whether individuals or states, have general humanitarian duties to aid those in great need; these exist quite apart from any past or ongoing relations with people in need. In addition, our past and ongoing relations involve certain basic requirements of fair dealing or at least decent treatment in our interactions with others. The violation of those standards, especially when powerful states like the US wrongfully and illegitimately harm the basic interests of poor and vulnerable states, may create obligations of restitution and repair. We must also consider that more specific obligations arise from our undertakings as a country in signing on to international agreements, such as the refugee convention. And finally, the very existence of the state system itself, as a global regime of governance from which we benefit, implicates us in the plight of refugees and gives rise to additional grounds for claims by those seeking asylum and resettlement.

This essay is a brisk survey of the moral and practical terrain, aiming to clarify the issues but not to settle them. It is intended as an introductory discussion for those interested in the debates around who should qualify for refugee status, especially in light of the predicament of Central Americans fleeing disorder. It pays special attention to the claim that the US has reparative obligations to Central American countries owing to US interventionist policies. Some of these historical relations, spanning much of the 20th century, are discussed by Enrique Camacho-Beltran, under the rubric of providing “a more complete normative panorama” of the migration crisis now confronting the US (^{Camacho, 2022, pp. 159-188}). The catalogue is extensive and includes CIA involvement in coups, abusive practices of US corporations, unfair trade agreements, US contributions to climate change and worsening environmental disasters, and drug laws that help enrich and empower murderous drug cartels. We might add to this extensive catalogue those asymmetric, neocolonial power relations that scholars like ^{E. Tendayi Achiume (2019)} argue undermine Third World sovereignty to the benefit of wealthy developed countries.

One additional consideration should be noted. Even progressive policy makers in the US who are sincerely motivated to repair historical injustices and provide succor to the neediest of would-be migrants must confront other factors with potentially profound significance. I refer to the various threats of domestic backlash associated with white nationalism and nativism. ^{Michael Blake (2020)} refers to these as “the bigot’s veto”.

What then should political leaders do? Simply act on their moral duties as best they can discern them and leave the consequences to take care of themselves? And what about citizens deciding which candidates to support, for example in Democratic primaries? Should they favor the moral idealists or the more pragmatic realists? It is easy to urge that we need more “non-ideal” political theory, but it is a messy business, as my discussion will illustrate.

I begin by laying out the various relevant moral grounds mentioned above. I endorse the liberal view of domestic and global justice as defended by scholars such as John Rawls, Michael I. Blake, and Anna Stilz, and (to some degree) Michael Walzer (^{Rawls, 2019}), members of self-governing political communities have special obligations to one another, but they also have outward looking duties and obligations based on general humanitarian considerations and also more specific relations with others.

I then examine the decades-long controversy surrounding the question of “who is a refugee”, and its relevance to the Central American migrants. Whereas the canonical 1951 Convention conception of refugeehood can be claimed only by those with a credible fear of persecution by their home state revisionists have argued that this arbitrarily excludes those -such as many of the Central American migrants- fleeing mortal peril owing to generalized disorder and insecurity, extreme poverty, or prolonged natural disasters such as those following from climate change. Having discussed the revisionists and their critics I then turn to a third position which I call “contingent revisionism” which asserts that claims to refugeehood should widen when states do not act on their duties to assist those in dire need.

I then discuss another ground for claims to asylum and resettlement: the notion that refugees should be thought of as “orphans of the state system”. The idea strengthens the force of states’ obligations to admit refugees though it is also apt to be misunderstood.

I turn finally to the issue of specific US responsibility for the Central American migration crisis and our reparative obligations. While some such obligations seem hard to deny, settling their scope and content raises complex and contestable questions. I conclude by advancing tentative suggestions about a way forward.

I hope in the end to have provided a useful overview of this complex terrain, highlighting some issues for further inquiry.

II. Justice among Members and Non-Members: the General Normative Context

The fullest mutual obligations of social justice hold among members of states because of the special relationship they share as members of collectively self-governing political communities. The institutions for which we are jointly responsible as citizens shape our interests and prospects comprehensively, from cradle to grave and beyond. Our government claims the authority to make final decisions on the most consequential matters. Legitimate states are answerable to their citizens in a special way, and they in turn bear special joint responsibility for what their government does in their name. The principles of domestic justice are designed to make the exercise of state power justifiable to the citizens over who it is exercised. When it comes to states’ obligations to secure fair equality of opportunity, for example, this is something that the Mexican government owes to Mexicans and the American government to Americans (^{Macedo, 2004, pp. 1721-38}).⁵

We stand in morally significant relations of various sorts to non-members as well. Our outward-looking relations with non-members are often significant but also different from those that face inward. Indeed, our obligations to outsiders while different in content may be in some instances more urgent. Among the most urgent claims that we may confront from non-members are those that advanced by people claiming to be refugees, as I explain below. And moral claims from non-members may even have priority over the claims of domestic justice when basic needs are at stake and claimants have good reason to press their claim against a particular state.

Our duties and obligations to outsiders fall into several categories, as mentioned in the introduction.

At the most general level, well off societies owe a duty of assistance to very needy or “burdened societies”, as Rawls calls them. Humanitarian assistance fulfills a universal moral duty of “mutual aid, which Walzer describes as a form of Good Samaritanism”: to assist the stranger fallen by the side of the road when we have more than we need and can relieve suffering at modest (or moderate) cost to ourselves. How much cost to ourselves? There is no precise answer. We should bear a “significant burden” (a reasonable or moderate cost) and help countries get to the point where they can provide decent lives for their citizens. An important feature of our universal humanitarian moral duties (or duties of beneficence) is that they would seem to have a target, which is helping societies get into a position where they can provide decent lives for their members. Once that threshold is reached (and again, it is admittedly vague), our duties to assist are at an end.

In addition to general humanitarian duties of aid on the part of the well-off toward those whose basic needs are unmet, we have a variety of moral duties and obligations arising from our relations with others who are outside our political community.

It seems right to say that we have duties of basic fairness or at least decency in our dealings with others, and these would seem especially important in the dealings of wealthy and powerful states in the dealings with poor and vulnerable ones. Duties of fair dealing with other persons and states include fair trade and also cooperation to solve common problems, such as climate change and other problems that are matters of serious global concern. And so too, we also have both shared interests and shared duties to promote institutions to facilitate and stabilize cooperation globally, especially on the most urgent issues, including those related to international political conflicts, economic and trade arrangements, and the global climate crisis. As others have argued, we should cooperate and institutionalize that cooperation, when we can do so at reasonable cost, so as to avoid the circumstances in which injustice will predictably prevail.

To fail to support institutions capable of regulating our relations with others on the basis of fundamental fairness, when such institutions are feasible and their cost reasonable, is to be complicit in allowing conditions to prevail in which the weak will be dominated and abused by the stronger.⁶

We also have reparative obligations arising from past and ongoing wrongful harms: unjust military interventions or our contributions to the climate crisis with uneven impacts globally.

Part of what we owe to non-members has been codified in the international human rights agreements which we have signed. We should respect and promote the human rights of all and do at least our fair share under our international commitments, and when we have entered publicly into international agreements, those duties become more specific obligations that we have undertaken and announced to the world.

Finally, I have come to agree with those who argue that the very existence of the global state system implicates us morally in the plight of refugees, and that is independent of our general humanitarian duties and our specific commitments under international law (discussed below). The existence of the state system, from which we benefit as members of decently functioning states, makes stateless people worse off in some respects by constraining their movements. I discuss this at greater length below

There is nothing original in this brief sketch of our outward looking duties and obligations. All of this is widely acknowledged by liberal democrats, social democrats, and others, and contested by those with more cosmopolitan sensibilities. The sort of global institutional architecture that we need in order to secure even a rough approximation to global justice is surely far more extensive than what we have. Nevertheless, once we do develop a more adequate global “basic structure”, it seems very likely that those supra-national institutions will remain much less comprehensive and less directly coercive than state institutions in relation to citizens (consider the European Union for example, in which states retain primacy) (Moravcsik, 2022, pp. 603 and 624). I emphasize all this simply to make clear my judgment that even with an urgent need to develop more robust global institutions to address the climate and migration crises, among others, we should not expect an end to the primacy of nation states, nor should we expect the dawning of an age of open borders.

III. Who is a Refugee?

Substantial controversy surrounds this question. It came up when a group of students and I met with the head of a refugee aid organization in Amman, Jordan, in January 2019. “That’s an interesting question”, he said, “if people are paying thousands of dollars to smugglers and they tell you their ultimate destination is Sweden, are they refugees?”.

What did he mean exactly?

There is little question that people fleeing famine and political instability are desperately needy. Yet desperate need is not the basis for classification as a refugee in international law. Article 33 of the 1951 Refugee Convention defines a refugee as someone who has crossed an international border because of a well-founded fear of persecution in one’s own country. At the core of refugeehood strictly conceived are persecution and “alienage”.⁷ “Alienage” refers to “a person who is outside the country of his nationality, or if he has no nationality, the country of his former habitual residence”.⁸ So far as persecution is concerned, the home state need not be the actual agent of persecution: refugee status could also be satisfied if officials simply stood by while private actors engage, for example, in genocide. Note too that not any form of persecution will satisfy the convention definition, which describes a refugee as a person who,

...owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership in a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such a fear, is unwilling to avail himself of the protection of that country.⁹

Those who are persecuted on one of these five protected grounds qualify under the “Convention conception” of refugeehood, as I shall call it.

Some countries have gone beyond the Convention conception and adopted wider definitions of refugeehood. Signatories to the Cartagena Declaration on Refugees of 1984 agreed to include “persons who have fled their country because their lives, safety or freedom have been threatened by generalized violence, foreign aggression, internal conflicts, massive violation of human rights or other circumstances which have seriously disturbed public order.” Signatories include all but one of the Central American and Caribbean countries and several Latin American countries.¹⁰ Similarly, the MERCOSUR countries of Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Peru, and Uruguay, also extend refugee protections to people on account of, “generalized violence, foreign aggression, internal conflicts, massive human rights violations, or other circumstances that disturb the public order”.¹¹ Colombia now provides refuge to over a million Venezuelan refugees.

People may apply for refugee status via state-based procedures, or they may apply to the offices of the United Nations High Commissioner for Refugees (UNHCR). Applicants are entitled under international law to a fair adjudication of their claims and, if judged to qualify, have a right to non-refoulement or non-return, as well as to refuge somewhere (not anywhere they wish).¹²

The UNHCR has the authority to adjudicate claimants’ status but not to resettle them in a new state, and this once again in no way guarantees that registrants will be resettled in the country of their choice. This path can lead to placement in a UN administered refugee camp, such as the overcrowded Moria camp on the island of Lesbos in Greece. Designed for 3,200, it grew to over 20,000 refugees who lived in crowded, dangerous, and generally appalling conditions until it was destroyed by fire in September 2020. Placement in refugee camps is very often anything but short term: people may be housed in a camp for as much as 20 years or more.¹³

Zaatari camp in Jordan has over 80,000 inhabitants, some of whom have been there since it opened in 2012. According to the UNHCR, as of April 2022 fewer than half have received even one dose of the Covid vaccine.¹⁴ The average stay in the Bidibidi camp in Uganda, one of the largest in the world, was around 10 years. “It was formally closed to new settlers, but the existing population is in it for the long haul -it’s becoming a city” (^{Strochlic, 2019}). The structures are becoming more permanent and foreign aid organizations are ceding local control to inhabitants where possible. Uganda is providing development assistance and sees the camp as a net positive (^{Betts, 2017}). With 270,000 residents, Bidibidi is a rare but far from insignificant example of what is possible when refugees are welcomed and integrated.¹⁵

IV. Shacknove’s Revisionism

In an influential 1985 article, Andrew Shacknove argued that the conception of refugeehood at the base of the 1951 international law convention is arbitrarily narrow owing to its historical origins in the mid 20th century European experience, especially Hitler’s genocidal policies against Jews, gypsies, and others. The moral and conceptual background of the conventional definition is that legitimate states are expected, at a bare minimum, to protect citizens’ basic rights and secure their fundamental interests in return for compliance with law. State persecution severs irrevocably this normal protective relationship between legitimate states and their citizens. Yet, as Shacknove pointed out, persecution does not exhaust the ways in which this “normal bond” between states and citizens can be severed.¹⁶

Shacknove proposed broadening the conventional conception of refugeehood in two ways. First, he argued that a well-founded fear of persecution based on one of the five protected grounds is a sufficient but not a necessary condition of refugeehood. Drawing on the conception offered by the Organization for African Unity, Shacknove proposed to include not only state persecution but also state collapse or weakness: people can be forced to flee not only from state ferocity or tyranny but also state frailty that issues in chaos and human misery. State frailty or failure may arise from natural disasters, or misgovernance, or some combination of these and other factors. The crucial thing, argued Shacknove, is that in these cases, as with state persecution, the normal protective bond between states and citizens is severed.

Shacknove’s second revision was to argue that alienage -which is to say, flight and arrival at the doorstep of another state- is not essential. People in great need may be internally displaced or suffering on the ground in their home territory, therefore, “Alienage should be considered one manifestation of a broader phenomenon: the access of the international community to persons deprived of their basic needs” (^{Shacknove, 1985, p. 283}).

So, the crucial criteria for refugeehood on Shacknove’s revisionist account are that the normal protective relationship between the state and its citizens is severed and that people’s basic needs are unprotected whether owing to state persecution or state failure, “tyranny or chaos”, and that they are available to international assistance: meaning that either they come to us, or we can go to them and assist them in protected areas in their own states (^{Shacknove, 1985, p. 283}).

To sum up, on Shacknove’s revised conception refugees’ status should be granted to those who meet three criteria:

Shacknove’s argument is relevant to the Central American migrant crisis, for it would appear that few of those families fleeing political instability, food shortages, or lack of opportunity are being persecuted by their state on one of the five protected grounds specified in the refugee convention (race, religion, nationality, membership in a particular social group or political opinion).

What about those fleeing gang-related violence? Determining the scope of claims under the convention is not always easy. “Membership in a particular social group” has proven the be a sort of “elastic clause” in the determinations of some immigration judges. For example, gangs prey upon abandoned street children by pressuring them into unwanted gang activity or, especially in the case of girls and young women, sexual activity. Matthew Lister points out that “particular social groups” deserving of protection may well include abandoned street children preyed upon by gangs, and also those with gang-related tattoos. Such claims may be especially likely to succeed when people have sought the protection of governments without success or have previously suffered from violence without obtaining government protection (^{Lister, 2008, p. 827}). The Convention conception is not without flexibility.

Should those fleeing persisting poverty, political instability, or generalized disorder nevertheless have a claim to refugeehood based on Shacknove’s wider conception? This is important because the legal status of refugee is a privileged position in international law. Those who qualify are not mere supplicants for aid but have rights claimable under international law: a right to asylum, a fair adjudication of their claims and, if granted, rights to non-return, and permanent resettlement.

One further status is relevant to the current migration crisis at the Mexican-US border. As an alternative to refugeehood, citizen-residents of some Central American countries and some other countries have been granted Temporary Protected Status (TPS) by the US. TPS status is granted to citizens formerly resident in a designated country who are in the US at the time of designation. This category of protection against return has been granted on the basis of “ongoing armed conflict (such as civil war), an environmental disaster (such as earthquake or hurricane), or an epidemic, other extraordinary and temporary conditions.”¹⁷ Fourteen countries have qualified for TPS status in the US, in the past or currently, with some being so designated first by Congress and others by the Secretary for Homeland Security. TPS country-by-country designations are for up to 18 months and once granted are renewable. To be eligible, a person must have resided in the US continuously since his or her state was designated as eligible for TPS. People from Honduras and Nicaragua resident in the US were first granted TPS status in 1999, and those from El Salvador in 2001. TPS status for those individuals has been renewed since. Newer arrivals from those states could only qualify for TPS if their country were “redesignated” allowing newer arrivals to apply during designated periods.¹⁸ The Trump administration sought to terminate TPS status for people from El Salvador, Haiti, Honduras, Nepal, Nicaragua and Sudan, but the move was enjoined by a federal judge. The Biden administration has reversed that policy and continued protections for those granted TPS status in the past.¹⁹

All rights granted under TPS, unlike those available to refugees, are temporary and revocable, even if in practice long-term. Few if any new arrivals qualify for TPS. We will see more about this later.

On the face of it, Shacknove’s proposed expansion of the category of refugee seems like an obvious moral improvement. People can be in urgent need of assistance on grounds other than state persecution: when fleeing starvation for example. Indeed, peasants fleeing drought can be more urgently in need than, for example, a businessman fleeing persecution.²⁰ Shacknove further points out that his revised and expanded conception “accounts more exactly for those persons who are in fact taxing asylum states”, thereby furthering the “erosion of minimum order” in receiving states like Lebanon, Jordan, and elsewhere (^{Shacknove, 1985, p. 277}); a designation of people fleeing persistent extreme poverty and disorder as refugees might help spread the burden. It seems plausible, moreover, to attribute the arbitrary narrowness of the 1951 Convention to its “Euro-centric” origins. It also seems telling that the basis for a broader conception would be found in Africa, a more likely source of insight into the needs of the global south.

Of course, a byproduct of Shacknove’s widened conception is that many more people will qualify for refugee status. There are currently over 53.2 million “internally displaced persons” in the world: counting them would nearly triple the number of refugees worldwide.²¹ Nearly 700 million people in the world live on less than $1.90 per day:²² how many of them might qualify as refugees? The resettlement system is already overwhelmed.

^{Shacknove (1985)} recognizes this problem, and characterizes prudential concerns about further straining wealthy countries’ limited generosity as the danger of “premature cosmopolitanism” (p. 281).

There are good reasons to try and gain a clearer understanding of the reasons for and against the widened conception. We need to sort competing claims based on sound priorities. A too hasty embrace of the wide conception may contribute to both system overload and backlash. It may even be that among some migration activists and migrants themselves, the attractions of the widened conception and hopes that it will be acted upon contribute to increased migration flows.

There is a powerful case for resisting Shacknove’s proposal with political implications for the status of Central American migrants.

V. Resisting the Revised and Expanded Definition

Shacknove’s expanded conception of refugeehood helps call attention to the moral urgency of assisting people who are not persecuted on one of the protected grounds but who are in great need. There remains, however, a central question: if people are not persecuted by their home state, why is asylum and resettlement in another country the appropriate form of assistance? ^{David Miller (2016, p. 80)} points out that Shacknove’s expanded conception says nothing about home state complicity in creating the dire need. A state that is incapable of securing its citizens basic needs may welcome assistance. If so, providing aid may be a preferable response, on several grounds.

One reason is cost-effectiveness. Refugee adjudication and resettlement are very expensive. The alternative “aid and development approach” of helping needy people in their home countries is, as Matthew Lister notes, “more likely to be able to help more people” (^{Lister, 2013, pp. 645-671}) and, in addition, with less of a risk of the sort of political backlash that often greets refugee resettlement.²³ Lister joins Thomas Pogge in citing yet a third reason to help poor people “in place”: those most likely to be able to find their way to wealthy countries to apply for refugee status are unlikely to be the “truly economically destitute”. The refugee agency head I mentioned at the outset also observed this.²⁴ That last consideration should be underscored: by encouraging the flight of the relatively able and well-resourced, we may be worsening the condition of those left behind. We may not only be depriving the sending country of needed skills and able bodies, we may be providing a safety valve that could lessen the domestic pressures and resources available to be brought to bear for social and political reform.

Let’s keep that last safety valve concern in mind.

Lister argues that so long as “we work to meet our obligations” through development assistance and aid, as a principled matter we “have no obligation as such to grant refugee status and asylum to those suffering from severe poverty”.²⁵ This is an important point: in the face of dire need abroad, states with duties to assist generally have a choice of means. Resettlement abroad is only one option, and we have now surveyed several reservations concerning resettlement.

Of course, we may worry that these reservations about resettlement will be all too appealing to public officials looking for excuses to turn away those seeking asylum even while having no intention -or no ability- to provide sufficient and effective aid. Let us keep in mind the worry that aid will be insufficient, for it will be important later.

The aforementioned considerations are important but we have not yet stated clearly the core of the principled case against Shacknove. Lister and Max Cherem pointed out that persecution by one’s home state is a unique justificatory basis (or nearly unique basis) for the specific remedies of the international refugee convention, which include the procedural right to fair adjudication of one’s claim and then, if successful, substantive entitlements to non-refoulement (or non-return) and permanent resettlement in another state where one’s basic rights will be protected. Persecution on group-based grounds, Cherem insists, uniquely entitles a person to the “very specific and durable remedy” of membership in a new state because it decisively severs the normal relationship between states and their citizens. This simply may not be true of people with “unfulfilled basic needs” (^{Cherem, 2016, pp. 184 and 185}).

In addition, the “protected grounds” specified by the 1951 convention (race, religion, nationality, political persuasion, or membership in a particular social group) make sense, even while they should not be seen as exhaustive. These are all deeply rooted and enduring aspects of our identities: features that, as Lister says, we cannot easily change and should not have to change. Persecution on the basis of these sorts of grounds is therefore “more serious or threatening” (^{Lister, 2013, p. 670}) and less possible to evade, than persecution on other grounds, such as, for example, owning particular assets that are coveted by high public officials. For these same reasons, we should extend the protected grounds to, for example, sexual orientation and gender identity as among the “particular social groups” covered by the refugee convention.It is by persecuting people on grounds that are deep-seated and immutable or fundamental, such that people cannot or should not be expected to change them, that the state irrevocably signals that it “has actively turned against them” and is their enemy. Then the state repudiates all concern for their well-being in a way that is deep and irrevocable (^{Lister, 2013, pp. 661 and 662}). Membership in a new state is the only possible remedy when the normal protective bond between states and their citizens is not merely severed but repudiated by states in this way.

In contrast, people who are merely in need, no matter how desperate the need, cannot claim that membership in a new state is their only resort. Their state lacks capacity, but it has not renounced concern for them as it does when it resorts to persecution on the immutable or deep-seated group-based grounds, so the possibility of their being assisted at home is not foreclosed, as it has been for bona fide refugees.

An additional consideration bolsters the importance of both persecution and alienage (or flight) for refugeehood and the duties to which it gives rise. Humanitarian intervention may seem to be an alternative means of helping those who are persecuted by their home states but who have not fled from it. However, intervening against a policy of state persecution is liable to require military means, and Lister rightly points out that such interventions very rarely pass the test of “proportionality”. Interventions often either require the intervenors to bear unacceptable costs, or they risk imposing greater and disproportionate harm on the “residents of the offending state”, as compared with the harm that might be prevented (^{Lister, 2013, pp. 663 and 664}). Such would seem to be the case with Western military intervention in Libya, and also the second US war in Iraq insofar as that 2002 invasion was partly, even if secondarily, justified by humanitarian concern for the domestic victims of Saddam Hussein’s dictatorship. In both cases, the costs of the intervention have been severe and long lasting. Thus, there is wisdom in the refugee convention’s emphasis on both persecution and alienage.

Whereas Shacknove seemed to claim that it was an advantage of his broader conception that it included more of those who are “taxing” the system of refugee resettlement in Turkey, Jordan, Lebanon and elsewhere, Lister and Cherem describe this as one of its central flaws. Given that the insufficient number of slots available for refugee resettlement, allowing the merely needy to qualify risks displacing those facing persecution who have no other remedy. When states feel overburdened by refugee claims, they often turn asylum seekers away, and direct them toward processing by the United Nations or elsewhere. While the UNHCR can adjudicate refugee claims it cannot, as we saw above, provide the remedy of permanent resettlement nor can it compel states to provide resettlement. The human costs of over-inclusiveness and the misallocation of scarce resources are seen in the hundreds of thousands of people languishing, often for decades, in UN-sponsored refugee camps. As Cherem puts it: “camp refugees can’t go home, yet every country rejects their appeal. They exist in a limbo between a repudiated membership and hopes for one that may never come” (^{Cherem, 2016, pp. 193 and 194}). When needy migrants fill slots allocated for refugees that is an unfairness to the refugees because membership in a new society is their only remedy, whereas the needs of “refugee-like” strangers in great need can be remedied by aid, or Temporary Protected Status, or in other ways. There is, Cherem asserts, a distinct unfairness when the “specialized tool” of refugee law is “used for something it wasn’t made for” (^{Cherem, 2016, pp. 191, 195 and 196}).

VI. “No Recourse” to One’s own State?

It turns out that, contrary to initial moral intuitions, there are significant weaknesses in Shacknove’s proposed broadened conception of refugeehood.

But there is one clear category of people in need who do not fit the Convention criteria of persecution and alienage who obviously do need the “specialized tool” of permanent resettlement because they can have no recourse to the protection of their own state. These are the climate refugees whose homelands have been degraded beyond repair or which face the near prospect of being submerged under the seas: Buxton calls these “total land loss” climate refugees. They also require membership in a new state, or the relocation of their political community elsewhere: for protection of their basic needs they have no recourse to their home states, at least as it is currently located (^{Buxton, 2019, pp. 193-219}).

A crucial consideration in the case of climate refugees is whether one has a realistic, reasonable, or adequate prospect that one’s own state will become able to secure one’s basic interests. If one does have that prospect -perhaps with outside assistance- to some adequate degree, then one may not have a good claim to permanent resettlement.

What does an adequate prospect of recourse to one’s own state for protection require? Luara Ferracioli has recently proposed a way of sorting and delimiting some of the relevant cases. She argues that many people might be said to have no decent prospect of securing some of their basic needs. These may include the low wage worker without adequate health insurance in the US who faced the prospect of falling below the poverty line as the consequence of an illness, and also the homeless (or unhoused) person in Sydney or Melbourne who is unable to access decent housing because of high immigration and rising prices (^{Ferracioli, 2022, p. 86}) In these cases a crucial consideration is that the channels of political protest and reform are not closed off to those individuals. Their political agency has not been altogether undermined, and so their membership in their society is not (as she says) meaningless.

Thus, in assessing whether potential claimants to refugee status have adequate recourse to their own state to eventually secure their currently unmet basic needs, important weight should be given to political agency: can basic rights be fought for with some reasonable, realistic, or decent possibility of success? If so, that should be the favored option.

Ferracioli cites two important considerations in favor of that conclusion. One is that, “human rights protection is not «manna from heaven»… [it] is always an ongoing political project on the part of citizens”. Human rights protections and decent governance more broadly often need to be fought for, and we should encourage those with a modicum of agency to be and remain engaged politically. Indeed, we should also worry about the out-migration of those with the capacity and resources for agency from societies that stand in need of repair: the exit of the relatively young, energetic, and resourced can worsen the prospects of those left behind. This addresses the safety valve worry I mentioned above.

It is not unreasonable for the international community to incentivize members of struggling states to exercise their agency in favor of reform at home: in Albert O. Hirschman’s famous formulation, to exercise voice and loyalty to the other members of their community rather than exit (^{Buxton, 2019, pp. 193-219}).

Ferracioli also mentions additional categories of persons who might not be especially good candidates for exercising agency. She mentions pervasively misogynistic societies, such as Afghanistan, with high rates of gender violence and in which women are very often denied freedom of choice in politics, marriage, and family life. Such women, she suggests, ought to be regarded as having no realistic prospects for political agency and therefore as candidates for successful refugee claims (^{Buxton, 2019, pp. 193-219}).

Exactly when one has “no recourse” to one’s own state for protection, or “no adequate prospect of success” in reforming one’s own society in concert with others, are difficult questions to answer. There are obvious dangers here of instrumentalizing people or treating them as resources for political reform rather than autonomous agents with their own lives to lead. Finally, many of those who flee from their home countries may already have profound feelings of remorse and guilt, even if unwarranted. So, I raise these issues to flag the need for further inquiry, not to draw any conclusions.²⁶

VII. A Third Conception: Need-Based Refugeehood as a Contingent Claim

The debate we have just surveyed is relevant to the status of many of the Central American migrants, including those whose plight is a consequence not of persecution but of generalized violence associated with drug cartels, gangs, generalized political incapacity and corruption, and natural disasters including some that are worsened by climate change. And then too we have the issue of specific responsibility on the part of the US, which I address below.

Under Shacknove’s broadened conception many of the Central American migrants would qualify for refugee status who may not under the narrower Convention conception. In light of the preceding discussion, it could also be interesting to consider whether those in flight have adequate prospects for political agency, but I cannot pursue that question here.

Matthew Gibney suggests a possible way of bridging some of the gap between Shacknove and his critics, and in a way relevant to the Central American crisis.

Gibney disagrees with Lister on the adequacy of offering “temporary protection” to those fleeing persisting “random or generalized violence”. Surely the crucial thing, he says, is that like the persecuted, many equally face “death if they stay where they are” (^{Gibney, 2018, pp. 1-9-30}). On that basis -facing mortal peril- they are equally entitled to protection.

Yet there is the now familiar rejoinder: those fleeing generalized violence, unlike the persecuted, have not been repudiated by their governments. Corrupt or incompetent officials may be failing to protect them, but other forms of assistance may be possible, and indeed may be welcomed by the remaining uncorrupt officials. Temporary protection accompanied by assistance to frail governments could be adequate even in these cases of people fleeing mortal peril.

In response Gibney suggests that refugees are persons whose basic “human rights cannot be protected except by moving across a border, whether the reason is state persecution, state incapacity, or prolonged natural disasters” that the state cannot cope with. Yet he adds, taking account of the possibility of aid in place: “People who can be helped in their own state -such as victims of natural disasters and poverty- have no right to asylum if states are willing and able to help them where they are. If protection is not forthcoming, however, individuals may legitimately claim to be refugees moving to protect their [most basic] human rights” (^{Gibney, 2018, pp. 1-9-30}).²⁷

This proposal accepts provisionally that aid can be a substitute for the flight to refugee status, and then conditions the granting of status on whether the states who can assist actually step up and provide effective assistance. While victims of generalized violence or mayhem, or natural disasters, can in theory be helped in place, the crucial thing on this view is whether they are in fact helped and their plight relieved.

The resulting account of refugeehood has the advantage of building in an incentive for states to give aid, and it thereby addresses a worry I mentioned above: that states would eagerly embrace the narrow conception of refugeehood (whose premise is that people who are needy but not persecuted can be helped via aid) but then not provide the aid. On this “contingent conception” of refugeehood, if effective aid is not forthcoming in the face of persistently unmet basic needs -or enduring mortal peril- then people have the right to refugee status.

The upshot is that states with the capacity to help those in dire need can preclude claims to refugee status by sending effective aid. States thereby acquire an incentive to provide aid as an alternative remedy for those fleeing general insecurity, state collapse, or a natural disaster.

Neat as this proposal may be in theory, its implementation in practice requires that states actually buy into it: commit to it and act on it. Attractive as it may be, it has no teeth.

In addition, it must be noted that government-to-government foreign aid has a lousy track record (^{Easterly, 2006}). Nevertheless, there may be things that wealthy states like the US can do to help Central Americans in place, I will return to that issue in the end.

VIII. Refugees as Orphans of the State System?

Near the outset I provided a brief general account of the grounds for states’ moral duties and obligations to people in great need. To recap, the first was the familiar general humanitarian “duty of rescue” or what Walzer calls “mutual aid,” which we owe to the needy stranger, per the Good Samaritan parable: we who have much should offer assistance to those in great need when feasible without serious risk or great cost to ourselves. This is a very broad but also a fairly weak principle in that it can be claimed by anyone in great need, but the duty it implies against any particular state holds only up to a point. Once we have borne a sufficient burden, and assuming that we have no substantial relation to the needy strangers’ plight, we can say “no more”. We are morally required to bear some cost but not great cost.

We also have the more specific and focused obligations arising out of our signing the Refugee Convention. Those obligations are more focused because the convention specifies the relevant categories of persons who are eligible, and tells those people what they have a right to expect when they present themselves to a state signatory. By signing we have announced and undertaken that we will provide those asserting claims under the convention with a fair hearing and, if successful, asylum and resettlement. We are now obligated to adjudicate the claims of those who present themselves to us with a credible case.

Before we get to the issue of specific obligations of repair to Central Americans arising out of past US actions, I want to consider one other ground of states’ responsibility for refugees independent of the two just mentioned.

Refugees exist in part as a consequence of global system of sovereign states. They are in effect orphans of the state system, as Joseph Carens first argued: they lack “effective membership in one state with no positive entitlement to join another” (^{Carens, 1991, pp. 18-29}) This, says Gibney quoting ^{David Owen (2020)} (who has developed the position at length), “poses a normative problem for the legitimacy of the state system, particularly if… we conceptualize «the international order of states» as a «global regime of governance»”.²⁸If this is correct then the duty to “respond to refugees” and provide them with protection, is not merely a general humanitarian duty holding among strangers but rather, “part of the «political obligations» of states generated by the common system of governance that they uphold”.²⁹ Gibney observes that, “The legitimation of this mode of governance requires a collective obligation on the part of states to incorporate refugees back into the system, if necessary through the provision of asylum”.³⁰ These observations seem to me correct: they have some force, though they can also be overstated as it seems to me that Emma Haddad does:

This book maintains… that the existence of modern political borders will ensure the constant recreation of refugees. Accordingly, it regards the refugee as a contemporary concept that was made a permanent feature of the international landscape with the consolidation of the modern system of nation states. (^{2008, p. 3})

It goes too far to say that “the existence of modern political borders will ensure” the constant recreation of refugees. Refugees are not an inevitable result of the state system any more than orphans are the inevitable result of the family system, at least if by “inevitable” we mean necessarily created or presupposed by the system. The state system no more requires refugees than the family system requires orphans. Refugees are created by failures or abuses of governmental power in particular places. Those failures are (again) not an inevitable result of the state system, nor are they presupposed by it. A world of states without refugees is perfectly conceivable, and what we should be aiming for.

Nevertheless, it is true that stateless persons are left out of the state system and unprotected by it. As compared with the state of nature in which everyone is in the same condition, today’s stateless persons are in some ways worse off, because more constrained as a consequence of the state system than they would be in the state of nature Stateless persons cannot range freely over the earth as they could have in the state of nature. Many are denied the option of moving and resettling in a new state by states who are not taking anything like their fair and proper share of refugees, and also not acting as they should to repair the conditions that generate refugees. They also have the prospect of being better off if they can secure membership in a legitimate and effective state, which is much better than the state of nature.

We who are members of decent and legitimate states benefit greatly from the existence of the global state system, and to enforce its terms we exercise power over members in our political community and non-members, including stateless persons. We must justify the global state system to all who are directly affected by it, including those who are left out of it, and especially those over whom we exercise power. We owe everyone on earth a justification for the state system in part because other systems are conceivable: one global state, or a regime of free movement across states. So Haddad is correct to say that refugees “are the result of erecting boundaries, attempting to assign all individuals to a territory within such boundaries, and then failing to ensure universal representation and protection”.³¹ So yes, refugees are owed incorporation into the global system of states partly because so many others benefit from the state system that leaves them out, and because that system makes them in some respects worse off -more constrained- than they would be in a global state of nature, and finally because we exercise power over them in enforcing the terms of entry into and membership in our state.

What do these observations add to our discussion?

Insofar as the claims to of stateless persons arise from the general circumstance that the world is now organized by states with borders, their claims are against states in general, and not specifically against any particular state. Their claims are in that respect diffuse rather than focused. Yet they are not merely claims for humanitarian assistance: the global state system implicates us and our states in their plight, and an adequate response to their claims is a precondition of the legitimacy of the global state system. This makes our moral duties to refugees more stringent. There are still limits to any particular state’s obligations -a fair share or reasonable cost proviso- but I think we can go further.³² Because these obligations arise out of the imposition of the state system, and thereby implicate us, we have a stringent obligation to create and maintain a systemic response, that is, effective coordinating institutions (^{Montero, 2022}). Claims arising from the relation we share as subjects of a global state system may be addressed most pointedly to those states that are not rendering adequate assistance to the needy and not supporting and promoting improved global governance for all.

I have so far discussed the generalized duties and more specific obligations to assist those in need arising from humanitarianism, from international legal obligations, and from the existence of the state system. One topic remains: the more specific obligations of repair on the part of states arising out of responsibility for the conditions causing people to flee their home states and seek asylum and refuge. I turn to these next.

IX. Reparative Obligations, the US, and the Central American Migrants

First a brief word on recent migration trends to the US.

The combined decline of labor demand associated with the Great Recession of 2008-9, and the increased difficulty in crossing the border have meant that, since around 2012, “net flow of migrants from Mexico to the United States have essentially stopped”.³³ The reasons for the cessation of Mexican migration to the US are several, and include the decline of economic opportunities in key industries like construction in the USA; increased border security; improvements in economic opportunities in Mexico; decline in fertility in Mexico and with it a somewhat lower dependency ratio.³⁴

Over the last decade, with lower migration overall across the Southern border, a much larger portion of those seeing to enter the US have been families from the so-called Northern Triangle states of Honduras, El Salvador, and Guatemala. Since 2014, more than 2 million people are estimated to have left these three countries, the large majority of whom (80-90%) have sought entry to the US.³⁵ Flows have varied from year to year. In 2014, approximately 239,229 people from the Northern Triangle were apprehended at the U.S.-Mexico border; in 2019 apprehensions at the border almost tripled to 609,775 people.³⁶ Whereas an estimated 691,000 people left the region in FY2019, that dropped to 112,000 in the Covid year of FY2020, before rising again subsequently.³⁷ Specific obligations of repair -or reparative obligations- depend on a given state’s past or ongoing involvement and responsibility for the circumstances that give rise to refugee situations or forced migration of various sorts. Those circumstances include relatively weak state capacities, corruption, instability, generalized disorder, the power of criminal organization, and climate change. Responsibility for harms to governance capacity in any of these forms (or others), may give rise to reparative obligations to rectify the harms, at least where the policy involves wrongful or illegitimate harms.

Reparative obligations are distinctive in their force and extent as compared with generalized humanitarian duties and claims arising from the existence of the state system.

How so?

Whereas humanitarian duties are in principle claimable against anyone, but they are also diffuse -claimable against no one in particular- excepting perhaps those who are close enough to be accessible to would-be asylum seekers, or otherwise specially situated to help. Proximate countries who bear no special responsibility can fall back on limitations concerning the costs they must bear.

Claims grounded in the “orphans of the state system” argument are also claimable against everyone, but they have greater stringency than general humanitarian duties because those who benefit from the state system are implicated in the plight of those who are excluded and hemmed in by borders. The claims of those left out of the state system are diffuse, in that they are not specifically directed at particular state, but these claims strengthen the reasons for states to sign on to refugee resettlement arrangements and to make good on their commitments.

In contrast to the previous, reparative obligations are claimable against specific responsible agents -they are focused rather than diffuse- and in addition the “low cost” proviso does not apply: the obligations depend on the extent of the harm caused, the debts may be substantial. While reparative obligations may not require satisfaction in the specific form of asylum and the granting of refugee status, they may provide an especially powerful case for activating the contingent revisionist account described above. That is, if a responsible country does not make good on its reparative duties via aid and effective development assistance, it may have an especially powerful default obligation to accept migrants, at least assuming that this provides general benefits to the political community to which duties are owed.

The US and other Western states such as Australia have sometimes taken their reparative obligations seriously, including in response to Western military intervention in Vietnam, Cambodia, and Laos in the 1960’s and 70’s. After the fall of Saigon in 1975, the US Congress and the Ford administration authorized 140,000 refugee admissions. It soon became clear that the refugee crisis would be far greater. Eventually, 2.5 to 3 million people fled these war-torn countries, and it is estimated that as many as 50,000 drowned at sea. From 1979-1999 the United States accepted an additional 500,000 refugees from Vietnam.³⁸ Those numbers contributed to the US leading the world in refugee admissions. Given the horrors inflicted by the US military the response may still be judged inadequate.

We have been far less responsive in the case of Iraq in the aftermath of our 2002 invasion.

Obligations of repair for past and ongoing wrongful harm may then appear then as a promising basis for justifying claims to entry on the part Central American migrants. A practical problem arises however. In order to identify and specify reparative obligations it is, as Gibney says, “first necessary to explore the specific and sometimes complex causal relationships that exist between specific states, structures of power, and particular groups of refugees” (^{Gibney, 2018, p. 54}). In many cases the relevant causal relationships and questions of culpability, will be multiple, complex, and subject to reasonable disagreement.

Enrique Camacho-Beltran provides an impressive bill of particulars with respect to US involvement in the Central American region, and in El Salvador in particular. Early in the 20th century, Teddy Roosevelt asserted the right to exercise an “international police power” in Latin America, which involved violations of national self-determination and various forms of manipulation to promote US interests in the region. American and British nationals “owned most of the coffee plantations and railways which had detrimental effects on the local economy”. When indigenous farmers rose up to support a guerrilla insurgency led by Farabundo Martí in 1932, “the US sent naval support to contain the peasant rebellion and support dictator Maximiliano Hernández Martínes”. The US intervened several more times to support Martinez and other corrupt dictators. In 1981, “the US-trained Atlacatl battalion was involved in the El Mozote massacre where almost a thousand unarmed civilians (women and children included) perished. During the 80s, an estimated number of 80,000 people died in this US-fuelled war”. The CIA provided millions to help influence the 1984 election. On the economic side, to mention just one factor, the Central American Free Trade Agreement increased US “influence over domestic trade and regulatory protections”, over the protests of “unionists, farmers, and informal economy workers” (^{Camacho-Beltrán, 2022}).

So, in the case of Central America, and specifically that of El Salvador, it seems clear that US past and ongoing involvement in encouraging democratic overthrows, looking the other way at authoritarian regimes’ human rights violations, enactment of arguably unfair trade agreements, and refusal to curb the abusive powers of US corporations, add up to a substantial debt, based on wrongful harms, even if the precise magnitude of the debt is hard to specify.

Another factor influencing poverty in Central America for which US bears part of the blame is climate change. Yet here too there are complexities and uncertainties. US contributions to climate change are part of a larger pool to which many have contributed. How should responsibilities be apportioned? Rebecca ^{Buxton (2019}) considers three classes of agents liable for shares of reparative obligations for climate change: the polluters, the beneficiaries (that is, consumers or consuming nations), and the able. All three categories seem relevant, with qualifications. Some of the polluters may be very poor, as in India, and poverty may be at least a partial excuse for polluting. On the other hand, insofar as the polluters are also beneficiaries and rich -and therefore able- the case for obligations of repair seems strongest.

One general conclusion would seem inescapable: the United States is obliged to do more than it does -much more- both with respect to climate mitigation and climate-induced migration.

These are yet additional relevant factors driving Central American migration. Migration is being driven by generalized disorder, gang violence, and “feminicide,” much of which seems related to illegal activity associated with drug cartels. The drug cartels are enriched and empowered as a side effect of the US war on drugs, which turns drug production and sale into a highly lucrative criminal activity. Most of the illegal guns seized in Mexico are also US made.

But US anti-drug laws, even if unwise, may nevertheless be legitimate policies. The vast majority of governments in the world make the possession and use of drugs like cocaine illegal. Fewer than 10 countries have legalized the possession of small amounts of cocaine for recreational purposes (the US and a few others have legalized it for medical uses), and even they generally prohibit the cultivation, transport, and sale of cocaine. US policies concerning cocaine are not unusual.

The costs imposed on Central American states and Mexico as a consequence of our drug laws would seem to be considerable. The US should consider the externally imposed costs when weighing the overall justifiability of such laws, if we do not that would seem to be a moral failing. But is it clear that our drug policies are illegitimate? I am not sure. Mexico has its own war on drugs, notwithstanding that in 2019 a judge in Mexico ruled that two people had a right to use (but not sell) recreational cocaine.³⁹

The question of responsibility for the plight of the Central American migrants is complex. Some argue that poverty, corruption, and economic underdevelopment in Central America owe not only to the malign influence of outside actors but also to choices and factors that are endemic to the region and its culture, including endemic local corruption and high rates of feminicide.

Luis L. Schenoni and Scott Mainwaring, in an article cited by Camacho-Beltran, provide an empirical examination of hegemonic influences over democratization and democratic breakdown in Latin America after 1945. Schenoni and Mainwaring argue that: “While past work has attributed responsibility to the US for the waves of democratic breakdowns from 1948 to 1956 and 1964 to 1976, an examination of the 27 breakdowns from 1945 to 2010 gives reason to doubt this interpretation” (^{Schenoni & Mainwaring, 2018}).

Indeed, contrary to Camacho-Beltran, Schenoni and Mainwaring provide partial exoneration for the US role. Their research suggests that the US may have played a role in certain breakdowns when it sensed a communist threat, and the US at least weakly supported 12 of 27 authoritarian coups, but they argue that US support was by no means a sufficient condition for the success of such coups. Even when there was support, US embassies’ messaging on the ground always opposed authoritarianism. Furthermore, given that the US opposed many coups, the article takes a largely neutral stance, arguing that while US support may have been helpful in certain cases, the US was not responsible for triggering or ensuring the success of either wave of democratic breakdowns: “The waves of authoritarianism that rocked Latin America from 1948 to 1956 and from 1964 to 1976 were not a result of consistent US indifference toward democracy or support for authoritarianism” (^{Schenoni & Mainwaring, 2018, p. 272}).

In another assessment of US responsibility for mass migration out of the Northern Triangle states, Michael Shifter and Bruno Binetti survey and credit many of the same factors cited by Camacho-Beltran: CIA involvement in coups, US drug policy, trade and economic policies that concentrate gains at the top. While they also attribute blame to corrupt and self-serving elites in these countries,⁴⁰ that fact does not obviate US culpability: responsibility is not “zero sum”.⁴¹

I discuss these essays not to debunk Camacho-Beltran’s overall argument, but to suggest that the project of assessing, gauging, and assigning responsibility is bound to be complex and contested. Likely the US does far less than it should to address the root causes of migration from Latin America, but it is not easy to say what should be done. I would join those who condemned the Trump administration for combining highly punitive border security measures with efforts to terminate $500 million in aid to Central American states in retaliation for their failing to stem the flow of migrants. The Biden administration is making an effort to address the region’s problems by engaging regional governments and especially grassroots and faith-based civil society organizations in a “Root Causes Initiative” to improve economic development and governance.⁴² Deliberative engagement with Central American states and societal representatives seems like a far more promising and humane path, and more consistent with redressing past harms. But whether these efforts will bear fruit remains to be seen.

X. Conclusion: From Moral Reflection to Action?

I have spent most of this essay surveying normative debates around the general question of who should qualify for refugee status. These principled debates are relevant to the question of who among those fleeing from the Northern Triangle states might qualify for refugee status. Some undoubtedly do qualify, and many others should be admitted on the basis of policies required to satisfy our humanitarian duties, obligations arising from the existence of the global state system, and duties of repair for past US state policies.

On the other hand, insofar as the US accords Temporary Protected Status to those fleeing natural catastrophes and generalized disorder -that is, to those facing mortal perils that may be temporary, and who have not been renounced by their own states- this seems not unreasonable. The problem with TPS is that many people have been in this “temporary status” since the 1990’s. Many are now more or less fully integrated into US society, have had children and families here. For long-term TPS recipients, as with the long-term undocumented, there should be a path to full US citizenship. Another problem is that TPS has been a narrow window open to a select few only.

More pressing than the problems of refining our moral and legal assessments is the challenge of getting countries, especially powerful countries like the US, to act.

Ideally we should apportion responsibility fairly across global contributors to the problem and devise enforcement mechanisms. But all such international efforts are plagued by contestation, uneven compliance, and shirking on commitments. These are especially nettlesome problems in the absence of a trusted 3rd party with authority to decide and the power to enforce.⁴³ Powerful countries like the US should support the institutionalization of such authorities in the form of multilateral organizations with teeth, as we have done in the case of the World Trade Organization, the international protocol on ozone-depleting gasses, and on some other matters. The problem is that whereas elites in powerful states came to perceive an interest in lowering trade barriers and protecting the ozone layer, this is not yet the case with respect to refugee flows and climate change, among other global problems. The people of the entire globe, and not only the Central American migrants, are held hostage to the deeply dysfunctional politics of the US, which must therefore be added to the list of things that threaten humanity.