¶¶Òõ¶ÌÊÓƵ

Will international refugee law still be relevant in 2033?

The implications of future megatrends for the protection ofÌýpeople forced from their homes

Jane McAdam AO

Keynote address

Kaldor Centre Conference, 20 November 2023

Ìý

I have been asked to reflect on whether international refugee law will still be relevant in a decade’s time. Questions about the future of international refugee law have been asked almost since its inception. The first universal refugee treaty, the 1951 Convention relating to the Status of Refugees, originally limited protection to refugees who fled ‘as a result of events occurring before 1 January 1951’,[i] although the drafters did express the hope that the treaty would ‘have value as an example exceeding its contractual scope’[ii] and be applied to other refugees as well. For the first five decades of its existence, UNHCR’s mandate was limited to five-year terms, which were renewed successively until the General Assembly decided, in 2003, to make the agency’s mandate permanent, or at least ‘until the refugee problem is solved’.[iii]

Ìý

Despite political attacks, heightened border controls and increasing numbers of people on the move, the international refugee regime, both legal and institutional, has proven to be resilient. And there is every reason to believe that it will remain so. This morning, Aarathi queried whether our 20^th century frameworks can tackle 21^st century challenges. Well, the foundational principles of international refugee law are timeless, stemming as they do from centuries-old notions of sanctuary and shelter.[iv] Especially in light of the predictions we have heard today about future drivers and trends of displacement, their importance will only increase in the coming years, not diminish.

Ìý

But here is the rub: while international refugee law will remain relevant, the question is whether it will be able to do the work it was cut out to do. Will people be able to access territory to claim asylum? If they can, will they have access to legal status determination procedures? Will such procedures be sufficiently rigorous and nuanced to capture the complexity of movement and the compounding nature of risk? Will decision-makers – especially if they are not human – be able to understand how multiple drivers of movement intersect and coalesce, and how they relate to core protective principles? Will the public be open to accepting those in need of protection? Will they be sufficiently well informed about why we should? It is here that our vision becomes less than 20/20.ÌýÌý

Ìý

If we look to history, we can see that refugee law has been remarkably robust. Although drafted in the 1950s, the Refugee Convention has proven to be adaptable to changing contexts and capable of a dynamic interpretation. For example, the refugee definition is now understood as being able to encompass gender-based persecution, persecution linked to sexual orientation and identity, as well as persecution in the context of disasters – circumstances that its drafters would never have envisaged.

Ìý

While international refugee law on its own cannot resolve the displacement we see today or in the future, it does offer a principled, ordered framework for protection which can ‘serve both as the essential premise for international involvement, and as the measure of accountability for the assessment of particular actions or policies.’[v] Still, as Guy Goodwin-Gill has noted, ‘the law alone does not provide solutions, and much remains to be done’.[vi]

Ìý

One question is what future exists for international refugee law, as opposed to national iterations of it.[vii] Rather than using the Refugee Convention as a blueprint to guide positive action, some States look for the grey areas to confine and restrict their obligations as far as is arguably possible. Australia has arguably led the way on this by creating a ‘self-contained statutory framework’ largely devoid of references to the Refugee Convention. In 2014, then Immigration Minister Scott Morrison stated that: ‘This parliament should decide what our obligations are under these conventions—not those who seek to direct us otherwise from places outside this country.’[viii]

Ìý

There is a further question about how far refugee law will take us when there are many others also in need of protection. The development of the principle of non-refoulement in human rights law is an obvious complement, extending protection to people who face a real risk to their life, or who face the prospect of torture or other cruel, inhuman or degrading treatment or punishment. This has particular relevance in the context of climate change and disasters, with international bodies already recognizing that ‘without robust national and international efforts’, ‘the effects of climate change … may expose individuals to a violation of the[ir] rights … thereby triggering the non-refoulement obligations of sending states’.[ix] Given the unlikelihood of States adopting new binding instruments to protect those on the move, the precise scope and contours of the non-refoulement principle will become ever more important.

Ìý

Indeed, one of the most important developments over the past decade has been our growing understanding about how climate change intersects with other drivers of displacement, such that it amplifies existing risks and vulnerabilities. Some people fleeing the impacts of disasters and climate change will be refugees, because a disaster makes a risk of persecution more acute, or exacerbates ‘existing patterns of discrimination and marginalisation’.[x] Others will not satisfy all the elements of the refugee definition, but may still be protected by human rights law because their life is at risk or they face such dire circumstances upon return. We go into these issues more in the Kaldor Centre Principles on Climate Mobility, launched last week.

Ìý

Most people who do move because of the impacts of climate change and disasters will never cross an international border but will stay within their own country. Indeed, over the coming decades, many more people may be trapped and unable to move at all. Whereas historically, we have tended to view displacement as a sign of vulnerability, in some cases, immobility may be an even stronger indicator – particularly if people’s predicament remains unseen by policymakers. How the world deals with those who are stuck may present an even greater humanitarian challenge than assisting those on the move.[xi]

Ìý

The vast majority of the world’s population cannot lawfully cross a border: most people do not hold passports, which also means visas are inaccessible, and some passports enable far greater access than others, entrenching a global mobility divide.[xii] Projections about the use of artificial intelligence (AI) and other technologies may embed this even further. The Janus-faced nature of biometrics and predictive analytics means they can be harnessed in all sorts of ways, some positive, some less so. AI can assist with the provision of humanitarian assistance, but may also subvert free and informed consent: for instance, refugees in Jordan who had to have their irises scanned in order to receive weekly rations felt they had no alternative but to comply.[xiii] As Louis said this morning, certain biases inherent in algorithmic profiling means difference is rendered as threat, while at the same time, machine-assisted decision-making may be able to provide more nuanced approaches to uncertainty than humans can. Smartphones and social media can help those on the move access and share vital information, receive assistance, stay in touch with family, and be powerful global advocates, but these innovations can also make them more vulnerable to government surveillance and detection. To what extent will access to territory affect access to solutions?

Ìý

As my colleague, Fleur Johns, has written:

Ìý

It may be that automation allows international institutions to anticipate and temper violence and suffering like never before. Yet the very same changes may threaten to undermine international law’s and institutions’ operations, or narrow their effective jurisdiction, insofar as they heighten inequalities, magnify distrust, and impede communication and connection.[xiv]

Ìý

As Niamh mentioned earlier this afternoon, we already have the technological capacity to forecast risks, so how do we ensure that it is used to anticipate and protect, rather than intercept and deflect?

Ìý

For those who can access refugee status determination procedures, automated decision-making poses further challenges. ‘Even the best analytical models can only be as good as the historical data that they are trained on’.[xv] There is a risk that machines will take too blunt an approach and deny protection to people in need, undermining both human rights and due process. In part, this goes to system design, which depends on human input and the values underpinning it.[xvi] As Shayar said earlier, perhaps a better focus would be on assistive technology as a tool for human decision makers.

Ìý

In the asylum context, beyond the obvious concerns about wrongful denial of protection claims, there is a related issue about the ongoing development of refugee law: if decisions about people’s legal status are made without full, human deliberation, the capacity to develop more nuanced understandings of key principles and concepts may be lost – and with it, the notion of refugee law as a living, dynamic body of practice. For instance, our initial understandings of the impacts of climate change on displacement were not nearly as fully formed as they are now. Had we programmed a machine early on, we may well have screened out deserving claims because of our own human fallibility in properly understanding the issues.

Ìý

And yet, on the flipside, machine learning is now enabling major advances to our understandings. Next-generation approaches to modelling climate mobility at a global scale use ‘advanced statistical methods designed to accommodate the complex relationship between climatic factors and human mobility’.[xvii] While the results are not yet at a point where they can reliably inform future policy,[xviii] some interesting findings are emerging. For instance, large-scale studies to date reveal that in impoverished communities, ‘adverse climatic changes tend to affect migration only weakly or even negatively, reflecting resource constraints that compromise the ability to move.’ [xix] This is contrary to the common assumption that those in the most difficult circumstances will be the first to leave.

Ìý

All these considerations come at a time when communities all over the world are feeling the pressures of shrinking resources, rising costs, worsening disasters and the existential threat of climate change. Even those in relatively privileged circumstances may not feel particularly fortunate. The plight of others further afield may not be in plain sight – and if it does come into view as people seek protection elsewhere, then – as our last panel explored – the reception that awaits them remains unclear. The combination of the global megatrends we have discussed today could either engender a new humanitarianism and communitarianism, or instead lead to hardened nationalism and xenophobia.

Ìý

Reflecting on the panel just now, how do we create a sense of connection, rather than disconnection, of shared humanity rather than difference?

Ìý

Notably, not a single State has ever withdrawn from the Refugee Convention, despite perennial protestations about its ongoing utility and relevance. Even the most jingoistic politician must realize that without it, there would be greater disorder in addressing global displacement. As Volker Türk, now the UN High Commissioner for Human Rights, once noted, ‘departures from the fundamental principles of international refugee protection have neither reduced nor stalled refugee movements’, but have resulted in ‘ineffective management of large-scale influxes, the diversion of refugee movements, [and] the creation of tensions between states as burdens and costs are shifted from some onto others’.[xx]

Ìý

Conclusion

Ìý

To foretell the future accurately requires a crystal ball, but to foresee it in general terms requires at least a solid understanding of the past. As Magdalena said this morning, we cannot deal with the future unless we acknowledge what has gone before. It is in taking this long view of international refugee law that I feel confident in saying there is no doubt that international refugee law will still be relevant in 10 years’ time. Whether it will remain sufficiently dynamic, living and nimble in its response to evolving circumstances will depend on scholars, decision makers, policymakers and lawyers constantly demonstrating its adaptability and applicability in the same way as has occurred over the past century, as well as displaced people themselves explaining how a range of factors combine to force them to move. There are certain patterns and trends that are likely to continue, but there will also be some disruption. Anticipating and planning for this is key. And of course, refugee law alone is not, and never has been, the answer. It is one element – albeit a very important one – in a complex of protective principles and other laws, policies and practices concerning mobility.

Ìý

International refugee law will not look identical in 2033 as in 2023 (and I say that with some trepidation, knowing that we have a 5^th edition of The Refugee in International Law to write in this coming decade). However, the fundamental principles on which it is based will remain – and some may have been extended in yet unanticipated ways, especially as we continue ‘to scrutinize the outer limits of the principle of non-refoulement when it comes to protecting those at risk’.[xxi]

Ìý

My hope is that 10 years from now, I will safely be able to misquote Mark Twain to assert that reports of the death of international refugee law were greatly exaggerated.

Ìý

Endnotes:

[i] Convention relation to the Status of Refugees (adopted 28 July 1951, entered into force 22 April 1954) 189 UNTS 137, art 1A(2).

[ii] Final Act of the United Nations Conference of Plenipotentiaries on the Status of Refugees and Stateless Persons (1951), Recommendation E.

[iii] UNGA res 58/153 (22 December 2003) para 9.

[iv] See eg Norman M Trenholme, The Right of Sanctuary in England: A Study of Institutional History (University of Missouri 1903); Peter Gatrell, The Making of the Modern Refugee (OUP *) *.

[v] Guy S Goodwin-Gill, ‘The International Protection of Refugees and Asylum Seekers: Between Principle and Pragmatism?’ (Keynote Address, Andrew & Renata Kaldor Centre for International Refugee Law Conference, UNSW Sydney, 3 November 2014); see also Guy S Goodwin-Gill, ‘Editorial’ (1993) 5 International Journal of Refugee Law 1, 5–6.

[vi] Goodwin-Gill Ìý41.

[vii] See eg Stephen Meili, ‘The Constitutional Right to Asylum: The Wave of the Future in International Refugee Law?’ (2018) 41 Fordham International Law Journal 383.

[viii] Morrison, Hansard (25 September 2014) 7.

[ix] Teitiota, para 9.11.

[x] Matthew Scott, ‘Finding Agency in Adversity: Applying the Refugee Convention in the Context of Disasters and Climate Change’ (2016) 35(4) Refugee Survey Quarterly 26, 28.

[xi] Government Office for Science (UK), Foresight: Migration and Global Environmental Change: Future Challenges and Opportunities (2011) 25.

[xii] See eg E Tendayi Achiume, ‘Migration as Decolonization’ (2019) 71 Stanford Law Review 1509; Max J Andrucki, ‘The Visa Whiteness Machine: Transnational motility in post-apartheid South Africa’ (2010) 10 Ethnicities 358. See also the Henley Passport Index <https://www.henleyglobal.com/passport-index> accessed 26 September 2023.

[xiii] EDRi, ‘The Human Rights Impacts of Migration Control Technologies’ (EDRi, 12 February 2020) <https://edri.org/our-work/the-human-rights-impacts-of-migration-control-technologies/> accessed 21 September 2023.

[xiv] See Fleur Johns, ‘Data, Detection, and the Redistribution of the Sensible in International Law’ (2017) 111 American Journal of International Law 57, 58.

[xv] Robert Beyer and Andrea Milan, Climate Change and Human Mobility: Quantitative Evidence on Global Historical Trends and Future Projections (IOM 2023) 21.

[xvi] See Monika Zalnieriute, Lyria Bennett Moses, and George Williams, ‘The Rule of Law and Automation of Government Decision-Making’ (2019) 82ÌýModern Law ReviewÌý425.

[xvii] Beyer and Milan (n 17) 10.

[xviii] ibid 21.

[xix] ibid 11.

[xx] Volker Türk, ‘Prospects for Responsibility Sharing in the Refugee Context’ (2016) 4 Journal on Migration and Human Security 45, 47.

[xxi] Jane McAdam, ‘Displacement in the context of Climate Change and Disasters’ in Costello, Foster, and McAdamÌý847.