The 10 days at Evian, and what they tell us about the next refugee crisis

On 6 July 1938, President Franklin D. Roosevelt opened a conference at the Hotel Royal in Évian-les-Bains, on the French shore of Lake Geneva, by inviting the assembled delegates to declare their governments' willingness to admit refugees fleeing the German Reich. Over the next ten days, representatives of 32 states would take the floor one by one. Twenty-nine of them said, in effect, no. The United States would not loosen its quota regime. Britain pointed at Palestine and washed its hands. The Dominions refused outright. The Dominican Republic, the lone country to take the question seriously, offered to settle perhaps 100,000 settlers on agricultural land it did not yet have. Roosevelt left Évian convinced that the most powerful nations on earth had washed their hands of the problem in unison.

A century on, the structural pattern is more important than the moral indictment. Évian was not a failure of conscience so much as a failure of coordination, a clear case of states, acting rationally inside their own domestic constraints, producing a collective outcome that none of them wanted. The same arithmetic keeps reappearing, from the boat people of 1979 to the Syrian crossings of 2015 to the Channel crossings of last summer. The lesson of Évian is not that leaders are uniquely cruel. It is that the international refugee regime, as built, is a non-régime, a diplomatic shape that produces the worst plausible outcome because no single actor bears the cost of refusal and no single actor captures the benefit of admission.

A conference designed not to decide

The call for Évian came from Roosevelt in March 1938, weeks after the Anschluss, when Austria's 185,000 Jews lost the partial protection they had still held under a sovereign government. The conference was framed as humanitarian. It was also carefully designed to be non-binding. The United States, which had convened it, did not want a document it would then be expected to honour. Each delegation was invited to make a statement. No votes were taken. No targets were set. No enforcement mechanism was even sketched.

Deutsche Welle's reconstruction of the ten days records the rhythm: country after country explaining why this category of refugee, from that part of Europe, at this moment, was a problem for someone else. Australia, Canada, France, Switzerland, the United Kingdom and the Irish Free State all said variations of no. Only the Dominican Republic, under President Rafael Trujillo, and a handful of Latin American states offered anything concrete, and the Dominican offer was hedged with conditions that would, in the end, deliver a small fraction of the 100,000 slots initially floated. The conference adjourned on 15 July 1938. The question it had been convened to answer was left, formally, open.

What this produced, in the months that followed, was a predictable redistribution. The countries that had said yes absorbed what they could. The countries that had said no watched the problem migrate toward them via secondary movement, irregular routes, and the slow grinding pressure of stateless populations with nowhere legal to go. The Évian outcome was, in technical terms, a textbook collective-action failure. Every state had an individual incentive to free-ride on the hospitality of others. The aggregate result was near-total refusal.

The counter-narrative: the states that did move

The dominant memory of Évian is the refusal. It is worth pausing on the exceptions, because they are doing more of the historical work than is usually credited. The Dominican Republic under Trujillo, a dictator whose antisemitism in his own country is well documented, nevertheless extended citizenship to roughly 700 Jewish refugees in the period that followed, an offer that depended on the personal intervention of a Dominican envoy in Berlin and a small group of European Jewish relief agencies. Costa Rica, Nicaragua and a handful of other Latin American states took in smaller numbers. The Philippines, then a US Commonwealth, accepted several hundred German-Jewish refugees via a sponsorship scheme organised by the American Jewish Joint Distribution Committee. None of these numbers moved the aggregate. All of them changed individual lives by the thousand.

The structural point is that the countries which did accept refugees were not the rich, the powerful, or the geographically convenient. They were, in most cases, peripheral economies whose admission policies were set by small bureaucratic decisions and personal initiative rather than grand strategy. The lesson is uncomfortable for any contemporary state that insists it cannot move because of capacity: in 1938, the states that moved were the ones whose officials had decided, before the conference opened, that they were going to try.

The structural frame, in plain language

Refugee regimes, the international architecture for managing forced displacement, have not changed as much as the surrounding politics have. The 1951 Refugee Convention, drafted after the Second World War in a Europe still half-depopulated, set the legal floor. It did not, and was not designed to, redistribute the cost of receiving displaced people. The richer the destination state, the more it can argue, credibly, that the truly needy are elsewhere. The poorer the transit state, the more it can argue, credibly, that the system is gamed.

What Évian demonstrated, and what subsequent crises have repeatedly confirmed, is that the international community can convene, deliberate, and still produce a coordinated non-decision. The states of Évian did not meet to refuse refugees. They met, ostensibly, to share them. They refused not because they hated the people on the boats, in most cases, but because the cost of saying yes was domestic and immediate, and the cost of saying no was diffuse, foreign, and slow. When the structure pays you to refuse, you refuse.

The same pattern shows up in the politics of climate displacement, where the states most responsible for emissions are the least exposed to current flows, while the states least responsible are the most exposed. It shows up in the politics of mixed migration, where transit states absorb enforcement costs that the destination states would otherwise pay. It shows up in the politics of Dublin-style regulations, where the front-line member state of first arrival bears the administrative weight and the wealthier interior states avoid it. Each is a micro-Évian: a coordination failure dressed up as a policy choice.

What is different, and what is not

The 88 years since Évian have produced real changes. The 1951 Convention gives refugees legal status, not just charity. The UN High Commissioner for Refugees, born in the wake of the conference's failure, has a real budget and a real mandate. The European Union's Common European Asylum System, however imperfect, is a genuine attempt to share responsibility. Private sponsorship programmes, pioneered by Canada in 1979 and now spreading, build the kind of civil-society capacity that the Dominican Republic's 1938 offer depended on. The toolkit is larger.

What is not different is the political economy. The states that lose elections over refugee admissions are still, mostly, the destination states. The states that win elections by tightening border policy are still, mostly, the destination states. The transit states are still being asked to perform the labour of containment on behalf of the destination states. The international community, when it gathers to discuss displacement, still tends to design conferences that produce statements rather than commitments. The shape of Évian is still the shape of the room.

The contemporary question is whether the next large displacement event, whether from climate stress in the Sahel, from a new conflict in the Middle East, or from the slow grinding of authoritarian consolidation in a half-dozen countries, will meet a different architecture. The honest answer is that the architecture has not been built yet, and that the political incentives to build it are still, in most receiving states, on the side of refusal. Évian did not fail because the world was uniquely evil in 1938. It failed because the structure paid the world's leaders to fail. The structure is still there. The question is whether anyone has an interest in rebuilding it before the next crisis forces another ten days at the lake.

This publication treats Évian not as a moral parable but as a structural case: the diplomatic pattern repeats whenever the cost of admission is borne domestically and the cost of refusal is borne elsewhere. The wire treatment of refugee anniversaries tends toward the commemorative; Monexus reads the record for what it tells us about the next crisis, not the last one.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://en.wikipedia.org/wiki/%C3%89vian_Conference
• https://en.wikipedia.org/wiki/Dominican_Republic_and_the_Holocaust
• https://en.wikipedia.org/wiki/1951_Convention_relating_to_the_Status_of_Refugees