The AI Journal

In December 1921 the Soviet government published a decree that did something no government had quite managed before. It did not exile the roughly 800,000 Russians who had already fled the revolution and the civil war that followed it; they had left on their own. It did not confiscate property, most of them having none left to confiscate. It simply declared, by administrative fiat, that they were no longer Russian. They kept their passports, where they still had them, as the paper of a country that no longer claimed them and that they could not claim back.

A person with no state is not a new figure in history. What was new was the number, the geography, and the fact that the great powers of Europe, who had just spent four years building institutions to prevent exactly this kind of mess, now had several hundred thousand people parked in Constantinople, Berlin, Paris and Harbin who could not cross a border, open a bank account, or be deported anywhere, because no state would issue them a document and no state was obliged to take them back. Fridtjof Nansen, the Arctic explorer turned League of Nations High Commissioner for Russian refugees, convened a conference in Geneva in July 1922 to deal with the practical end of this: travel. The result, the Nansen certificate, is usually told as a tidy bureaucratic fix. Fifty-two states eventually recognised it. It let a stateless clerk get on a train.

That is the story everyone tells, and it is true as far as it goes, which is not very far. The certificate itself, as the historian Claudena Skran documents in her study of the inter-war refugee regime, carried almost no substantive rights (Skran 1995). It was not a work permit or a right of residence, and it committed no state to let the holder stay, return, or ever acquire a nationality again. The 1926 arrangement that extended the scheme mainly cleaned up the terms on which holders could re-enter the country that had issued the paper, which tells you what the document was actually for: it let the authorities track a population moving across borders they otherwise had no legal standing to cross. A useful instrument for states. A thin one for the people carrying it.

What is easy to miss, reading the certificate as a travel pass, is what Nansen’s office had to invent to issue even that much. Before 1922, “stateless” was not an administrable fact. It was something a consular clerk discovered by failing to find a country willing to claim a person, not a status a bureaucracy could stamp. To recognise the certificate at all, fifty-two governments had to agree, in effect, on a procedure for certifying that a specific person belonged to no state, a determination made without that absent state’s consent or even its presence in the room. That is not a travel document. It is a new administrative object, built because the travel problem could not be solved without it. The paper is the part that looks like the achievement. The category is the part that had to exist first.

The sequence that followed shows which one actually travelled forward. When the office extended coverage from Russians to Armenians in 1924 and later to Assyrians and Saar refugees, it was extending the category, not the card; the document for each group looked the same because the underlying determination, statelessness independent of persecution, was the same kind of fact. When the 1933 Convention Relating to the International Status of Refugees was drafted, its Article 3 became, on Gilbert Jaeger’s account as a former UNHCR protection director, the first non-refoulement clause in treaty law: a binding promise not to turn a recognised person back (Jaeger 2001). That promise is not about transport. It is a legal consequence attached to a status, and a status needs a stable definition before anyone can be bound by what follows from holding it. The Nansen Office closed in 1938, its passports superseded within a few years by newer documents. The category did not close with it. The 1951 Refugee Convention, in Article 1A(1), simply folded every existing Nansen certificate holder into its own definition of refugee by reference, inheriting the determination rather than re-litigating it.

Here the case for the paper, rather than the category, gets its strongest run. The document secured no right to work, no right to remain, no right of return; Skran’s account is blunt that it offered holders little beyond a crossing. If the achievement is measured in rights delivered, there was no achievement, only forty years of states agreeing to recognise a status while declining to attach anything to it that cost them something. Hannah Arendt, writing from the vantage of 1951, made the sharper version of this objection: a right held only by courtesy of states that owe you nothing is not a right at all, and the stateless of the inter-war years lived, in her phrase, on charity rather than on law, their survival never a matter anyone was bound to secure (Arendt 1951). On that reading, the category was not an achievement either. It was a label that let governments feel they had solved a problem they had in fact only agreed to name.

The objection is correct about what the category failed to deliver and wrong about what that failure shows. A right needs someone bound to honour it, and nobody was bound to honour the Nansen certificate’s claims on a person’s future. But a right also needs an object: a determinate person recognised as falling under it, independent of the one authority, a home state, previously assumed to make that determination. That object did not exist before 1922 in any form a second state could act on. The 1954 Convention Relating to the Status of Stateless Persons was written specifically because the 1951 Convention covered only the stateless who were also refugees fleeing persecution, leaving everyone stateless for other reasons uncovered; drafting that gap required treating statelessness as a condition separable from refugeehood, which is the exact distinction the Nansen Office had to operationalise thirty years earlier to issue a certificate to a Russian émigré who had fled no one in particular. Arendt is right that the card bought no safety. She is describing the thinness of the right that got attached to the category, not showing that the category was empty. The thinness is the reason the category had to keep being refined by the conventions that followed it, in 1933, in 1951, in 1954, each one working on the object Geneva had defined and none of them going back to redefine it from scratch.