Evidence & history
Rights without a state: from the Jewish experience of statelessness to today's refugee system
The real question is not which people is better at surviving. It is this: once someone loses the effective protection of their own state, who makes sure they can still live as a normal person? I checked each step against treaties, UN figures, and museum archives.
In theory human rights belong to every person. In practice most of them only work through a state that recognises you. Europe's Jews learned this in the 1930s; the modern refugee system was built partly in answer. It protects people from being sent back, but it still cannot promise anyone a way out of waiting.
Two words that are not the same
A refugee is someone outside their own country because of a well-founded fear of persecution, who cannot or will not rely on that country's protection. A refugee may still hold a passport. A stateless person is someone no state considers a national under its law. A stateless person may never have crossed a border.
The two overlap: the 1951 definition explicitly covers refugees "not having a nationality". Mixing them up hides the real problem, which is not one label or the other, but losing the state that turns rights on paper into rights you can use.
The modern order assumes a quiet chain: birth gives you a legal identity, a state recognises you, documents prove it, and those documents unlock work, property, school, travel, and a vote. Break any link and most other rights stop working, even though they are still "yours".
Why the Jewish experience became the extreme case
The precedent came earlier: in 1922 the League of Nations created the Nansen passport for Russians whom Soviet decrees had stripped of nationality. The Jewish case pushed the problem to its limit. A common claim needs correcting here: the 1935 Nuremberg citizenship law did not make German Jews stateless. It demoted them to nationals without political rights. The collective stripping came with a 1933 law and then the Eleventh Decree of November 1941, which cancelled the nationality of Jews living abroad, deportees included.
Before that, the exits closed. At Evian in July 1938, 32 countries met to discuss Jewish refugees and almost none agreed to take more; in 1939 the St. Louis, with 937 passengers, was turned away by Cuba, the United States, and Canada. About six million Jews were murdered in the Holocaust. Hannah Arendt, writing in 1951, drew the lesson: people only noticed a "right to have rights" when millions had lost it.
Modern Zionism's answer, from Herzl's 1896 pamphlet onward, was a political home "secured by public law" so that Jewish safety no longer depended on the goodwill of others' majorities; Britain endorsed a "national home" in 1917. One explanation should be dropped outright: that Jews were "naturally good with money, so they bought the land". Jewish wealth varied enormously, and by 1947 Jewish individuals and bodies such as the Jewish National Fund held only about 6 to 7 percent of Mandate Palestine's land. Land purchase was not how the state was made.
A solution for one people, a displacement for another
In November 1947 the UN General Assembly voted to partition Palestine. In the 1948 war, according to the UN's own history, over half of the Palestinian Arab population fled or were expelled; UNRWA counts more than 700,000 people. Their descendants are among the roughly 6 million Palestine refugees the UN counts today.
They also sit in an odd legal corner: Art. 1D of the 1951 Convention excludes people receiving help from another UN agency, which for Palestine refugees means UNRWA rather than UNHCR. After 1948, a large number of Jews also left or were pushed out of Arab and Muslim countries; about 850,000 is the commonly cited figure, though estimates and the balance between expulsion and emigration are debated.
This is not to say the two histories are the same. It is to say that solving one group's safety without designing another group's rights can create a new problem that lasts for generations.
The system we have, and where it stops
The 1951 Convention (146 parties) and its 1967 Protocol (147) define who is a refugee and forbid sending them back to where their life or freedom is threatened. The 1954 Convention (99 parties) gives stateless people a status; the 1961 Convention (82 parties) tries to stop statelessness being created in the first place. UNHCR protects and seeks solutions, but only by working through governments; it is not a world government and cannot hand anyone citizenship.
What none of these do is guarantee an exit. The three traditional solutions are going home voluntarily, integrating locally, or resettlement to a third country. At the end of 2025, 117.8 million people were forcibly displaced, 70 percent of refugees had been in exile for five years or more, and at least 4.5 million stateless people were reported, with the true number higher.
Resettlement shows the gap most clearly: about 2.5 million refugees need it in 2026, while about 37,000 left through UNHCR-assisted resettlement in 2025, or about 81,800 if all resettlement and sponsorship arrivals are counted. Each state can give a reason why it should not be the one to take more. Add those reasons up and the refugee still has no answer. That is a collective action problem, not a gap in any one law.
Neither closed nor borderless
From here on this is design thinking, not a description of what international law already guarantees. A fuller floor of rights would include a verifiable identity even without a valid passport, a basic right to work, schooling for children, basic health care, a legal way to hold property and money, and a long-term exit that cannot be left pending forever.
Full closure traps people who really are fleeing persecution. Unconditional openness ignores real questions of housing, schools, budgets, integration, security screening, and scale. The harder and more useful question is how to keep borders and democratic self-government while making sure that losing a state's protection no longer means losing the right to a normal life.
Compressed into one chain, the lesson reads: lost protection, lost status, long marginalisation, a second generation that inherits the problem, and conflict that hardens into territory and nationalism. A better system cuts that chain early. The full evidence map linked on this page lays out the flows, the treaty table, and every source.