How to Move Borders Without a War

Sovereignty and self-determination should be negotiable—and not under the barrel of a gun.

An illustration showing a dove moving a red line by carrying it in its beak
Tyler Comrie / The Atlantic

When U.S. Secretary of State Antony Blinken talks about defending the “rules-based order,” he is referring to the post–World War II liberal, internationalist, market-based system that, bolstered by human rights, seemed to reach its apotheosis with the end of the Cold War. One of that system’s foundational principles is the rejection of war as a way to change borders.

In exchange for giving up war, states gained guarantees for their territorial integrity. Which worked, or seemed to: International wars declined in frequency, though that owed more to the nuclear-superpower standoff than the Charter of the United Nations. Protection against intervention gave states something else too: a free hand against internal threats. International law enables this territorial conservatism by treating secession as a political matter left to states’ discretion. Many states forbid changes to their borders or require the whole population to agree to any change. Some, such as China, criminalize even peaceful advocacy for autonomy or independence.

Never a very pretty bargain, then—but as long as it worked, perhaps worthwhile. Now, though, that bargain is breaking down.

After the Cold War, the use of force by major powers has become routine—in the Balkans, the Sahel, the Middle East, Afghanistan, and now Ukraine. Putin invokes America’s recent, frequent wars—the historian Samuel Moyn notes that more than three-quarters of postwar U.S. interventions have happened since the Cold War—to justify his aggression. In international law, enough wrongs can make a right: Russia’s war is a violation, but it also contributes to the decay of the rule-based order that has already set in.

We are left with half the bargain: rigid borders without peace. The choice for the overwhelming majority of the world’s peoples is their current state or nothing, which is why, when the cause becomes urgent enough, violence seems the only option.

The slide back to a more violent order may be hard to halt, but a rethink about the other half of the bargain is still possible. No state should be coerced into ceding territory, but peaceably changing borders can have a positive outcome that better reflects the wishes of the communities living within them. Helping states create peaceful mechanisms for change can even be stabilizing, because it mitigates some of the causes of violence.

What doesn’t work—what history shows doesn’t work—is to stake the world’s stability on a rigid, unchanging order.

In 1939, the historian E. H. Carr wrote an influential critique of the international community’s first attempt at a rules-based order in the 20th century—the League of Nations system. The name we give that era, interwar, tells you how the experiment ended. Carr excoriated its reliance on legal mechanisms to regulate global politics: Courts and treaties inevitably entrenched the preferences of their designers—in that case, the victors in World War I—and kept down other powers such as the U.S.S.R. and Germany, which, not coincidentally, launched the next war.

Carr was no worshipper of violence. He wanted peace, and his message holds true for our own postwar experiment in rules-based order: If we don’t create peaceful pathways for change, we can expect the other kind.

Peaceful territorial change might sound implausible, but it’s been done. After World War II, a consensus emerged that colonies deserved independence, within their existing borders, as an exercise of self-determination. Sometimes colonial rulers resisted—as they did in Algeria, the Dutch East Indies, Mozambique, and Kenya—but mostly, independence was accomplished without open warfare. In many cases, Britain, France, and others handed over governance to local elites without violence. A series of UN General Assembly resolutions smoothed the path, declaring—preemptively, before war or crisis—that independence was normative.

Decolonization is not usually thought of as a form of secession, but it was—severing bonds between sovereign and subject. The moral case for self-determination, for independence and self-governance, was compelling, but the new legal order also made three brutal simplifications: Self-determination offered independence in exactly the borders colonies already had; self-determination applied to a territory’s whole population, not minorities within it; and the right to form new states involved only colonial peoples, defined as geographically separate, ethnically different groups.

The result was to limit self-determination’s seemingly radical potential to challenge borders. It offered independence to colonies but nowhere else. Combined with the new rule of territorial integrity, it meant that states, including newly independent colonies, would have rigid borders, whether their populations liked it or not. This was a game of musical chairs, and the music had stopped in 1945: If your group already controlled a state, you could keep it; if a colony, it would become a state, and you could keep that. But Kurds were out of luck; so were Kosovars, Tibetans, and every community that was a minority in someone else’s country with the right to defend its territorial integrity against internal rivals.

Unlike colonial populations, minorities are integrated into the state, but many suffer discrimination. Uyghurs aren’t colonial subjects, though that nuance may escape them as they sit in reeducation camps, forbidden from speaking their language or practicing their religion. Even in rights-respecting countries, minorities can be permanent political losers just because of demography.

But international law and the global order give minority communities few resources to negotiate the terms of their oppression. Without a right to exit, what kind of bargain can they make? Such de facto internal colonies can have complex, intertwined demography—Han live alongside Uyghurs in Xinjiang, Serbs alongside Albanians in Kosovo—so separation would mean that some people end up on the wrong side of any line. Of course, not drawing a new line means, mathematically, leaving even more people stuck in a state they don’t feel a part of.

Strict limits to self-determination seemed necessary to the new order’s architects—both in the former colonial powers and in the newly independent states—because the logic of self-determination was too compelling: If groups were forever contesting borders, that would jeopardize any stability. But this assumption was, and is, poorly tested and overstated.

Many countries’ elites fear that a right of exit would lead to endless fracture. That is not necessarily the case. When groups are given a choice, the reasons to remain integrated can overcome the arguments for separation, which is why—in Quebec, Scotland, and New Caledonia—people have voted to stay. When states agree to a political process that could lead to territorial division, violence is rare; much secession-related violence is actually resistance by the state.

The violence surrounding separatism has actually given rise to the current system’s one, controversial way to secede, but if you’re looking for a peaceful mechanism, you’re not going to like it. Remedial secession would allow exit in order to escape extreme persecution. It could even justify armed intervention, as happened in Kosovo in 1999, when NATO bombed Serbia to protect Kosovars. But this humane response is open to abuse, as Putin’s pretextual claims in Ukraine show. It also has real limits: It doesn’t help communities that are just ordinarily oppressed. It also generates perverse, horrifying incentives: Separatists have a stronger case if the persecution gets worse.

Moreover, remedial secession is not a peacetime mechanism. Kosovo is a sobering reminder of how morally urgent the need for a peaceful process is: If Kosovars deserve a state, they deserved it before the war, the million refugees, and the thousands of bodies, not because of them. Kosovars believe that—why else did they start fighting? If you agree, you should be troubled by how narrow the gate is through which people must pass to get a hearing, let alone to succeed in seceding. How narrow, and how bloody.

Separatism does, unquestionably, produce turbulence, if only because elites and majorities don’t want to see their state divided and many will fight to oppose even peaceful efforts at secession. If preserving stability were always an overriding consideration, then suppressing peoples’ aspirations for greater autonomy—even supporting states that suppress them—might be regrettable but necessary. That’s the bargain the global order struck. But if that bargain really is breaking down—if the rules-based order isn’t actually ensuring stability—it’s not clear that it is worth it. And because a lot of the violence in this world happens between people trying to govern themselves and states trying to stop them, creating peaceful pathways to negotiate that process could actually reduce violence and produce more stability.

One example of what a peaceful process can look like is a system of referenda. Communities within states could petition their state to hold a secession referendum on a portion of the state’s territory that they would define. If they won a majority (or possibly supermajority), the state would be obliged to negotiate independence. This model is similar to the plebiscites that were held after World War I, and to the process that Canada’s Supreme Court proposed for Quebecois separatism in the 1990s. It also applies to what Catalan separatists mean by a “right to decide,” although in the latter two cases, existing units of the state—Quebec and Catalonia—are the focus of separatist efforts, rather than a self-defined territory.

A self-defined plebiscite model has clear advantages. It depends not on ethnic or religious identity but on democratic consent—activists have to persuade their neighbors. Letting groups define the territory they hope to win a vote on avoids reliance on predefined territories within the state that can limit many minorities’ ability to make claims, because many states gerrymander internal boundaries to avoid creating units in which minorities predominate. (Kurds in Turkey are an example: Kurds predominate in the southeast, but Turkey doesn’t have major subdivisions like states or provinces.)

But a model that doesn’t rely on existing boundaries needs other ways to set limits: One way is to require a minimum population, to avoid micro-secessions. One might require at least 1 million people in the territory, but the precise limit is arbitrary: Make it lower, and more groups benefit; higher, and claims that succeed will have greater political weight. Finally, a plebiscite that leads to a negotiated exit can be framed as a human right grounded in a democratic desire, not something a group only gets as a remedy after extreme persecution.

There are other models. Secession can be based on existing territorial entities, as the U.K. has provided in different ways for Scotland and Northern Ireland. States can choose their own approach that realizes a right to exit in ways consistent with their constitutional traditions. A start would be to revoke laws that criminalize separatism or proclaim a country’s indivisibility. All workable models assume that local peoples’ wishes must be taken into account. To do otherwise—subordinating them to those of a countrywide population—is tantamount to simply reasserting the existing borders.

In any model, other states must be free to extend recognition to communities that achieve peaceful separation, or to support them if their state resists. But that doesn’t mean licensing military intervention: Protecting states against external threats is the core function of the global order, and wars based on bogus or highly contested claims of liberation, such as the one Russia has undertaken in Ukraine, should always be resisted. A right to decide gives a group a human-rights claim, not an expectation of military intervention by outsiders.

The ultimate goal is not necessarily to encourage a wave of secessions. The real value of a right of exit lies in giving minorities leverage to negotiate the conditions for going or staying. Creating that leverage might be a better guarantor of peace and stability than today’s rigid, failing order.

That order’s most beleaguered victims—the millions of refugees who have fled not a foreign invader but their own rulers—can testify to what its vaunted stability really means. So can the hundreds of millions more who do not flee but remain in what we call peace but cannot call justice: in their homes but in someone else’s country.

The world is a dangerous and unjust place, and it’s no longer clear that insisting on rigid, unchanging borders makes it less so. What we are seeing daily in Ukraine, Syria, Cameroon, Myanmar, Ethiopia, and in a dozen other blood sacrifices is proof that however daunting it is to imagine change, Carr was right: We either do this in peace or we do it in war.