We need to redefine what we mean by "ethnic cleansing" far beyond our conventional understanding of the term, so that it includes all the ways states constantly control their demography: systems of government, immigration policy, border controls, citizenship laws, and more. The term should be understood broadly, and as part of a still wider suite of policies, called "ethnic engineering," by which identity-based states ensure identity-based control of territory.

And the international sovereign order, and the stability it gives us, could not exist without it.


1. What we talk about when we talk about ethnic cleansing

  1. A similar problem applies to “terrorism,” which also has no internationally accepted definition and probably never will. I like to say, only partly in jest, that the only definition everyone would agree on is that terrorism “is what they do to us.”

  2. A note on terms. I use the term “ethnic cleansing” to describe the forced removal of any lineage-based identity group, be it ethnic, religious, sectarian, linguistic, or what have you. It is my view, as of this writing and pending further exploration of the matter, that the effects of identity-based security dilemmas operate similarly on all of these categories.

When a new term is invented in response to a specific event, we often struggle to decide how to apply it elsewhere. “Genocide” was Raphael Lemkin’s attempt to describe the Holocaust. Once the term entered common usage, the debate began. What other genocides were there? Did the Turks commit genocide against the Armenians? Did the United States commit genocide against the Native Americans? Were the Khmer Rouge genocidal? Was Stalin? What about Caesar's conquest of the Gauls? Etc.

The same is true of ethnic cleansing. The term was first used to describe the breakup of Yugoslavia in the 1990s, and that is where our minds often go when we consider what it means. So the genocidal massacre of more than 7,000 Muslim men and boys at Srebrenica in 1995, which horrified the world and finally moved the United States to decisively intervene in the conflict, is the indelible symbol of ethnic cleansing.

This way of thinking about ethnic cleansing is a mistake, akin to observing the color red and assuming you have the whole spectrum. It misses nearly the entire picture. Genocides are rare. Ethnic cleansing is endemic. Benjamin Lieberman’s book “Terrible Fate: Ethnic Cleansing in the Making of Modern Europe” details just how ubiquitous the tactic was to building the international order Europe enjoys today, but it’s not just Europe. Demographic control is integral to the nation-state project.

The United Nations Office on Genocide Prevention and the Responsibility to Protect notes that the term ethnic cleansing “has not been recognized as an independent crime under international law.” Nor is there an official definition. I suspect I know why. Governments understand that defining ethnic cleansing would be a fraught business, and would expose the many ways states attempt to control their demography.

The best we have so far is probably from the UN Commission of Experts studying violations of international humanitarian law in the former Yugoslavia, which defined the term as “rendering an area ethnically homogenous by using force or intimidation to remove persons of given groups from the area.” This definition is useful because it emphasizes the intent and effect of the act more than the tactics employed. This is the definition I want us to use now. When you think of ethnic cleansing, I do not want you to think of Srebrenica anymore, at least not in isolation. Genocide is only the most brutal means of ethnic cleansing, getting rid of a group in your territory by exterminating it entirely. Many if not most states have a strong interest in maintaining or solidifying their current demographic structure, which was seldom achieved peacefully. Ethnic cleansing is not an anomalous war crime, an extreme and unusual depravity in international affairs that the Bosnian Serbs invented, any more than the apocryphal apple that fell on Isaac Newton's head marked the commencement of gravity. Ethnic cleansing is our shared history, and it is built into the way the international system works.

2. Why do people ethnically cleanse each other?

  1. Academically, this is not actually true, and works such as Timothy Snyder’s “Black Earth: The Holocaust as History and Warning” have been edifying in understanding the mindset behind Nazi crimes. In the public discourse, however, I find there’s little room for this kind of subtlety, and any exploration of motivations is shouted down as apologia. Public debates tend to focus instead on the tactics, the sequence of steps the Nazis took in their extermination campaign.

  2. This can also happen if the group poses no threat themselves but is tied to a much larger identity group that does threaten the identity-based integrity of the state. This would partially explain Myanmar’s targeting of the tiny Rohingya minority, who are mostly Muslim, even though they are demographically insignificant and pose basically no threat to the state at all.

  3. Other sources report significantly fewer. Displacement figures for both Croats and Serbs vary depending on whom you ask.

  4. My paper about this incident is here, though since writing it I have found a greater disparity in reported displacement numbers for both Serbs and Croats than I report there.

  5. It would have been really nice if Western policymakers had figured this out early in the Syrian conflict, and realized that “Assad must go" was a potentially genocidal demand.

  6. The census can be a highly destabilizing factor for the same reason, as it effectively ranks identity groups numerically, allowing larger or growing groups to flex their muscles and frightening smaller or relatively declining groups. Some countries, such as Nigeria, have had to “de-ethnitize” their census just to stop civil wars and military coups from breaking out each time they run one. Perhaps the increasingly diverse and brittle United States should consider the idea.

Apart from obfuscating much of the story, there's another reason why lumping ethnic cleansing in with genocide is analytically unhelpful. Genocide is so odious that the perpetrator’s motivations are basically irrelevant. “But why did Hitler hate Jews so much?” is the wrong question. We wave garlic and crosses at genocide. It is pure evil and we must stop it and defeat it. It is almost like a corollary to Godwin's Law. The first invocation ends all political arguments.

Ethnic cleansing, however, has a very identifiable cause in almost every case, and one that we can empathize with even if we are horrified by the result. It arises, almost without exception, from an ethnic security dilemma and a moment of political revisionism, often after the collapse or breakup of a polyglot entity such as an empire or totalitarian state. Disparate groups fear that other groups will unite to dominate them or seize lands they claim for themselves, so they preemptively mobilize to defend themselves and secure their claims. Generally, people don't ethnically cleanse because they're evil, they do it so that no one will do it to them.

The ethnic security dilemma is zero-sum and self-reinforcing. In 1991, when Yugoslavia broke up, Croatia seceded, and then an enclave in Croatia called the Krajina, where Serbs held a slight majority, counter-seceded. The Krajina was diverse and had been so for generations. Under Yugoslav rule, it didn't matter who lived there, as it was part of one multi-ethnic communist dictatorship. But the dispute of whether it would belong to a Serb-dominated or Croat-dominated political unit rendered the very existence of each group there a threat to the other. In order to ensure Serb control, the breakaway Serbs, backed by Belgrade, kicked out nearly the entire Croat population, possibly as many as 250,000 people. In the summer of 1995, the military tables turned, and the Croat army stormed the Krajina and took it over. Nearly the entire Serb population—estimates range from 150,000 to as many as 300,000—fled. Most never returned, and today Croatia’s Serb population is about a third of what it was before the war. For both sides, the fear of being ethnically cleansed incentivized ethnic cleansing and proved self-fulfilling. This is typical.

Truth and reconciliation do not end the ethnic security dilemma. In fact, states often subsequently forget they ever committed such acts, or even celebrate them. If the ethnonational claim to a territory is inconvenienced by the presence of others, the forced removal of those others undermines the national origin myth, and therefore must be expunged from memory. It cannot be taught in schoolbooks. It cannot be acknowledged or apologized for. The surviving populations move on with life as normal, as if nothing ever happened. The thing about ethnic cleansing is, you don't notice the people who are no longer there. The whole point of the undertaking was so that you wouldn't have to think about them anymore.

Democracy does not solve the ethnic security dilemma, it exacerbates it. Getting rid of the Rohingya polls quite well in newly democratic Myanmar. It was the 1990 elections in the Yugoslav republics that brought in nationalists who broke up the state and cleansed it of its diversity. The 2005 Iraqi elections handed power from the minority Sunnis, who had largely run things under Saddam Hussein's dictatorship, to the majority Shia, who cleverly and successfully campaigned on a platform of being more numerous, while the Sunnis largely boycotted the proceedings. The inevitable civil war erupted the following year, and ended less because of the American “surge” and more because the Sunnis realized they were going to lose. During the surge, the United States paid the two main Iraqi factions to not fight each other, and stamp out their respective unsavory elements, until after American forces left. The instant they were gone, sectarianism returned, and the Islamic State invasion in 2014 caused a re-run of the war, again ending with Sunni defeat. (One could imagine this dynamic continuing.) Next door, the minority-controlled Syrian government of Bashar al-Assad took note. When the Arab Spring came to Syria, the government demonstrated that there was literally no limit to what it would do to prevent a similar outcome. “Democratic” in the context of the ethnic security dilemma is good for majoritarians and bad for minorities, and incentivizes ethnic cleansing to ensure electoral success. The fact that the surviving Hutu genocidaires, who linger on in the ungoverned spaces of the Democratic Republic of the Congo, call themselves the “Forces for the Democratic Liberation of Rwanda” is telling. Democratic liberation is terrific if you’re an ethnic militia whose ethnicity constitutes 85% of the population. Not so if you’re 15%. Ethnic cleansing is demographic gerrymandering. It is democracy by other means.

Your own personal aversion to ethnonationalism, and your diverse group of friends, is irrelevant here. A great many Yugoslavs had, and still have, no interest in separatist projects, and intermarriage between groups was common before 1991. Yet the war came anyway. The ethnic security dilemma operates at the collective level. The existence of a nationalist leader and his or her constituency, no matter how limited, is terrifying to other groups and causes them to react, setting off a chain reaction in which voices of moderation are bulldozed.

3. So what constitutes ethnic cleansing?

  1. It’s technically possible, as of this writing, that Congress could still step in to save the day here.

  2. Let’s highlight two important differences here. First, the tactics reportedly used by the Myanmar regime could also qualify as war crimes, while a bureaucratic deportation regime like that of the United States would not. Moreover, the DACA recipients are generally recognized as citizens of somewhere, even if they've barely lived in that somewhere. Not so the Rohingya, who are rejected by Bangladesh (as well as basically everybody else) and are stateless. We should not minimize these differences. Nevertheless, both acts should be understood as ethnic cleansing.

  3. The development investments, it appears, often backfire, as Africans save up the money from working in EU-funded projects in their home countries to finance the expensive and perilous trek to Europe.

  4. In general, it's a bad look to outsource your immigration enforcement to slave traders.

  5. Though they did reportedly massacre some who stayed.

  6. See Laura Silber and Allan Little’s “Yugoslavia: Death Of A Nation.” Penguin, United States, 1997, p. 358.

Since ethnic cleansing entails any method of identity-based demographic control, we have to define the term outwards, far beyond what is conventionally understood. States are creative about this sort of thing.

The most indisputable form of ethnic cleansing is the kind that apes the Balkan tactics that spawned the term. A military or paramilitary campaign of terror to force most or all of an ethnic group out of a territory, such as the current campaign in Myanmar against the Rohingya people, obviously qualifies. UN human rights chief Zeid Ra’ad Al Hussein told the Human Rights Council in September that the Rohingya exodus “seems a textbook example of ethnic cleansing.” Of course, since the term is relatively new and there’s no legal definition, there is no textbook. “Textbook” here means “something that resembles what happened in Bosnia.”

Accusing Myanmar of ethnic cleansing is not difficult. The next step is harder. At the same time that Myanmar is evicting nearly a million people who have lived in the country for a long time but that the state rejects as citizens, the Trump Administration is doing the same in the United States. Its decision to end the Deferred Action for Childhood Arrivals (DACA) recipient protections, presumably leading to a systematic deportation of some 800,000 people who have lived in the United States for nearly their entire lives, looks different in this light. A state has arbitrarily decided not to grant citizenship to people who live in its territory, and is now getting rid of them, sending them back to a country they barely know just because they share an identity with the people of that country. Because in America's case, nearly all the DACA recipients are from Latin America, and almost none are white.

The tactics used to carry out this program are of course very different. Myanmar reportedly burns villages, uses rape as a weapon, and plants landmines. ICE does not do these things. But this says less about the relative humanity of the two regimes and more about their relative wealth. Strong states with resources and territorial control have a monopoly on the use of force (or, borrowing from the UN, shall we say “force and intimidation”) and can deport millions under a legal regime by using the bureaucracy and law enforcement. Weak states that struggle to control their territory don't have this luxury, and resort to pogroms. Alex de Waal called Sudan’s Darfur policy “counterinsurgency on the cheap.” Myanmar’s Rohingya policy is “strong on immigration on the cheap.”

The two policies look physically different but from the perspective of the state their goal is the same. Nearly a million people of a specific background want to live in the country, and aren't welcome, and the state is forcing them to leave. The tactics used do not determine whether a policy is ethnic cleansing, the result does. Murder by hatchet and murder by administering a painless but fatal dose of poison are both murder. If Myanmar’s actions constitute ethnic cleansing, why don’t those of the Trump Administration?

I will go even further, however. The above is “positive ethnic cleansing,” or removing existing people from a territory. There is also “negative ethnic cleansing,” or preventing people who would naturally want to migrate to and live in and be a citizen of your state from doing so based on who they are. Negative ethnic cleansing can be just as violent and identitarian as the positive variety, but it’s so ingrained in our international system that many of us don't even realize it’s there. But migrants do. Negative ethnic cleansing includes any kind of identity-biased restrictive immigration policy. It is the policy of the successful ethnonationalist. Having secured identity dominance in her territory, she will now defend that territory against migration- or fertility-based demographic change.

Consider the recent Foreign Policy series “Europe Slams Its Gates.” The series chronicles the European Union’s attempt to stop the flow of migrants from North Africa through a combination of economic investments and security arrangements with foreign governments and militias. The effectiveness of these tactics is debatable. So is the morality of the security arrangements. But as with the DACA case, the real question here is the intent. And the intent is obvious. Europe wants to block a flow of migrants from Africa and the Middle East in order to maintain demographic stability. Absent the policy, tens of millions of people of African and Middle Eastern descent would be in Europe right now, living, working, intermarrying, and (depending on citizenship laws) becoming politically enfranchised citizens. Blocking them makes no economic sense. Flows of mostly young, working age migrants are a net economic positive for aging, industrialized countries facing pending labor shortages (a fact of life for nearly every country in Europe). So the only conceivable reason for Europe to spend enormous financial outlays on immigration restrictions that compel migrants to take dangerous illegal trips across the Mediterranean by boat, perishing by the thousands annually in the process, is to keep Europe demographically “European.” This is negative ethnic cleansing.

States can also legally codify ethnic purity through blood citizenship laws. A restrictive regime gives the state the legal standing to remove unwanted peoples and call it immigration policy. If a homogenous country requires parental citizenship and places high hurdles on naturalization, it will likely remain homogenous in perpetuity. If that country is aging so rapidly that it faces acute labor shortages and population decline, is wealthy enough that many would like to emigrate there, and still maintains that citizenship regime, we do not have to wonder why. We know.

There is a final element to ethnic cleansing, and that is, to borrow from Mitt Romney, the concept of “self-deportation.” Romney’s comment during the 2012 American presidential campaign drew chuckles, but self-deportation is not funny. Often a group isn't forced to leave their homes, but rather reads the writing on the wall and leaves of their own accord. This doesn't happen by chance, and we must ask ourselves if it is even useful to distinguish it from ethnic cleansing when it does happen. Often it is the result of persecution towards an impoverished minority, such as the undocumented Mexicans Romney was speaking of. Long before Myanmar's current ethnic cleansing drive, its policies towards the Rohingya were so unpleasant that a majority of the group's members had already left the country. But a changing political dynamic can also encourage previously dominant settler colonials to flee en masse.

Consider the pieds-noir of Algeria. In 1960, there were a million of them, many tracing their lineage back to France's seizure of the territory in 1830, or even earlier. In 1962, the year Algeria gained independence, nearly all of them were evacuated, and by 1970 fewer than 50,000 remained. If such a total removal of a population from a territory in such short order is not ethnic cleansing, the term has no meaning. The Algerians may not have actually removed most of these people. They didn’t have to. The changing political landscape and security situation compelled the pieds-noir to remove themselves. Similarly, Croatian leader Franjo Tudjman washed his hands of the cleansing of the Krajina after Croat forces seized it in 1995, causing virtually the entire Serb population there to flee. “I thought 60 to 70 percent of the Serbs would stay,” he later said, “that they would understand that democratic Croatia will guarantee their ethnic rights. So the Serbs themselves are to blame for their destiny.” In fact, the invading army left escape routes, and the Serb leadership and population took them without hesitation. If you change the political dynamics of a territory and suddenly an entire identity group packs up and leaves in response, is this appreciably different than formally throwing them out?

Ethnic cleansing is, in short, any policy that ensures demographic dominance of a group or groups in a territory or polity to the exclusion of another group or groups. Pogroms, discriminatory immigration systems, harsh border regimes, blood-based citizenship laws, and even dynamics in which identity groups feel sufficiently unwelcome that they simply remove themselves, are all part of a suite of policies intended to effect and maintain an ethnonational program. For my part, I see little difference between them. They all have the same purpose. They all force people to not be where they want, often with fatal results, and do so on identity grounds.

4. Ethnic engineering

  1. Unlike most other terms here, I made the first two of these up. It’s possible terms exist and I just don’t know them, but I’ve never heard an official term for either of these tactics, which is odd, as they are both common.

  2. Please see Anthony Howell and C. Cindy Fan’s "Migration and Inequality in Xinjiang: A Survey of Han and Uyghur Migrants in Urumqi,” Eurasian Geography and Economics, 2011, 52, No. 1, p. 123.

  3. Serbia faced a similar problem in the 20th century in Kosovo to that of the Israelis in the West Bank today. It did not end well.

  4. That is, unless you want to get creative about how you define Apartheid and Jim Crow segregation, arguing that the physical separation they require amounts to ethnic cleansing within a single polity. We could call this “ethnic zoning.” Consider American segregation. If our definition of ethnic cleansing is “rendering an area ethnically homogenous by using force or intimidation to remove persons of given groups from the area,” “separate but equal” would definitely qualify, as would redlining, and even gentrification and gated communities or suburbs. American whites might have to share a national polity with African Americans, but they will go to extraordinary lengths to avoid sharing neighborhoods, property wealth, school systems, equal treatment by law enforcement, etc. In this fashion, they Balkanize America, achieving political dominance and control of territory and wealth without formally splitting the country up or kicking people out. “States rights” in this framing is really about the right to ethnically zone, transferring more equitably distributed federal resources to local polities that can keep white money and land in white hands, with the “other” out of sight and mind, merely a tolerable, invisible reservoir of exploitable low-wage labor.

  5. Example: in the Maldives, Islam is the state religion and officially, the population is 100% Muslim. No human population ever became 100% anything naturally.

  6. Example: Khartoum’s attempt to enforce sharia law upon its largely non-Muslim southern population played a role in renewing the North-South war in Sudan, ultimately leading to South Sudan’s secession.

There are three notable policies that are not ethnic cleansing per se but are nonetheless part of the ethnonational program towards unwanted identity-based populations: ethnic swamping, ethnic subjugation, and ethnic assimilation. To understand these, we need to define the term “ethnic cleansing” outwards yet again. Let us instead think of “ethnic engineering,” different ways of ensuring a group’s dominance in a territory, of which “ethnic cleansing” is just a subset.

Ethnic swamping occurs when the state claims disputed territory, and wants to solidify its claim by settling members its own group until they demographically overwhelm the locals. Ethnic swamping is not ethnic cleansing, but a substitute. It is ethnic cleansing by other means. This tactic does not preclude ethnic cleansing or genocide: the settling of most of the countries of the New World after 1492 arguably involved all three, often in tandem. The difference with ethnic swamping is that you’re not actually making the group you dislike leave. You’re simply out-populating them.

While settler colonial projects almost by definition must attempt ethnic swamping, the most obvious contemporary example would be Chinese settlement policy towards Xinjiang and Tibet, in a country where Uyghur and Tibetan populations are hopelessly outnumbered by Han Chinese. By encouraging westward migration, either state-arranged or, increasingly, self-initiated, the Chinese government is making both groups a minority in the territory they would claim as their own, thus rendering impossible any claims to independence. Even the Dalai Lama has largely given up on the idea. While the extent of this practice in Tibet ignites stormy debate, Xinjiang’s numbers are available, and they are notable. In 1945, Uyghurs were nearly 83% of the population, to 6.2% Han. By 2008, it was 46.1% Uyghur and 39.2% Han. Whatever the motivations for such population transfers, the result demographically ropes the disputed region to the national core, locking in Chinese territorial integrity while allowing the state to dictate cultural dominance, regional wealth distribution, and security arrangements. And that is what ethnic swamping is all about.

The Israeli settlement expansion in the West Bank is a highly ineffective attempt at ethnic swamping. The Israeli policy bears some similarities to the Chinese one, though it has far less international recognition. However, it has a fatal flaw: the Israelis quite simply do not have the numbers. China has a reservoir of some 1.24 billion Han Chinese to swamp perhaps three million Tibetans and 11-15 million Uyghurs. By contrast, if one includes the Palestinian diaspora, Israeli Jews are already outnumbered by their Palestinian foes, and this will only increase in coming years, as the Palestinians have higher birthrates. Even a universal aliyah probably wouldn't be enough, as there are fewer than 15 million Jews on Earth. You can’t ethnically swamp if you are a minority. At least, you can’t do it democratically.

Of course, nobody likes to be ethnically swamped, not the Native Americans, not the Uyghurs, not the Tibetans, and not the Palestinians. They correctly see the arrival of newcomers as a sustained attempt to expropriate their claim to the land, and they know they will face discrimination and distrust under alien control. So to maintain such a program, one needs another tactic: ethnic subjugation.

Ethnic subjugation occurs when an identitarian government responds to a violent uprising by a group, particularly a regionally concentrated group, by collectively punishing the whole group. Ethnic subjugation is not genocide, but a substitute. It is genocide by other means. It is an attempt to destroy, in whole or in part, not a group, but a group’s capacity to be a political force. Not surprisingly, it is often mistaken for genocide. The Save Darfur Coalition routinely called the Sudanese government’s actions in Darfur genocide, but this labeling muddied the term and missed the point. Sudan had no reason or desire to exterminate the tribes of Darfur. It just wanted to suppress their revolt, and was short of money, so it armed the janjaweed and had them do it. A richer state with more advanced military and administrative capabilities does not get accused of genocide. I’ve never heard anyone suggest Russia’s Chechnya operations constituted genocide, but Moscow’s intent there was identical to Khartoum’s in Darfur. In each case, a regional identity-based revolt was forcefully suppressed. The ethnic group can continue to exist, but its political aspirations are destroyed and its humans are killed indiscriminately to drive home the point. Often, the poorer the state, the more indiscriminate the killing, hence de Waal’s term “counterinsurgency on the cheap.” Nigeria’s 1967-1970 defeat of the Biafran secession was another example, as was Sri Lanka’s crushing of the Tamil Tigers in 2009. There are a lot of these.

In all of the aforementioned cases, it was a minority enclave that was subjugated, but the tactic is also used by minority identity-based regimes to maintain authoritarian control over majoritarian uprisings, as in Bahrain and Syria during the Arab Spring. In fact, such regimes seldom have any other choice, and when the subjugated group is a legitimate demographic threat, the subjugation must be permanent and often legally codified. Apartheid, slavery, and Jim Crow were indefinite subjugation regimes. Minority dictatorships achieve the same ends without admitting what they are doing. Despite the death tolls from all of the above cases, none of this is ethnic cleansing, nor is it genocide. Whether it is a war crime or not is thornier.

There is still another way to “cleanse” unwanted minorities: assimilate them out of existence. Whether this can be done depends on the identity-based ideology of the state. It is easier to convert a whole population to a proselytizing religion or make them all speak your language than it is to make them look more like you if they don’t. Large states including France, Italy, Germany, Russia, and the United States have clearly done sufficient assimilation, each in their own way, to forge a national identity in much of their population that supersedes other identities. The unifying force can be a common enemy, religion, language, or loyalty to tsar/emperor/grand poobah. America, a nation of immigrants, has been unusually good at this. My own backstory is instructive: my ancestors were Poles, Jews, Russians, and Italians, none of whom were considered “American” or “white” or likely spoke much English on arrival. Three generations on, I am clearly American, generally considered “white” and speak only English. I mention these characteristics not to suggest that this is an improvement, but rather to note that I have been largely assimilated into the national, ethnic, and linguistic identities of this country’s demographically and historically dominant group, something my Jewish ancestors never experienced in their much longer sojourn in the Russian Empire. While I still retain certain cultural characteristics—a handful of Yiddish pejoratives and a fondness for pasta, mostly—I am far more American than I am anything else.

However, this practice has limits and dangers. Nationalizing programs are not as easy or accepted as they once were. Any attempt to wipe out a language or religion would be called “cultural genocide” today. In the identity politics hotpot of the United States, even making English the official language would draw fierce condemnation. America has also been far less amenable to assimilating people of color than it was to my forebears. If a group is disallowed such assimilation, or actually prefers to keep its existing customs, a nationalizing program starts to look less like assimilation and more like subjugation, and can often backfire. Also, assimilation granted can be taken away on a dime, as Jews have routinely discovered virtually everywhere. If you find yourself stealing the children of oppressed minorities to educate them in your language and school system, all in the name of a program that shares its name with that of the Borg from Star Trek, perhaps it is time to reconsider your program. Even so, assimilation is not cleansing—no one is forced to leave—so much as it is erasure.

Just as with “negative ethnic cleansing,” there can also be “negative assimilation,” where you block or quarantine a new identity, particularly a religion, to prevent it spreading. Japan once persecuted Christians during the Tokugawa shogunate in the 17th century. Today, Japan is 1% Christian. By comparison, South Korea is nearly 30% Christian, and the Philippines is 86% Roman Catholic and another 8% other Christian denominations. Anyone who has seen the movie “Silence” will understand the brutality such a policy requires, but demographically, even 400 years later, it's clear that it did what it was supposed to do.

5. Ethnic cleansing is part of the system

  1. And now, a brief aside on R2P. My fellow Americans, imagine if France conducted a humanitarian intervention in the United States to stop us from deporting DACA recipients and set up a humanitarian safe zone in, say, New Mexico, where they could seek refuge. Would we like this? No, we wouldn't. Would it be sustainable long-term? No, it wouldn't. And this is why we should be extremely skeptical about the Responsibility to Protect.

  2. Arguably, empires even exacerbated these security dilemmas, because they encouraged large migratory flows within their lands. The Ottoman Empire never planned to collapse, so it never stopped to consider what might happen to Turks who migrated to the lands that became Greece or Bulgaria, or vice versa.

  3. This is because states that have completed identity-based partition and separation can grant minority rights where they could not before, because demography is no longer a weapon to be used against them. A Serb family in Croatia in 1994 was a fundamental threat to Croatian territorial integrity, just by being there. A Serb family in Croatia today is somebody’s neighbor… albeit a neighbor who may face continued discrimination.

  4. This is why I argued in 2015 that Iraq and Syria would be more peaceful and stable if they Balkanized. The ensuing two years of carnage have done little to convince me otherwise.

  5. I should mention a couple of other ways to solve the ethnic security dilemma, but I’m footnoting them because they’re difficult to manage. One is constitutionally paralyzing the government to eliminate it as a political battleground. This has worked, sort of, in Bosnia, Lebanon, and Belgium, though it usually requires some form of geopolitical superstructure or external hegemon to keep it from disintegrating. Additionally, Nigeria’s elegant electoral de-ethnicization system has also been effective, but would be difficult to replicate anywhere else. Imagine if the United States took all identity indicators off its census; increased its number of states by a factor of 12; changed the electoral college to make it impossible to win the presidency without garnering minimum support among multiple regional and racial blocs; and got both Republicans and Democrats to agree to alternate between presenting a white candidate and a non-white candidate every other election cycle. I can’t imagine it either, but this is basically what Nigeria does. It’s an astounding achievement. Even so, building this system required half a century that included a civil war, innumerable military coups, and three separate stabs at civilian governance, and even now the country is plagued by Boko Haram, MEND militants, simmering separatists, intercommunal violence, and spectacular corruption.

  6. Most other examples have been less “return” than “revenge.” Both Krajina Croats and Kosovar Albanians were allowed to go home after being displaced by Serbs... and promptly counter-cleansed the Serbs. So if you support “unconditional right of return” for Palestinians, it might be worth taking a moment to consider what that would look like in practice.

  7. Some will protest at this juncture that the EU has free movement. But free movement does not equal free citizenship. My thoughts on how the EU and the Schengen zone elegantly sidestep ethnonationalism are here.

  8. Hi there, I see you object to this assertion because you read Samantha Power’s “A Problem From Hell: America and the Age of Genocide.” “But America intervened in Bosnia and Kosovo!” you say. Yes, we did. The Bosnia intervention did not stop ethnic cleansing: it codified an ethnic cleansing project already completed over the previous four years, and today nearly all Bosnian Serbs live in Republika Srpska and nearly all other Bosnians live in the rest. The Kosovo intervention did not stop ethnic cleansing either: NATO bombing was in fact what precipitated the Serb decision to ethnically cleanse the Kosovar Albanians—previously they had relied on ethnic subjugation—and NATO victory allowed the Kosovar Albanians to return the favor. By that time, the continued existence of Kosovo as part of Serbia in a multi-ethnic political entity—and that is the point of stopping ethnic cleansing, right?—was inconceivable, and the enclave unilaterally declared independence in 2008.

So if there’s no appreciable demographic difference between Myanmar’s treatment of the Rohingya and America’s treatment of DACA recipients, does this mean that nearly all immigration policy is a crime? It seems that the more internationalist-minded in the foreign affairs set would be ready to take such a step, especially with the current attempt to include ethnic cleansing—which, remember, still isn't actually by itself illegal under international law—under the Responsibility to Protect. I would actually go in quite the opposition direction. Demographic control is integral to maintaining today’s sovereign order. And today’s sovereign order is good. Demographic control is the price we pay for keeping it.

The international sovereign order as we know it today—a system where all states are free and independent within recognized borders—has been integral to the “Long Peace” the world has enjoyed since the Second World War. States do not fear conquest as they once did. Few borders have changed. Most of the ones that have have been due to the internal collapse of empires large and small: the Soviet Union, Yugoslavia, and the British, French, Austro-Hungarian, Russian, Nazi, and Ottoman Empires.

Each of these empires (save the short-lived Nazi one, obviously) was a tapestry of different groups who moved and intermarried and traded freely through it. The first problem with empires, however, was that they knew no borders, and sought to expand in any direction they could. This was inherently unstable, requiring almost continual peripheral warfare, a scramble for colonies, and ultimately imperial overreach and eventual destruction. The second problem with empires was that they suppressed ethnic security dilemmas through force, without getting rid of them. Only fear could keep the local systems in line. And each time an empire fell, its various pieces viciously ethnically cleansed each other into identity-based, territorial homelands. These spasms of violence created many countries we know today.

The nation-state system that was created, by contrast, has proven far more stable than the imperial system it replaced. Few states created in the age of nationalism no longer exist, and while some splinter, since the Second World War few have lost territory to violent acquisition by another state. In the 21st century very few states have actually gone to war with each other at all, or remained in a state of hot war for long. Today’s international news headlines are dire—highlighted by four active famines as of this writing—but when we read them, we forget that even now we live in easily and without question the most peaceful, prosperous, and stable moment in the history of humankind. The international sovereign order, to the extent we maintain it, works. In modern times we only have a few models to organize international affairs. This is by far the best one we have ever come up with. We remove blocks from this Jenga set at our peril.

It works not in spite of, but because a great many of its nation-states are identity-based in character. That is, they are homelands for a specific group of people, and that group constitutes an unassailable demographic majority, and thus has no reason to fear domination by other groups. This is true for most democracies, including nearly all non-principality European states. It is also true for most of East Asia. And while many post-colonial states have a patchwork of groups, particularly in Africa and the Middle East, these have often faced tremendous instability precisely because of their diversity, and furthermore most of them only achieved independence by forcibly removing the bulk of the colonizing humans, often including settler communities that had lived there for generations. Anti-colonialism can be a nationalizing, unifying force... for a while at least.

Solving the ethnic security dilemma through partition and population transfer has consistently led to more stable outcomes and more liberal societies, as in the former Yugoslav states today. This is not because ethnic partition is good, far from it, rather because solutions that don't fix the ethnic security dilemma are worse. Post-imperial states with an ethnic security dilemma that insist on staying together, such as Syria and Iraq, must use ethnic subjugation to maintain that unity. Let us not forget what ethnic subjugation often requires. Cities— Aleppo, Mosul—must be obliterated. Whole demographic groups must be indiscriminately slaughtered or starved into surrender.

Ethnic cleansing, as Lieberman argues, usually arises in times of insecurity, such as the decline or collapse of an imperial authority, and ethnic cleansing exacerbates that insecurity. However, it is difficult to look at the various outcomes in Eastern Europe, the Balkans, and the Middle East today and not conclude that ethnic cleansing, once completed, yields more stable outcomes than the previous imperial system. Either way, it is usually permanent. There is little precedent for a large-scale population displacement episode in which the ethnically cleansed were ever allowed to go home again. The reasons for this are obvious: allowing returns would recreate the ethnic security dilemma that caused the ethnic cleansing in the first place. Try to imagine Yugoslavia being put back together again; the reunification of the Indian Subcontinent; millions of Ashkenazi Jews returning to Poland, this Poland; or the restoration of Greek and Turkish populations forcibly transferred from each others’ countries in the early 1920s. It’s difficult. And that does not bode well for, say, the likelihood the Rohingya ever going back to Myanmar. It also throws into serious doubt the viability of any “unconditional right of return” for the Palestinian diaspora.

Hang on, actually, there is one example that comes to mind of a successful diaspora return following systematic displacement: the conquest of Rwanda by an invading Tutsi rebel army from Uganda in 1994. Costs of this successful return include the Rwanda Genocide and the Second Congo War, the destabilization of much of the African subcontinent and the deaths of several million people. Be careful what you wish for.

Ethnic engineering and its consequences are built into the international sovereign order, and the power of the state to control its own demography on identity grounds is manifest and in fact integral to the maintenance of that order. We do not and cannot seriously object to ethnic swamping or short-term ethnic subjugation programs, nor to blood-based citizenship regimes or discriminatory immigration quota systems. If you accept that Tibet is part of China, the question of which citizens of China live where is a purely internal matter, no more a question of international relations than gentrification in American cities. And no state has ever responded peacefully to a violent ethnic secessionist bid, nor could one reasonably be expected to. We cannot maintain an international sovereign order where states (or free movement zones such as the EU) cannot independently control their immigration policy or remove those without legal right to residency. It would be wildly impractical and a stability-shattering violation of sovereign rights to attempt such a thing.

But even if you could do it, such a policy would be disastrous, because in practice, almost any noticeable level of demographic change away from the dominant group in a country creates instability, regardless of the location. In the two 20th century instances in which it proved sufficient to tip the country’s demographic balance—Lebanon in the 1970s, Côte d’Ivoire in the 1990s—civil war soon followed. Even relatively small Muslim communities in Europe, buttressed by the arrival of a million Syrian refugees to a continent of three quarters of a billion people, have fed a right wing electoral wave, Brexit, and xenophobia the continent once thought banished. While few nationalist parties have actually seized power so far, they’ve changed and effectively won the debate on immigration. Marine Le Pen made the runoff in the 2017 French elections. The Muslim population of France is estimated at 8.8%. How will the French vote if it doubles? You can wish that people didn’t behave this way collectively, but it appears you can’t ask them to stop. They won't listen to you.

If we want sovereign and independent states—and, given the alternative, we should—we must accept a modicum of sovereign demographic control, not because we like it, but because it is going to happen regardless. If states can balance their demography, the demagogues will remain on the political fringe. If states are disallowed from this practice, the demagogues will feed on the resultant ethnic security dilemma and take over, and then they will restore demographic dominance through odious means. The outcome is the same, except in the latter case, you have demagogues in power.

Our foremost goal, rather than rectifying past ethnic cleansing or creating a utopian post-tribal world, should be to prevent ethnic security dilemmas. All other goals must be secondary. Ethnic security dilemmas make war criminals of us all. They destroy all capacity for peaceful coexistence or harmony within a polity. They incentivize an ethnonational program in which the state cleanses the unwanted or engineers them into oblivion. And once underway, this program cannot be stopped or undone. No one can restore the vanished communities of Salonica. No one, save perhaps an army of smugglers, can stop Europe from sealing its borders. No one can stop a determined American administration from limiting refugee flows and deporting overwhelmingly non-white immigrants by the millions. It is doubtful anyone can stop Myanmar’s operation either, for even if the campaign ended now, how could the Rohingya ever feel safe in Myanmar again? And if Myanmar’s army can run out a million Rohingya by torching their villages and mining the ground behind them, the ship has sailed.