The Trump administration is considering American intervention on the side of Israel in its war against Iran. My colleagues have persuasively written that US intervention is unjustified. But American involvement is increasingly likely, according to the betting markets and the intensity of bellicose words exchanged between President Trump and Iranian politicians. Like previous US military interventions, there is practically no public debate about the second- or third-order effects of war with Iran. One of the consequences of such a war could be the largest refugee crisis since World War II if the Iranian government collapses and a protracted military conflict or civil war ensues.

The Syrian civil war produced a refugee crisis that was a humanitarian, economic, social, and political disaster. That civil war started in March 2011, and by the end of 2012, 729,011 official refugees had fled Syria for neighboring countries and further abroad. The number of refugees then steadily climbed to a peak of over 6.8 million in 2021 before coming down to just under 6 million in 2024. And those numbers don’t include the 7.4 million internally displaced people.

At its peak, a quarter of Syrians were living outside of the country as refugees, and another quarter were internally displaced persons. The chaotic repercussions of such a massive refugee population disrupted the Middle East and Europe. Exaggerated fears of Syrian terrorism in the US likely helped Donald Trump win his first election in 2016. And that was for a country that descended into civil war with a population just shy of 23 million.

A governmental collapse, civil war, and protracted conflict in Iran could send a larger, more destructive, and more destabilizing refugee flow to neighboring countries and the West. Iran’s current population is over 92 million—almost exactly four times the size of Syria when it collapsed and sent a quarter of its population abroad as refugees. A similar refugee outflow from Iran would be about 23.4 million people and increase the worldwide refugee population by about 76 percent. That doesn’t include the roughly 3.5 million mostly Afghan refugees currently in Iran who would also likely skedaddle.

Besides their larger population, Iranians have more resources to flee and deeper cultural connections with the West than Syrians. Iranians today have a GDP per capita (PPP-adjusted) about three times higher than Syria did at the beginning of its civil war. About 900,000 Syrians were living abroad at the start of the Syrian civil war, about 4 percent of Syria’s population. As of a few years ago, there were about 4.5 million Iranian emigrants, equal to a similar percentage of Iran’s population.

Anecdotally, Iranians seem more familiar with Western culture for historical, linguistic, and other reasons. For instance, Farsi, the most common language spoken in Iran, is an Indo-European language that has more grammatically (not vocabulary) in common with English and German than Arabic, which is an Afro-Asiatic language. Iranians also seem more liberal than Syrians, which is why they’ve been rioting against their Islamic government for years. Still, it’s hard to be sure because their theocracy resulted from a homegrown movement that eventually took over and, therefore, must represent some significant share of the population. If it didn’t, that government would be unable to stay in power.

Iranian refugees wouldn’t just go to the West, although many would succeed. Those with fewer resources or fleeing more desperate circumstances would cross to neighboring countries. Almost 80 percent of Iran’s population lives in cities, and nine of the ten largest cities are in the western part of the country. Iran’s population is far more urban than Syria’s at the start of the conflict, and the collapse of public services in cities can produce more dire circumstances more quickly than in the countryside.

Refugees would most likely travel west into Azerbaijan and Iraq, two other Shi’a-majority countries, and relatively safe and prosperous Turkey. The politics in Iraq are opaque, but it’s not unbelievable to be concerned that a few million Iranian Shi’a refugees could disrupt the political system of a country recovering from a recent sectarian civil war between Sunnis and Shi’a. Turkey’s government welcomed many Syrian Sunni refugees, but it’s unclear whether it would be so welcoming to Iranian Shi’as.

A refugee surge from Iran equal to about a quarter of its population, or 23.4 million people, is likely the worst-case possible outcome for refugee flows. Only a protracted conflict, state collapse, the destruction of public services in Iran’s urban areas, and civil war on a Syrian scale would likely force that many people to leave—but Iranians have already started fleeing. Northern border crossings to Turkey, Azerbaijan, and Armenia are overwhelmed with Iranians trying to leave. The potential for nuclear disasters or the fear of radiation, whether well-founded or not, plus Iran’s highly urbanized population dependent on fragile services, could push refugee numbers up even faster than in Syria or past the worst-case scenario proposed above. Even refugee flows of one million Iranians by the end of this year and 2.3 million by the end of next year would also be a disaster.

Neighboring countries should begin to plan for the possibility of a refugee inflow from Iran. First, the Gulf States should take many in as temporary workers or just rich migrants for the duration of whatever conflict occurs in Iran. The Gulf States did this with Syrians during the civil war. Second, the United States, Canada, and other Western countries with large Iranian immigrant populations should fast-track family reunification permanent residency petitions and begin to rev up refugee admission and private sponsorship systems. Of course, the US won’t do this because it recently placed restrictions on Iranian immigrants and nonimmigrants, although immediate relatives and Iranian immigrants fleeing ethnic and religious persecution are allowed.

Third, neighboring countries should learn lessons from Jordan’s successful integration of Palestinian refugees in the 1990s, which increased that country’s population by about 10 percent in a single year. The main lesson is that Jordan liberalized its economy to accommodate such a large influx of new workers, and neighboring countries should start planning to do that to shrink the fiscal cost, reduce the potential of political destabilization caused by economic uncertainty, and mitigate the potential humanitarian catastrophe. Those countries should liberalize anyway. But preparing plans to do so in case of a refugee crisis is my nod to political practicality.

The Syrian civil war lasted almost 14 years and caused about a quarter of all Syrians to leave their homeland, resulting in tremendous humanitarian, economic, social, and political upheavals in the Middle East, Europe, and the United States. If the Iranian government collapses or descends into civil war because of US military intervention, the resulting flows of Iranian refugees could be many times larger and much more destructive than the Syrian refugee crisis. It’s almost unimaginable how disastrous it would be if chaos lasted for even a significant fraction of the duration of the Syrian civil war.

Supporters of bombing Iran should consider the possible second- and third-order consequences, such as the largest potential refugee flows in history, and whether attacking Iran is worth that risk. The United States, European countries, and other countries bordering Iran should begin to plan for such an outcome if Iran descends into chaos under a military onslaught.