On January 27, U.S. President Donald Trump signed an executive order temporarily banning migrants from seven Muslim countries—Iran, Iraq, Libya, Somalia, Syria, Sudan, and Yemen—and permanently halving the number of refugees admitted to the United States. A February 9 decision by the U.S. Court of Appeals for the Ninth Circuit  blocked its implementation pending a decision from the Supreme Court. But although the legality of the order is important, so too is the question of whether it will achieve its goal to “protect the American people from terrorist attacks by foreign nationals admitted to the United States.” The answer is a resounding “No.” A thorough look at terrorist attacks by foreigners on U.S. soil demonstrates that Trump’s order will do almost nothing to improve national security but will impose a great cost on Americans. 


Last September, I analyzed the risk of terrorism by foreign-born perpetrators in the United States from 1975 until the end of 2015. Over that period, a total of 154 foreign-born terrorists murdered 3,024 people in 15 domestic terrorist attacks. That means the annual chance of dying in an attack by foreign-born terrorists on U.S. soil was one in 3.6 million. The risk varies tremendously based on the type of visa used to enter the country. Foreigners who entered on the tourist B visa, including 18 of the 19 hijackers on 9/11, were the deadliest, killing 2,834 people over that 41-year period. Terrorists who came in on the Visa Waiver Program, which applies to nationals from developed nations, killed zero people, although they did make a few attempts.

Trump’s executive order didn’t look at the most dangerous visa categories in recent history but rather singled out all migrants from seven Muslim countries for a temporary 90-day ban and focused the administration’s greatest scrutiny on the refugee program. But despite all of the attention paid to refugees, the chance of being murdered in a terrorist attack by a refugee on U.S. soil is only one in 3.64 billion per year. Only three such murders have ever occurred—all three instances were in the 1970s, and the attackers all came from Cuba.

The chance of being murdered in a terrorist attack by a refugee on U.S. soil is only one in 3.64 billion per year.

For a sense of proportion, the chance of being murdered in a normal homicide in the United States during the same 1975–2015 time frame was one in 14,219 per year, which is an astronomical 255,906 times greater than the chance of dying in a terrorist attack by a refugee on U.S. soil. But those sobering comparisons didn’t stop Trump from temporarily suspending the refugee program for 120 days, slashing its annual cap from 85,000 admissions in 2016 to only 50,000 in 2017, and barring all Syrians. The United States allowed in an average of 79,329 refugees per year from 1975 through the end of 2015. The new cap of 50,000 represents a 37 percent cut in that annual refugee intake.  

Protestors at John F. Kennedy airport demonstrate against U.S. President Donald Trump's executive order banning travel from seven Muslim-majority countries, January 2017.
Joe Penney / Reuters

Migrants from the seven banned countries have also killed zero Americans in terrorist attacks on U.S. soil. However, 17 nationals from those countries did commit or were convicted of attempting an attack here. This list includes a goofy plot hatched by five students at Mankato State University to kidnap Minnesota Governor Al Quie in the winter of 1979.  

The administration’s recent bizarre claim that terrorist attacks are underreported could be prompted by the public’s extreme fear of personally being harmed in a terrorist attack. A shocking survey by the Public Religion Research Institute found that 47 percent of Americans are “somewhat worried” or “very worried” that they or someone in their family will be a victim of terrorism. This degree of fear, far out of proportion to the actual threat, can be partially blamed on a universal mental shortcut called the “availability heuristic.” Our minds use recent or dramatic events to understand ideas or concepts, which leads us to form false explanations by biasing us toward newer or more sensational information. With regard to terrorism, it is easy to immediately remember 9/11, recent attacks in Europe, or the Orlando nightclub shooting. All the attacks that didn’t happen do not register as prominently, nor does the fact that few people know someone personally who was killed by terrorists. Thus, terrorism seems much more dangerous than it really is. 


Regardless of the scale of the real or imagined foreign-born terrorist threat, the U.S. government has an important role in screening out terrorists who attempt to enter the United States. But a rational terrorist screening should be based on a few facts. 

First, policymakers should recognize that the United States already successfully screens out terrorists, which is why so few Americans have been killed on U.S. soil in terror attacks. Another possibility is that the low numbers of attacks here are the result of a much smaller foreign terrorist threat than most policymakers imagine. For instance, it is hard to take seriously pronouncements by officials such as South Carolina Senator Lindsey Graham, who said that Islamic State (also known as ISIS) terrorists are an “existential threat to the homeland,” when only 22 Americans have been killed by foreign-born terrorists on U.S. soil since ISIS became active in 2013. Every one of those deaths is a tragedy, but they do not amount to an existential threat. It makes no sense to argue, as Trump has, that the foreign terrorist threat is huge andthat there is “no system to vet them,” unless the entire value of the vetting is deterrence. 

Second, no security procedures are going to be completely effective at stopping terrorists. Steven Camarota, director of research at the Center for Immigration Studies, correctly noted that “in a nation as large as the United States, it is impossible to prevent terrorists from entering the country 100 percent of the time.” A committed terrorist or foreigner who decides to become a terrorist after arriving can be only partly defended against. Regardless of any actions taken today, there will be another day when an American will die in a terrorist attack on U.S. soil committed by a foreigner. And terrorist attacks committed by the U.S.-born children of immigrants, such as the 2016 murder of 49 people by Omar Mateen at the Pulse nightclub in Orlando, are dramatic but rare. American Muslims are well assimilated and have opinions very close to those held by the American mainstream, especially when compared with their coreligionists in Europe.  

Third, the benefits of additional security should be weighed against the costs of that security. There is a point at which additional security to screen out foreign terrorists will actually result in more deaths, as every dollar spent on screening refugees is a dollar that could have been spent elsewhere on reducing an even greaterrisk, such as homicide or domestic terrorism. The government has limited security resources, so it should spend them in ways that minimize violent deaths of all kinds—not just those caused by foreign terrorists. 

These statistics should inform U.S. policy on terrorism. Government regulators frequently estimate how much it will cost to save a single life—often referred to as a “statistical life”—under new safety rules that they propose. The key insight behind that estimate is that human life is certainly very precious, but not infinitely so. Indeed, everything people do that slightly increases their chances of dying, such as driving a car, would be unthinkable if they placed an infinite value on their own lives.

Not a single life would have been saved had Trump’s executive order been put in place 41 years ago.

The tricky part is figuring out how much that statistical life is actually worth. Many security experts place a high cost of $15 million on each statistical life. Using this estimate, a new terrorism prevention rule would be rational if the value of statistical lives saved by the rule were at least as high as the cost it imposed. If Trump’s refugee restrictions reduce the already low chance of dying in a refugee terrorist attack by a further 50 percent, to about one in 5.5 billion per year, then it would cost Americans about $525.5 million per life saved in lost economic output that would have accrued to U.S. natives—about 35 times as great as the benefit. But the larger cost would be the 510,000 fewer refugees resettled in the United States than would have been without the restrictions. Those tremendous costs would buy the United States one fewer murder committed by a refugee-terrorist over the next 17 years

These calculations aside, however, not a single life would have been saved had Trump’s executive order been put in place 41 years ago. Just 17 terrorists from the seven Muslim countries included in Trump’s travel ban have been convicted of planning or carrying out terrorist attacks on U.S. soil from 1975 to the end of 2015, and none resulted in fatalities. Likewise, refugees are less likely to kill Americans in terrorist attacks than are foreign-born terrorists who entered on most other types of visas. Trump’s executive order is not a rational response to the actual threat of foreign-born terrorism on U.S. soil. It provides almost no benefit to national security, yet it comes at tremendous cost. 

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now