Olivia Enos, David Inserra, and Joshua Meservey of the Heritage Foundation published an interesting Backgrounder last week about the U.S. refugee program. We agree with many, though not all, of its conclusions and think that it serves as a wonderful example of policy experts grappling with a difficult policy question in a nuanced and thoughtful way – two characteristics often lacking in Washington, D.C.

However, the Backgrounder’s claim that 61 refugees were convicted of Islamist “terrorism‐​related” offenses since 2002 has earned a lot of attention from the media. David Inserra was kind enough to send us a complete list of the refugee terrorists he and his colleagues counted. Here are the facts about these 61 people:

• None of these refugees killed anyone in a terrorist attack on U.S.-soil.
• Only five (8 percent) were refugees who attempted or planned an attack on U.S. soil. The other 56 (92 percent) of the list were either not refugees or not terrorists targeting U.S. soil.
• At most 50 were actual refugees who may have committed terrorism offenses, out of the 2.1 million refugees admitted since 1989, which is the earliest year that anybody on the list entered as a refugee. At least eleven (18 percent) of the refugee terrorists reported by Heritage were either not actually refugees or not convicted of terrorism offenses.
• Only five (8 percent) entered as refugees since 2008.
• Only five (8 percent) were likely refugee security vetting failures who entered as adults or older teenagers and committed an offense soon after entering.
• The 50 refugees represent just five of the 124 nationalities of refugees admitted since 2002 (4 percent). Three‐​quarters of the refugees who committed a terrorism offense came from a single nation.

The security threat from refugees is minuscule, concentrated among a few Somalis, and has little to do with vetting.

The Non‐​Refugees and Non‐​Terrorists

The Backgrounder’s use of terrorism‐​related offenses is problematic as it is not synonymous with actual direct or indirect support of terrorism. There is no definition of a terrorism‐​related offense in U.S. statutes but there is a broad working GAO definition: that it relates to “terrorism, homeland security, and law enforcement, as well as other information.” As far as we can tell, the term terrorism‐​related is used to describe a conviction for any offense that results from a terrorism investigation – even if it is for crimes that bear no relation to terrorism such as buying stolen cereal. David Inserra told us that “Our [Heritage’s] inclusion criteria wasn’t based on convictions for terrorism offenses because people could be involved in that sort of activity and not ever be convicted. We were trying to find the happy medium between overly‐​restrictive and too loose definitions.” Thus, we are working with different definitions and the reader should keep that in mind.

Six individuals on the Heritage list were not convicted of terrorism offenses. The government dismissed its complaint against Al‐​Hazmah Mohammed Jawad. Aws Mohammed Younis Al‐​Jayab and Ali Mohammed Al Mosaleh were charged with making false statements. Abdi Mahdi Hussein was convicted of failure to follow financial reporting requirements and was “not charged with any terrorism offense and was not alleged to have knowingly been involved in terrorism activities,” according to the FBI. Yusra Ismail was charged with stealing a U.S. passport—not terrorism—and Saynab Hussein was convicted of perjury.

At least five individuals on the Heritage list did not enter under the U.S. refugee program. One of Heritage’s errors is due to an error that Alex Nowrasteh made in his 2016 Policy Analysis (corrections will be made in future editions) that listed Nuradin Abdi as a refugee instead of an asylum recipient. Nima Ali Yusuf was also an asylum recipient. Harlem Suarez was a Cuban national would have received asylum by virtue of the Cuban Adjustment Act, not through the refugee program. Mahmoud Elhassan was a Sudanese immigrant whose mother sponsored him for a green card. Jasminka Ramic was a Bosnian immigrant who “immigrated to the United States as a legal permanent resident.” There may be others who should not be included in Heritage’s list as well as unknown individuals who should be counted.

This means that only 50 of the 61 people on the Heritage list were refugees who may have committed actual terrorism offenses. At the time of the publication of the Heritage Background, charges against 12 of the 50 remaining refugees were still pending. As Al‐​Hazmah Mohammed Jawad’s case shows, we cannot assume guilt based on the charges alone. But for the purpose of this post, we will assume these 12 cases will all end in convictions for terrorism offenses.

Few Refugees Plan Attacks on U.S. Soil—and Even Fewer Succeed

The text of President Trump’s original executive order cutting the refugee program justified itself as an attempt to “protect the American people from terrorist attacks by foreign nationals admitted to the United States.” It is remarkable, then, how few of these refugees actually planned or committed an attack on U.S. soil. As Heritage writes, “Five refugees successfully committed a terrorist attack inside the U.S., two of which occurred after the U.S. reformed the program post‑9/​11.” Three of the five refugees planned to commit an attack but were unable to follow through. Mohamed Osman Mohamud was caught in an FBI sting where he attempted to detonate a fake bomb. Similarly, Fazliddin Kurbanov’s bomb‐​making activities were “closely monitored by federal agents during the investigation and no terrorist attack occurred,” the Justice Department stated. Hysen Sherifi plotted to attack a military base but the FBI broke up his ring before they could do any damage. Abdul Artan and Dahir Adan were two refugee terrorists who actually committed their attacks but fortunately, they failed to get anybody killed other than themselves.

No refugee terrorist has killed a single person in a terrorist attack on U.S. soil since 1976—four years before the creation of the modern U.S. refugee program in 1980. From 1975 through the end of 2015, the annual chance of being murdered on U.S. soil in a terrorist attack committed by a refugee was an astronomical 1 in 3.64 billion per year.

Vetting Failures are Very Uncommon in the Refugee Program

The Heritage Backgrounder’s findings show just how uncommon vetting failures actually are. Heritage notes that the cases are mostly concentrated among “refugees who were resettled at a young age—often known as the ‘1.5 generation’—or the children of resettled refugees, who radicalize and then commit terrorist offenses.” That indicates that there is an assimilation problem for some refugees and not a vetting problem from the refugee program as a whole. We have entry dates for 47 of the 50 refugees, and 29 of them (62 percent) were juveniles at the time of their entry to the United States.

Additionally, many of the adult refugee terrorists were most likely not vetting failures who came intending to commit terrorism in the United States. We can’t read anyone’s mind, but the fact that the vast majority appeared to have shown no interest in an offense long after their arrival is probably about the best evidence we could get. Of the refugees for whom we have entry dates, 78 percent of them lived in the country for more than seven years before they committed an offense. Sixty‐​one percent lived here for more than a decade.

The seven‐​year point seemed particularly significant because the longest period between the entry of a refugee and his offense in a confirmed refugee vetting failure case—in which the government had evidence of radicalization prior to entry—was six years (Abdinassir Mohamud Ibrahim). If we assume that all offenders who entered as adults or older teenagers and committed their offense before their 7^th year in the United States are vetting failures—regardless of whether there is evidence of that—we can get a rough estimate of the number of vetting failures.

By this measure, it appears likely that just five on the Heritage list were vetting failures: Abdinassir Mohamud Ibrahim lied about his clan membership on his refugee application; Waad Ramadan Alwan and Mohanad Shareef Hammadi fought with insurgents in Iraq; Fazliddin Kurbanov plotted an attack just four years after entering as an adult; and Abdul Razak Ali Artan committed an attack just two years after entry. Only Kurbanov and Artan intended to carry out an attack on U.S. soil, out of the approximately 2.1 million refugees who have entered the United States since 1989.

Nationalities of Refugee Terrorists

Since 2002, the United States has resettled refugees from 124 nations. Yet only five nationalities are represented on the Heritage list: Somalia (37), Bosnia (7), Iraq (2), Uzbekistan (2), and Kosovo (1). However, President Donald Trump’s executive order shuts down refugee resettlement for all countries. Notably, Syrian refugees are absent from this list entirely, despite Donald Trump’s overwrought rhetoric on the subject—as are five of the six nationalities targeted by the order: Libya, Yemen, Sudan, Syria, and Iran. Somalia is the only country that has sent refugee terrorism offenders to the United States and that is specifically listed on President Trump’s executive order.

Somalis constitute only about 10 percent of the refugee program. Even Somali refugees have not proved dangerous to Americans. Almost all of the Somali terrorism offenders raised money for or went to join terrorist organizations abroad. Two did commit terrorist attacks in the United States but fortunately failed to kill anyone. Nobody from Somalia has killed anybody on U.S. soil in an act of terrorism for as far back as we have data.

Conclusion

The Heritage Backgrounder makes a valuable and interesting contribution to the debate over the future of the American refugee program. As a side effect, it shows just how small and manageable the refugee terrorist threat is.