What is the relationship between immigration and terrorism?
Balancing global utility and national security over the long term.
To put my cards on the table, I’m an advocate of high levels of immigration. I’m saying that upfront so there is no confusion as to whether I’m trying to place myself in some kind of faux-centrist between those who support open borders, and immigration restrictionists. The reason I think the latter are wrong is mostly laid out in my post on the economic impacts of immigration, and this post is one of the two reasons why I think the former requires some limits.
Is there a link between immigration and terrorism?
Choi and Salehyan (2013) look at the link between hosting refugees and domestic/international terrorism. The simple correlation coefficient between the two (see graph above) is 0.46 – which isn’t menial. But simple correlations aren’t the be all and end all. They undertake a negative binomial regression, and the results are shown below.
What this means is that when a country experiences an increase of one and two standard deviations of refugees above their mean value, the frequency of terrorism incidents increases by 18% and 40%, respectively. The limitation of this study, however, is that it looks at refugee flows from 1970 to 2007. With the odd exception, the largest refugee outflows in that period have not be to the West. This matters because one of the main casual mechanisms they provide for the findings above is that attacking refugee camps ‘provides opportunities for looting [of aid supplies] and theft by violent groups’. But that isn’t to explain away all of the effect: the effect is weakened in OECD countries but it still persists.
As show above, although its significance levels weaken in the OECD countries, as compared to the non-OECD countries, the refugee variable remains statistically significant. This isn’t a unique finding: Milton et al (2013) find the same result using a different method (i.e., they utilise a set of directed-dyad statistical models to assess the relationship between refugees flows and terrorism) and different data sets (using the ITERATE dataset and then an alternative dataset). The mechanisms that are put forward in this are far more relevant to Western refugee populations (even though they may not be in camps):
…it is also likely that the refugee camp is isolated and segregated from society at large, thus making the actions inside the camp hidden from the authorities. This privacy creates an opportunity for terrorist cell development, terrorist recruitment, and weapons laundering and distribution to go undetected.
Their results (shown in Table 1 above) show there is a positive relationship between refugee inflows and transnational terrorism. Indeed, moving from the 25th percentile (in terms of the volume of refugee inflows) to the 75th percentile, leads to a 15% increase in probability of transnational terrorism (0.28 to 0.43). But look again at Table 1 above and specifically at “capability ratio”. The capability ratio measures the relative strength of two dyads: the host state of the refugee and the destination state. Where the capability ratio is high (likely because the destination state is stronger), there appears to be a negative association.
This may be an extension of the OECD result from the previous study: stronger states are simply better at monitoring and incapacitating terrorists. Bove and Bohmet (2016) is a slightly different study. The studies discussed above look at refugees, and as explained the findings are about refugee inflows which are more directly related to countries outside the West. By contrast, Bove and Bohmet look at migrants, not refugees. Here is the mechanism they discuss:
... social bonds play the most important role as they provide “mutual emotional and social support, development of a common identity, and encouragement to adopt a new faith” (Sageman, 2004, p.135). The potential pool of terrorists is, in fact, formed by clusters of e.g., friends or worshippers, who are connected via strong ties... the presence a pre-existing social framework is a somewhat necessary requirement for joining, forming, or engaging with terrorist groups; sometimes, these networks exist long before any members engage in terrorist activities... migration flows and diaspora communities provide those linkages, nodes, and pre-existing social networks
The study has another measure which looks at all migrants (Migrant Inflow (In)) and here are the results for that measure:
Migrant Inflows (ln) is indeed negatively signed and statistically significant at the 5 percent level across Models 1-5. As a result, migration as such – independent from or not weighted by the terror level in the country of origin – actually leads to a decrease in the number of terrorist attacks by 0.5-0.6 percent when the number of migrants coming into a country is raised by 10 percent.
There is, in essence, a conflict decreasing effect of (aggregate) immigration on terrorism. This doesn’t displace the general finding above though: another measure they look at is ‘Wy: Migrant Inflow’ which measures migrants from terror-prone states. Focussing on that measure, the results are the same:
The empirical results demonstrate that immigrants are indeed a vehicle for terrorism to travel from one country to another, i.e., the level of terrorism “at home” increases with a larger number of immigrants from countries of origin where terrorism has been present... the mean of the number of terrorist attacks by 8 percent and 17 percent, respectively, when raising Wy: Migrant Inflow by one unit... terror events in one country travel to another state via the inflow of migrants
This isn’t cherry picking, its consistent and dispersed throughout the literature: Dreher, Gassebner and Schaudt (2020) examine migration from 183 origin to 20 OECD countries in a dyadic setting and find a larger number of foreigners leads to more terrorist activity in the host country.
So, TLDR: yes, there is a relationship between immigration and terrorism.
“But terrorism is so rare!”
It is tempting for liberals to accept the results above, and then say “but terrorism is very rare” without elaborating further. On that front, well, it is true that terrorism is very rare. The following table from the Cato Institute shows the low probabilities involved in dying of immigrant-terrorism in the United States.
There are similarly low levels of deaths caused in the UK by terrorism:
I don’t find this line of argument decisive for two reasons. First, I fear this doesn’t pass the “care about large, near-existential risks, even low probability test”. Scott Alexander explained this by reference to earthquakes in a post back in 2016:
Suppose I’m trying to make an argument that earthquakes are totally not a problem for Haiti at all, that there’s no need to invest in earthquake preparedness, and that Haitian people who worry about earthquakes are stupid. I make a graph showing that since January 13, 2010, fewer Haitians have died per year from earthquake-related causes than from crazy furniture-related mishaps. This is totally 100% true. Look at those stupid Haitians, worrying about something that on average never hurts anybody!
(the Haitian earthquake of January 12, 2010 killed about 100,000 people)
I’m sure there are a zillion small Richter 1.0 and Richter 2.0 earthquakes in Haiti all the time. I’m sure our monitoring interval of January 13, 2010 to present picked up lots of these and correctly noted that they don’t kill anybody. The only Haitian earthquakes anyone needs to worry about are the outliers.
Second, and relatedly, I think the impact of terrorism must be understood well beyond the immediate and direct fatalities. The impact of terrorism must be understood as an attack on the liberal, market-based constitutional order that requires a commensurately large dedication of resources. Terrorism is not just a means, as Phillip Bobbitt persuasively writes, it is itself the end for various nefarious groups.
How does terrorism affect the case for immigration?
Given the large global utility, and the significant economic benefits, which arise from high levels of immigration, we should consider whether there is a way of minimising the impact of terrorism arising from migration. Rather than seeking to deny, or downplay, the impact of immigration, or immediately banning all Muslims, we need to accept there are trade offs.
First, we should be grateful to our security services for the work they do to keep the rare events even rarer than they otherwise would be. In the U.S., the overwhelming majority of terrorism is thwarted:
In the UK, between 2013 and 2017, 13 attacks were prevented; since 2017, 31 late stage attacks have been foiled. This also gives me a chance to talk about one of my favourite studies: Jesse Lehrke & Rahel Schomaker look at the effectiveness of counterterrorism strategies in the fight against terrorism from 2001 to 2011. They look at "defences" (measured by legislation which assisted with things like surveillance). There is a large and significant effect of such measures reducing terrorism.
It’s worth noting that the 9/11 attacks were carried by those on temporary tourist visas, not migrants. The case for strengthening tourist vetting and controls is stronger than the case for reducing immigration. That said, those who support high levels of immigration would do well to ensure they are not undermining efforts for the police, nor security services to carry out surveillance, monitoring and arrests; those decry the impacts of immigration, would do well not support policies which cut funding for the security services.
Second, we need to be mindful that those immigrants who turn to terror in the West, are unlikely to have been angels in their origin states. This is important for two reasons. As we saw above, there is a reduced impact of terrorism in strong, particularly rich and/or Western, states. In addition, given the terrorism-reducing impact of Western states, it is likely that immigration to the West reduces the overall level of global violence. This is a natural corollary of the findings above, but we have specific evidence for the proposition as well.
One particularly neat study design is looking at how migration impacts the effect of natural resources on the onset of civil conflict, finding a significant reduction. Another neat study design looks at the impact of refugee returnees finding that when they return violence increases (a finding which contrasts with the fact that a civil war has never been caused by immigrants in a developed country). Movement of flesh and blood persons is not even the only mechanism here, we know that remittances, and potentially diffusion of norms, also reduce civil conflict. Financing of terrorism deserves a post of its own, but in addition to that general finding, it suffices to note that Piazza and LaFree (2019) find that Islamist terrorist organizations are more likely to refrain from high-casualty terrorism when they are dependent upon diaspora backing.
Third, even a crude cost benefit analysis giving each life a $15m cost finds that an immigration moratorium would have to prevent 2,333 deaths annually to be worth the economic cost. As I said above, I don’t think we should only look at lives lost as the decision argument, but its nonetheless a consideration in balancing national security and global utility. I glibly referred to global utility above, but its worth adding adding meat to this point. There are likely to be over 3.3 million Muslims in the UK. MI5 is currently watching 43,000 people suspected of terrorism related activities. Let’s assume they are all Muslims, and then lets double the figure of potential terrorists. Even that is roughly 2.5%. The utility gains to the remaining 97.5% - millions of people - must be part of the consideration here.
Let’s be upfront about what we are saying: immigration will increase the chance of terrorism in the West, but there are trade offs. Immigration is likely to give signifiant utility gains to migrants themselves, positive economic benefits, but also reduce the global level of serious conflict and violence. The less we hide away from the trade off, the more we can be honest about what state resources need to be prioritised. Milton Friedman famously said that you cannot have open borders with a welfare state; in light of the above, what he should have said was that you cannot have open borders without a strong security state.