Columbia University

Faculty of Arts and Sciences

Does Terrorism Work?

An Interview with Political Scientist Page Fortna

June 29, 2015

In her recent paper, Do Terrorists Win? Rebels' Use of Terrorism and Civil War Outcomes,” Harold Brown Professor of U.S. Foreign and Security Policy and Chair of Political Science Page Fortna questions the effectiveness of terrorism. The paper, published last month in International Organization, has generated an important dialogue about the seemingly simple question at the center of Fortna’s research: does terrorism work? In her paper, she argues that the political agendas of terrorist organizations consist of goals that they are unlikely to achieve. She defines a terrorist group’s ability to fully achieve its political goals as “success,” which she measures in relation to civil war outcomes. This idea of measuring the “success” of terrorism has triggered an argument about what that success really means. The following is a Q&A with the author about the paper as well as the political conversation it has sparked:


In your paper, you write that, contrary to popular belief, the majority of terrorist violence today actually takes place among groups fighting civil wars. This type of terrorism is the most common, and yet it receives far less media attention than mass attacks like the one that took place in Paris earlier this year. Can you speak to this focus on the few “big” terrorist attacks while the most common terrorist activity remains almost unheard of?

Page Fortna: First, I should clarify that I don't claim that most terrorism takes place in civil wars—only that most terrorism is domestic rather than transnational. A lot of that domestic terrorism takes place in civil wars, but some of it takes place in lower level domestic conflicts that don't meet our thresholds for calling something a civil war.

For this study, I'm looking at only large-scale civil wars, so I definitely don't want to claim that I'm capturing most terrorism, though in other work I'm looking at lower levels of civil conflict as well. But that said, I agree that there has been more focus on the type of terrorism that is least common—transnational attacks.

I think this is true for at least two reasons. One is that until recently the best data on terrorism focused on transnational terrorism rather than domestic terrorism. The second reason (perhaps the cause of the first) is that most terrorism scholarship happens in the West, and the threat of terrorism in the West was, historically, perceived to be transnational. We understandably focus on that which we think threatens us the most, but from the perspective of humanity as a whole, domestic terrorism is the much larger problem.

As the New York Times reported recently, there is now more appreciation of the fact that even in the US, the threat of domestic terrorism (e.g. white supremacist extremism) is as big if not bigger than transnational (e.g. Islamist extremism) terrorism. Hopefully, that will lead to a greater focus on domestic terrorism as well.

You mentioned that some researchers believe that terrorist violence gives rebel groups the “power to hurt,” which can often persuade governments to respond to their threats and cooperate with them. It is interesting to consider how terrorists gained that power since they are ultimately unlikely to achieve the goals laid out in their threats. Do you think that this is the result of the mainstream focus on acts of mass terrorism?

PF: It is true that terrorism is a cheap and relatively easy way to inflict a lot of hurt. Where I differ with those researchers who think this makes terrorism effective is whether that "hurt" translates into governments responding in ways that get opposition groups what they want. As I argue in the paper, terrorism is more likely to backfire by making the government less likely to negotiate, by undercutting political support for groups that use terrorism, and by rallying support for governments taking a hard line against them. One important distinction that probably accounts for the different findings in the scholarly research is what counts as "terrorism."

For example, Thomas' article (which uses the phrase "power to hurt") is looking at the effect of lots of different types of violence against civilians combined, whereas I am looking more narrowly at what we quintessentially think of as terrorism—that is, deliberately indiscriminate violence against civilians, where the whole point is the random nature of the violence. Unfortunately, I think some types of violence against civilians can be effective—for example, targeting civilians to get them to collaborate with your own side, or to deter them from collaborating with the other side. But I think it is useful to separate different types of strategies of violence against civilians to assess their effects.

According to data collected by New America about “homegrown extremists” and the number of lethal terrorist incidents that take place in the US, there have been nearly twice as many deadly right-wing attacks than deadly Jihadist attacks since September 11, 2001. In light of the recent conversation about domestic terrorism, do you find any similarities between the acts committed by rebel groups outside of the US and “homegrown” American extremists?

PF: The recent conversation has been a reaction to the terrible events in Charleston, South Carolina. As far as we know so far, the attack at the Emanuel AME Church was carried out by a "lone wolf" terrorist, not directed by a larger political organization. It's not clear to me how strategically the alleged attacker, Roof, was acting. And I do think that there are important differences between strategic political organizations (even ones with goals I find reprehensible) and lone wolves, so I wouldn't want to push the lessons of my own work for this type of attack too far.

But I do think that the fact that the effect of the attack, at least so far, has been drastically to undercut popular support for the cause Roof espoused (witness the dramatic change of opinion on the Confederate flag) is fairly typical. Big attacks like this generate a lot of attention, but they don't generally help the cause of the people who carry them out. If white supremacists think that attacks like this one will start a race war that they will somehow win, they are pretty sorely deluded.

While ISIS’s rise as a violent terrorist group was a bit too late to be included in your data set, your research is generating an interesting conversation about how it can be applied to the group. Would you consider ISIS to be a “successful” rebel group, or one whose prominence is based merely on their “power to hurt”? Have we simply overestimated their potential to “succeed”? 

PF: I consider ISIS to be successful in the sense that it has grown to be a large and prominent group, and it controls a lot of territory. But it's too soon to tell whether it will succeed in accomplishing its ultimate goal of establishing a fundamentalist Islamic "Caliphate," and it is the ultimate success of groups in achieving their political goals that I focus on in my research. If this case were included in my data, it would currently be coded as an ongoing war—the midpoint in the scale of success that I use. If I try to predict whether ISIS will be successful ultimately? I think it is extremely unlikely that ISIS will end up in a negotiated settlement with Syria or Iraq, largely because of its use of terrorism.

So to succeed it will have to take power or establish a separate state militarily. I'm not an expert on ISIS, but my own view is that the military successes it has had, particularly the territorial successes, have come despite rather than because of its use of deliberately indiscriminate violence against civilians.

The distinction among types of violence is important again here—I think the willingness of ISIS to be quite brutal, and very publicly brutal, with those whom they see as collaborators with their enemies (that is, discriminate violence against civilians) has probably contributed to their success, as have severe weaknesses in the governments they are fighting against. But when they have targeted people indiscriminately, they have lost rather than gained support.

How can we shift the focus from large-scale terrorist activities to a more realistic portrayal that sheds light on the idea that most of terrorism’s power lies in our overestimation of it?

PF: I hope that my research can contribute, at least a little bit, to just such a shift. One of the things terrorism is really good at is getting attention. And psychologically, we tend to be much more afraid of the most spectacular dangers than the most common ones (for example, we're more afraid of flying in planes than driving in cars, even though the latter is more dangerous, because the former makes the news and most deadly traffic accidents do not). And I think that because terrorism is so effective at scaring us, there is a tendency to think that it is effective at generating political change; that is, to overestimate, not just its power to hurt, but especially its power to transform.

But if you step back from the news coverage and actually compare groups that use terrorism to those that don't, you can see that it turns out not to be politically effective in the long run. The media is going to continue to cover the big spectacular terrorist attacks, and they are going to continue to generate fear. So I think it is up to social science researchers to do the more long-term, sober, analysis of what the effects of terrorism really are. It's also up to us to try to assess the causes terrorism—something that gets speculated on by pundits quite a bit, but where we don't have a lot of solid research that compares groups who use terrorism to those who choose other means to effect political change—that's the question I'm working on now.

Q&A Compiled by Kristi DiLallo