If you have tried to debate people who are convinced that there is no link between Islam and Jihadist terrorism then chances are that you have heard of Robert Pape and his studies of suicide terrorism.
Pape is a scientist who has studied suicide terrorism and concluded that it is not Islamic fundamentalism that is motivating the suicide terrorists, but rather foreign occupation. He has collected all known suicide terrorist attacks for the past 30 years, and on the basis of this data he claims to show that what almost all suicide terrorist campaigns have in common is not Islam, but a desire to get rid of a foreign military presence.
The theory not only fully exculpates Islam from any major role in the terror, but it also dovetails nicely with the idea that most of the problems we see in world are of our own making—which is comfortable to believe, because then we can easily do something about it. If only we stopped bombing and occupying the middle east, the problems of terrorism would go away.
Let me be honest up front: I am highly sceptical of the idea that Islam is completely blameless with respect to terrorism. We live in a world in which most—if not close to all—global terrorism is committed by Muslims, who cite their religion as the explicit motivation for their atrocities. If a terrorist attack was to occur tomorrow, we would not really be able to guess in advance, what the assailants socio-ecomonic background would be, where he came from and grew up, or what kind of personality he had. But we can guess with almost certainty that he will be a Muslim and that his last words will be Allahu Akbar.
Any credible theory of terrorism has to grapple with that fact. If the religion of Islam is to be considered completely unrelated to that fact, we would have to find a rival explanation for why this relation occurs, even though it is not causal. And such a rival theory must be supported by mountains of evidence. But most people rarely give such a coherent rival hypothesis; they just assume that such a theory must exist, because they are uncomfortable with the idea that there should be anything inherently bad about Islam. Pape, on the other hand, purports to have exactly such a theory.
If ever there was something that could change my mind on this topic, it would be a theory like the one Pape has put forward. That is, a theory that explains the apparent overrepresentation of Islamic terrorism, using some overlooked factor(s) that is indeed independent of the religion. A theory that not only has a plausible explanation for the phenomenon, but is also supported by a substantial amount of evidence. Pape claims to have just that, and thus he deserves to be taken very seriously.
Luckily, Pape has made all his data readily available for the public—a hallmark any scientist who is confident in his own conclusions—so everybody is invited to check his evidence for themselves. Let us see take a look then, and see how well his theory holds up to critical scrutiny.
Robert Pape is a professor of political science at the University of Chicago as well as the founder and director of the Chicago Project on Security and Terrorism (CPOST), which has created an open database of every suicide terrorist attack since 1982. On the basis of that data, Pape has written two books: "Dying to Win: The Strategic Logic of Suicide Terrorism"(2005) and "Cutting the Fuse: The Explosion of Global Suicide Terrorism and How to Stop It" (2010). The main conclusion of both books is that best explanation for suicide terrorism is not Islamic fundamentalism, but rather foreign occupation.
According to Robert Pape over 95% of all suicide attacks are driven by a strategic logic to compel foreign occupiers to withdraw their forces. That was the case back when he wrote the his first book "Dying to Win" (DtW), which looked at all suicide attacks between 1982 and 2003; it was the case when he wrote his second book "Cutting the Fuse" (CtF), which dealt with the suicide attacks from 2004 to 2009; and it remains the case today. As late as the end of march this year, Robert Pape gave a lecture where he still claimed that 95% of all suicide terrorist attacks are caused by "military interventions".
While Pape's theory has evolved somewhat since his first book, a few key elements remain constant. First of all, foreign occupation has remained his main "risk factor" for suicide terrorism throughout all his research. Pape's definition of what constitutes a foreign occupation, however, has changed significantly—from the slightly dubious to the utterly absurd—over the course of his two books and subsequent lectures.
Secondly, during all of Pape research, the influence of Islamic fundamentalism is consistently downplayed if not flat-out dismissed. While Pape does not disregard religion completely—he claims it still serves a purpose for recruitment, and that a religions differences between the occupied and occupier can increase the likelihood that suicide terrorism will occur as part of a resistance towards occupation—he absolutely dismisses the idea there should be anything inherent in Islam, that is more conducive to suicide terrorism, as opposed to any other religion.
The apparent overrepresentation of Muslim suicide bombers, should, according to Pape, be seen as a result of the foreign occupations that we in the west have forced upon the Muslim world in places like Iraq, Afghanistan and Palestine. And the data he has collected apparently confirms this interpretation.
Pape's evidence
Pape believes his two conclusions—that foreign occupation causes suicide terrorism, and that Islam has nothing to do with—follows directly from the data he has collected. He says there would need to be hundreds of suicide attacks that he missed in order for his theory to collapse. So how does the data prove him right?
The strongest case for Pape's theory was made back in DtW; back when his theory, by virtue of the relative rarity of suicide attacks, still had a semblance of plausibility. During the period studied in that book (1982–2003) the leading suicide terrorism group was not an Islamic group; it was the Tamil Tigers of Sri Lanka (LTTE); a Hindu/secular group, that had nothing to do with Islam. What is more, many of the suicide attacks were perpetrated not by religious groups, but by secular groups such as the PKK in turkey. A third of the suicide attacks by Muslims during this period were by secular groups. And suicide terrorism in the Arab world, was not even limited to Muslims; 8% of the suicide attackers during Israel's occupation in Lebanon from 1982-86 were Christians.
Overall Islamic fundamentalism is only associated with about half of the attacks (still quite significant one might say, but let's ignore that for now), and can therefore be dismissed as the primary cause behind suicide terrorism. Foreign occupation on the other hand accounts for a full 95% of the attacks, and can thus be seen as the main reason or risk factor for suicide terrorism.
In CtF, Pape does not repeat his analysis of what percentage of attacks that are associated with Islamic fundamentalism—having already disproved Islam's influence in DtW, he probably felt there was no need for it—but instead focuses on the explosion of suicide terrorism that occurred since the war in Iraq, which he sees as the vindication of his theory about foreign occupation. After all, suicide terrorism escalated only after we invaded Iraq—precisely what Pape's theory would predict.
The problems with Pape's evidence against Islamist influence
The main problem with Pape's evidence against the influence of Islamic fundamentalism—which should be obvious to anyone who follows the news—is that it simply does not hold up anymore. While it might have been the case before 9/11, suicide terrorism today is almost completely dominated by Islamic fundamentalist groups; a fact which was already undeniable by the time Pape wrote his second book.
Why did Pape not do a similar analysis to the one he did in DtW, where he checked how many suicide terrorist attacks were committed by groups who adhere to an Islamist/Salafist ideology? I can only imagine it must be because the conclusions of such an analysis—as we shall see—would completely undermine his claim that suicide terrorism has nothing to do with Islam..
The only real quantifiable evidence that Pape has against the influence of Islamist extremism, is his analysis of the period from 1982–2003. Bear in mind that the number of suicide attacks in this period only constitute 7.5% of all the attacks in Pape's database—a rather insignificant fraction. Last year alone had more than one and a half times as many suicide attacks as the entire period studied in DtW. But even back in DtW, the Islamist proportion of the suicide attacks was 50%, a fact that can hardly be considered as evidence against the influence of Islamic fundamentalism.
A second problem with Pape's evidence, is that it does not bother to properly investigate the role of Islam itself as distinct from Islamism. Just because a Muslim terror group might ultimately have secular goals, does not mean that the terrorist's religion could not have an influence on his willingness to die for the cause. Pape does attempt a lackluster investigation of this in DtW, but it is ultimately hampered by the way he classified his data.
One has to seriously question how much effort Pape actually put into determining the religion of the suicide attackers. Pape has only determined the religion of roughly half of the attackers in the period from 1982 to 2003 (he has partly stopped classifying the attackers' religion later on—only around 7% are classified today). For example, he claims not to know what religion the attackers in Iraq have. And what is more, Pape cannot even guess the religion of suicide bombers from groups with an explicit Islamist agenda. Is it reasonable to assume that a HAMAS suicide bomber just might be a Buddhist? HAMAS, Islamic Jihad and Al Aqsa Martyrs Brigade alone constitute 50% of all the attacks, where Pape couldn't identify the religion, but could determine the group.
If all that wasn't enough, then I would also question how many of the people, whose religion Pape classified as "secular", are in fact secular. For example, lets take the first two secular attackers in Pape's dataset: Attacks by Ahmad Qasir and Bilal Fahs, which both took place in Lebanon. Ahmad Qasir would, it turns out, regularly go to the mosque, and Hezbollah would later claim his attack as their own. Bilal Fahs' fiancée said he spoke at length about the prayer leader of Jibshit, that he read Islamic books and watched war movies about Islam. In his last letter to the leader of AMAL he apparently wrote: "I will that my brothers in the movement all join hands in the jihad enjoined upon us the Imam-Leader [Khomeini]".1 Not only does this sound distinctly Muslim to me, but it even sounds like both guys might very well have been Islamists.
It turns out that if you attempt to properly estimate the religion of the attackers (I will describe later how I estimate their religion) that at least 85% of them are Muslim. And remember that we are still talking about the early period of suicide attacks. Today Muslims have a near monopoly on suicide terrorism, which is something you will never hear Pape admit. Had Pape cared to honestly show his readers just how many suicide bombers were Muslims (even during the early period of 1982–2003), I highly doubt that people would have found his argument against the influence of Islam very convincing.
What about Pape's claim that 8% of the Lebanese suicide attackers from 1982–1986 were Christians? Here we have a curious discrepancy between what Pape says in his books and what is actually supported by his data. In the database there appears to be only two Christian attacks: One by a member of the Syrian Social Nationalist Party, which does appear to be a genuine case of a Christian suicide attack, and another by a group called the Vanguard of Arab Christians.
The Vanguard of Arab Christians is a group that no one had ever heard of at the time, and have never heard from since. Their only attack was also claimed by another group called Forces of the Ali Ayyoub Group. The Christian radio station at the time said that both claims were "dubious". And in any case the attacker only killed other Christians. Overall, I would be very hesitant to conclude that this was a Christian attack, although it is of course possible.
Let's cut Pape some slack and say these two attacks were indeed perpetrated by Christians. That still does not square with the claims that Pape makes in his books. In DtW, Pape claims that there was yet another Christian suicide attacker, whom he even names. What is more, in CtF, Pape claims there was yet two more Christian suicide attacks in the same period, bringing the percentage of Christian attacks up to 15%.2
Not only is it strange, that the number of Christian attacks has changed from one book to the other (the overall number of attacks has not changed, in fact it is lower in CtF than in DtW), but none of these attacks can be found in Pape's database. The only Christian terrorist attacks in the entire CPOST database of suicide attacks, are the two cases I just mentioned above. A couple of weeks ago, I therefore decided to send CPOST an email asking for an explanation for this discrepancy. So far I have yet to receive an answer.
In the past seven years, more than a hundred Tibetans have self-immolated as a protest against the Chinese occupation of Tibet. Apparently there is no shortage of Tibetans who are willing to sacrifice their own life in resistance to the Chinese occupation, which has clearly driven them to the point of desperation. Yet not one—not even one!—has decided to take a few innocent Chinese lives with him in death. Pape is never confronted with this problem in his research, because he only looks at cases where suicide terrorism has already occurred, yet this is a serious problem for his theory, if he wants to claim that foreign occupation causes suicide terrorism.
The worst problem with Pape's theory, however, is his definition of what constitutes a "foreign occupation". Any unsuspecting reader might assume it was something akin to the Hague definition, or anything comparable in international law or the laws of war. But Pape's definition is nothing like that. Instead he defines occupation merely as meaning the "exertion of political control over territory by an outside group".4 A deliberately vague definition, which is able, as it turns out, to include almost anything.
The first instance where alarm bells should go off, is that Pape sees Al Qaeda's attacks against the United States as a response to America's "occupation" of Saudi Arabia. The only problem is that Saudi Arabia has never been occupied by America. Pape does acknowledge this, but he tells us that the fact the U.S. has had a military base in the country, housing some 12.000 troops since the first gulf war, is enough for Al Qaeda to "perceive" it as an occupation—and the perception is what matters. Nevermind the fact that these troops were present with the blessing of the Saudi government; nevermind that they were originally put there to defend the country against Iraq; and nevermind that the troops have never interfered with anything in Saudi Arabia's internal affairs, let alone killed any Saudi citizen. The mere presence of foreign troops is enough, Pape assures us, to make it a casus belli in the eyes of Al Qaeda, and thus cause them to use suicide terrorism against America.
Pape admits that his definition would mean that other countries such as Bahrain, UAE and Qatar, where the US has also stationed troops, would likewise be considered occupied—he says this almost as a prediction that we could expect attacks in these countries in the future. What he should have mentioned, however, is that it also includes countries like Germany, Japan and South Korea, all of which have much larger contingents of US troops than Saudi Arabia. Does Pape mean to tell us that we should expect German suicide attacks against the U.S.?
It gets worse. If it seems hard to fit Al-Qaeda's attacks within the theory of foreign occupation, then it would be completely impossible with the massive wave of suicide attacks that erupted in Pakistan around 2007 and still continues to this day. Pakistan suffers neither from a conventional military occupation, nor from the presence of foreign troops. How on earth does Pape manage to square the case of Pakistan with his theory? Well, he invents—and I promise you, I am not making this up—something he calls "indirect occupation".
Pape tells us that "in indirect occupation, the indirectly occupied country gives a higher priority to the goals of the indirect occupier than its national interest alone would warrant". "When Musharraf acquiesced to U.S. demands and shifted around 100,000 troops from the Eastern front against India to engage militant targets in western Pakistan, the alliance transitioned into an indirect occupation."5
One starts to wonder why Pape's theory does not account for 100% of the suicide attacks. I am having a hard time coming up with any country, that could not with some similarly tortuous ad hoc reasoning be considered occupied.
Therefore I was not surprised when Pape in a recent lecture, now says ISIS' attacks are motivated in part by the occupation of the Syrian government and in part by the military intervention of the international coalition. And when an incredulous audience member asked him about the Nigerian terror group Boko Haram, Pape merely said "It's the Christians" and that "we support them", presumably implying that the Nigerian government is occupying its own country, or at least is perceived to do so, by its Muslim population—and that we are to some extent responsible for it.
The final problem with Pape's theory is that when a country is first shown to be "occupied", then Pape just assumes that practically every attack within that country must be motivated by the occupation. Meaning that when purely sectarian attacks occur between Sunni and Shia Muslims in a country that just happens to be occupied (by Pape's twisted definition of course), then Pape will still consider the attack as "caused" by the foreign occupation. When Sunni militants blow up multiple Shia mosques in Pakistan, Pape will still code those attacks as "Pakistani Rebels vs. Pakistan & U.S. Allies"6, because, as he explains, "the broader goal of sectarian attacks is to undermine the Pakistani government."7
This is utter nonsense.
What sense then, can anyone make of Pape's claim that 95% of suicide attacks are caused by foreign occupation? Given his infinitely elastic definition of occupation, the claim becomes absolutely worthless.
Why did Pape not do a similar analysis to the one he did in DtW, where he checked how many suicide terrorist attacks were committed by groups who adhere to an Islamist/Salafist ideology? I can only imagine it must be because the conclusions of such an analysis—as we shall see—would completely undermine his claim that suicide terrorism has nothing to do with Islam..
The only real quantifiable evidence that Pape has against the influence of Islamist extremism, is his analysis of the period from 1982–2003. Bear in mind that the number of suicide attacks in this period only constitute 7.5% of all the attacks in Pape's database—a rather insignificant fraction. Last year alone had more than one and a half times as many suicide attacks as the entire period studied in DtW. But even back in DtW, the Islamist proportion of the suicide attacks was 50%, a fact that can hardly be considered as evidence against the influence of Islamic fundamentalism.
A second problem with Pape's evidence, is that it does not bother to properly investigate the role of Islam itself as distinct from Islamism. Just because a Muslim terror group might ultimately have secular goals, does not mean that the terrorist's religion could not have an influence on his willingness to die for the cause. Pape does attempt a lackluster investigation of this in DtW, but it is ultimately hampered by the way he classified his data.
One has to seriously question how much effort Pape actually put into determining the religion of the suicide attackers. Pape has only determined the religion of roughly half of the attackers in the period from 1982 to 2003 (he has partly stopped classifying the attackers' religion later on—only around 7% are classified today). For example, he claims not to know what religion the attackers in Iraq have. And what is more, Pape cannot even guess the religion of suicide bombers from groups with an explicit Islamist agenda. Is it reasonable to assume that a HAMAS suicide bomber just might be a Buddhist? HAMAS, Islamic Jihad and Al Aqsa Martyrs Brigade alone constitute 50% of all the attacks, where Pape couldn't identify the religion, but could determine the group.
If all that wasn't enough, then I would also question how many of the people, whose religion Pape classified as "secular", are in fact secular. For example, lets take the first two secular attackers in Pape's dataset: Attacks by Ahmad Qasir and Bilal Fahs, which both took place in Lebanon. Ahmad Qasir would, it turns out, regularly go to the mosque, and Hezbollah would later claim his attack as their own. Bilal Fahs' fiancée said he spoke at length about the prayer leader of Jibshit, that he read Islamic books and watched war movies about Islam. In his last letter to the leader of AMAL he apparently wrote: "I will that my brothers in the movement all join hands in the jihad enjoined upon us the Imam-Leader [Khomeini]".1 Not only does this sound distinctly Muslim to me, but it even sounds like both guys might very well have been Islamists.
It turns out that if you attempt to properly estimate the religion of the attackers (I will describe later how I estimate their religion) that at least 85% of them are Muslim. And remember that we are still talking about the early period of suicide attacks. Today Muslims have a near monopoly on suicide terrorism, which is something you will never hear Pape admit. Had Pape cared to honestly show his readers just how many suicide bombers were Muslims (even during the early period of 1982–2003), I highly doubt that people would have found his argument against the influence of Islam very convincing.
What about Pape's claim that 8% of the Lebanese suicide attackers from 1982–1986 were Christians? Here we have a curious discrepancy between what Pape says in his books and what is actually supported by his data. In the database there appears to be only two Christian attacks: One by a member of the Syrian Social Nationalist Party, which does appear to be a genuine case of a Christian suicide attack, and another by a group called the Vanguard of Arab Christians.
The Vanguard of Arab Christians is a group that no one had ever heard of at the time, and have never heard from since. Their only attack was also claimed by another group called Forces of the Ali Ayyoub Group. The Christian radio station at the time said that both claims were "dubious". And in any case the attacker only killed other Christians. Overall, I would be very hesitant to conclude that this was a Christian attack, although it is of course possible.
Let's cut Pape some slack and say these two attacks were indeed perpetrated by Christians. That still does not square with the claims that Pape makes in his books. In DtW, Pape claims that there was yet another Christian suicide attacker, whom he even names. What is more, in CtF, Pape claims there was yet two more Christian suicide attacks in the same period, bringing the percentage of Christian attacks up to 15%.2
Not only is it strange, that the number of Christian attacks has changed from one book to the other (the overall number of attacks has not changed, in fact it is lower in CtF than in DtW), but none of these attacks can be found in Pape's database. The only Christian terrorist attacks in the entire CPOST database of suicide attacks, are the two cases I just mentioned above. A couple of weeks ago, I therefore decided to send CPOST an email asking for an explanation for this discrepancy. So far I have yet to receive an answer.
The problems with Pape's evidence for his theory about foreign occupation
First of all, Pape is sampling on the dependent variable, meaning he only looks at cases where suicide terrorism has occured, and does not bother to seriously investigate all the other instances of foreign occupations where suicide terrorism did not occur.3 Pape cannot say that foreign occupation, or anything else for that matter, is the cause of suicide terrorism until he has investigated how suicide terrorism varies with respect to foreign occupations. Where are the Tibetan or Cypriot suicide bombers? just to name a few. Pape does not even attempt to investigate this, let alone quantify it.In the past seven years, more than a hundred Tibetans have self-immolated as a protest against the Chinese occupation of Tibet. Apparently there is no shortage of Tibetans who are willing to sacrifice their own life in resistance to the Chinese occupation, which has clearly driven them to the point of desperation. Yet not one—not even one!—has decided to take a few innocent Chinese lives with him in death. Pape is never confronted with this problem in his research, because he only looks at cases where suicide terrorism has already occurred, yet this is a serious problem for his theory, if he wants to claim that foreign occupation causes suicide terrorism.
The worst problem with Pape's theory, however, is his definition of what constitutes a "foreign occupation". Any unsuspecting reader might assume it was something akin to the Hague definition, or anything comparable in international law or the laws of war. But Pape's definition is nothing like that. Instead he defines occupation merely as meaning the "exertion of political control over territory by an outside group".4 A deliberately vague definition, which is able, as it turns out, to include almost anything.
The first instance where alarm bells should go off, is that Pape sees Al Qaeda's attacks against the United States as a response to America's "occupation" of Saudi Arabia. The only problem is that Saudi Arabia has never been occupied by America. Pape does acknowledge this, but he tells us that the fact the U.S. has had a military base in the country, housing some 12.000 troops since the first gulf war, is enough for Al Qaeda to "perceive" it as an occupation—and the perception is what matters. Nevermind the fact that these troops were present with the blessing of the Saudi government; nevermind that they were originally put there to defend the country against Iraq; and nevermind that the troops have never interfered with anything in Saudi Arabia's internal affairs, let alone killed any Saudi citizen. The mere presence of foreign troops is enough, Pape assures us, to make it a casus belli in the eyes of Al Qaeda, and thus cause them to use suicide terrorism against America.
Pape admits that his definition would mean that other countries such as Bahrain, UAE and Qatar, where the US has also stationed troops, would likewise be considered occupied—he says this almost as a prediction that we could expect attacks in these countries in the future. What he should have mentioned, however, is that it also includes countries like Germany, Japan and South Korea, all of which have much larger contingents of US troops than Saudi Arabia. Does Pape mean to tell us that we should expect German suicide attacks against the U.S.?
It gets worse. If it seems hard to fit Al-Qaeda's attacks within the theory of foreign occupation, then it would be completely impossible with the massive wave of suicide attacks that erupted in Pakistan around 2007 and still continues to this day. Pakistan suffers neither from a conventional military occupation, nor from the presence of foreign troops. How on earth does Pape manage to square the case of Pakistan with his theory? Well, he invents—and I promise you, I am not making this up—something he calls "indirect occupation".
Pape tells us that "in indirect occupation, the indirectly occupied country gives a higher priority to the goals of the indirect occupier than its national interest alone would warrant". "When Musharraf acquiesced to U.S. demands and shifted around 100,000 troops from the Eastern front against India to engage militant targets in western Pakistan, the alliance transitioned into an indirect occupation."5
One starts to wonder why Pape's theory does not account for 100% of the suicide attacks. I am having a hard time coming up with any country, that could not with some similarly tortuous ad hoc reasoning be considered occupied.
Therefore I was not surprised when Pape in a recent lecture, now says ISIS' attacks are motivated in part by the occupation of the Syrian government and in part by the military intervention of the international coalition. And when an incredulous audience member asked him about the Nigerian terror group Boko Haram, Pape merely said "It's the Christians" and that "we support them", presumably implying that the Nigerian government is occupying its own country, or at least is perceived to do so, by its Muslim population—and that we are to some extent responsible for it.
The final problem with Pape's theory is that when a country is first shown to be "occupied", then Pape just assumes that practically every attack within that country must be motivated by the occupation. Meaning that when purely sectarian attacks occur between Sunni and Shia Muslims in a country that just happens to be occupied (by Pape's twisted definition of course), then Pape will still consider the attack as "caused" by the foreign occupation. When Sunni militants blow up multiple Shia mosques in Pakistan, Pape will still code those attacks as "Pakistani Rebels vs. Pakistan & U.S. Allies"6, because, as he explains, "the broader goal of sectarian attacks is to undermine the Pakistani government."7
This is utter nonsense.
What sense then, can anyone make of Pape's claim that 95% of suicide attacks are caused by foreign occupation? Given his infinitely elastic definition of occupation, the claim becomes absolutely worthless.
What does the data actually show?
Pape allows free access to his complete database of suicide terrorism, which is accessible from the CPOST website. While that might seem really great at first, Pape has unfortunately not made it easy to see what the role of religion might be in relation to suicide terrorism.
One of the only parameters you can't search for in the database, is the religion of the attacker. Pape has likewise not made it obvious for the reader what attacks are motivated by Islamic fundamentalism—or foreign occupation for that matter. We are just supposed to guess, how Pape exactly classified these attacks.
The worst part of it all, is that Pape has made it deliberately hard for people to download his data. To be fair, you can actually download the datasets, but they do not include any of the relevant details, such as group and religion, thus making it very difficult to do any form of analysis on these parameters. If you want to see which terror group was behind a particular attack, or what religion the attacker had, then you have to manually check the details of the attack. Meaning that in order to do any statistical analysis on these parameters, you would have to manually go through the details of all 4933 attacks—so that's what I did!8
So what does the data actually show? Let's start with Pape's own classification of religion. Meaning attacks where he identified one or more of the attackers religion. These 355 attacks constitute roughly 7% of all the suicide attacks. For sake of simplicity, I have aggregated all different Muslim denominations together into just "Muslim".
Pape's own data on religion reveal that 69.5% of all attackers are Muslim, roughly 26.7% are Hindu, 3.4% are secular, and then we have a single Christian attacker (whom you might have a hard time spotting in the pie chart, but there is actually a very thin blue line at the very top).
A clear Muslim majority, but still a surprisingly high number of Hindu attackers. The group of secular attackers is, as we saw above, a somewhat dubious category, but even if we add it to the Muslim attackers, we still don't get above 75% of all attacks.
But let's move beyond Pape's own classification and attempt to recreate his analysis of Islamic fundamentalism. To do this I have looked through all of the 102 terror groups in Pape's data, and attempted to identify whether or not the groups adhere to Islamist ideology. Using primarily Standford's Mapping Militant Organizations project and Tracking Terrorism I have considered a group to be Islamist, if these sites described its ideology as Islamist, or if the goal of the group is to implement an Islamic state based on Sharia law. If I have been unable to find any reliable information on the ideology of the group, I have classified it as unknown.
Just to avoid any accusations of somehow having tipped the balance in favor of Islamism, I have deliberately tried to skew my classification in favor of secularism. For instance, a group like the Islamic Army of Iraq, is ostensibly an Islamist group that have fought along side ISIS. But the group is described as advocating a "softer" version of Sharia law, which doesn't really compare with the version espoused by ISIS. Its goals are also distinctly nationalistic in nature, as opposed to the global caliphate that ISIS has in mind. Fair enough, I classify that group as secular then.
What is more, if just one secular group has claimed credit for an attack, then I classify that attack as secular, regardless of how many other rival Islamist groups might have taken credit for the same attack. Meaning that whenever there is the slightest doubt that this could be a secular attack, I classify it as secular.
With those rules for classifying groups, I end up with 75 Islamist groups, meaning roughly 3 out 4 terror groups are Islamist. One should keep in mind, however, that although Pape's data contain 102 distinct groups, the vast majority of these groups are obscure groups who have only carried out one or a handful of attacks. The groups that really matter in the big picture, all have names you will recognize immediately if you are at all familiar with the subject.
One should also keep in mind, that Pape has only managed to identify the responsible group in roughly half of the suicide attacks. It seems nonetheless reasonable to assume that the unidentified attacks would likely be distributed along the same ideological patterns.
As the chart above reveals, the vast majority of all attacks are committed by Islamist groups. A whopping 91%! If one looks at the number of people killed instead of the number of attacks, then Islamist groups are responsible for 93% of all the people killed by suicide attacks. This finding goes directly against everything Pape has been saying for the past 15 years and yet it comes directly from his own data.
If you look at how the distribution has evolved throughout the years, one can see a surge in Islamist suicide attacks, starting roughly around 2001, that quickly eclipses secular terrorism. One might forgive Pape for thinking Islamist suicide terrorism was unimportant before 2001, but today it is simply inexcusable. By now Islamist groups have close to a monopoly on suicide attacks.
Had Pape bothered to do an analysis of Islamist influence back when he wrote CtF, he would already then have discovered that Islamist groups perpetrated more than the lion's share—83%— of all suicide attacks. There was absolutely no excuse not to have done such an analysis.
Keep in mind that these results come not from some vague definition of what constitutes an Islamist attack—unlike Pape's definition of foreign occupation. This is what you get with the most conservative definition of what constitutes Islamic fundamentalism and deliberately having skewed the data against it. The actual proportion Islamist attacks is almost certainly even higher.
What if we attempt to look at islam per se? If we look at religion alone, we can reasonably—or so I would argue—infer the religion of not only the secular groups, but also of all those attacks where Pape could not identify the perpetrators.
For the 15 secular groups, 8 can be assumed to be Muslim groups. Some groups like PFLP, the Lebanese Ba'ath Party and Communist Party have all had Christian members, I have chosen to classify those as "Islam/Christianity". But groups like Al-Aqsa Martyrs Brigade, AMAL, the Free Syrian Army and the various Kurdish groups are distinctly Muslim groups, and therefore these can safely be assumed to be "Muslim" groups.
As for all the unknown attacks, we can roughly infer the religion of the perpetrators by looking at the location of the attack. Of all unknown attacks, almost 90% took place in just 6 countries: Afghanistan, Iraq, Pakistan, Nigeria, Syria and Yemen. With the exception of Nigeria, these are all Muslim majority countries with an almost exclusively Muslim population. I think it is reasonable to expect all attacks in these countries to be perpetrated by Muslims. While Nigeria is not a Muslim majority country, all of its known suicide attacks have been committed by an Islamist group, Boko Haram, plus Pape has classified these attacks as part of the same campaign that Boko Haram has been waging.
Overall, if the attack has taken place in a region where Muslim terror groups are the only ones known to ever have committed suicide terrorist attacks, then I would think it is reasonable to assume that the unknown attackers are also Muslims. That means that attacks in Lebanon and Israel for instance cannot be assumed to be Muslim; but attacks in Pakistan or Chechnya can.
At this point we are down to so few attacks that I also took the time to sift through a good chunk of the sources for the remaining ones, which often have clues to the motive behind the attack. The unknown attack in Spain for instance was clearly related to the Madrid bombings and can thus safely be assumed to be Muslim attack.
With this rough estimate of the attackers' religion, we can now get an idea of just how many suicide terrorist are Muslims. The answer is 96%. Both when you look at number of attacks and number of people killed. While you might say there is a lot of assumptions included in this estimate, a simple reality check would confirm that this is probably even still a conservative estimate.
The truth is that the LTTE—which was defeated in 2009—was the only non-Muslim terror group that has ever waged a large scale suicide terrorism campaign. Apart from a handful of sporadic, unrelated attacks committed by non-Muslims, all other attacks have been perpetrated by Muslims, and the vast majority of those directly in the name of Islam.
Considering the fact that Pape has done everything he could to hide or obfuscate the influence of Islam, he has nevertheless not been able to find any significant amount of non-Muslim suicide attacks; if we consider LTTE an outlier he would have practically nothing. Feel free to disprove me here. If you can find more than 20 non-Muslims suicide attacks, that were not perpetrated by the LTTE, I would be quite surprised. But even so, those numbers would still be completely dwarfed by the Muslim attacks.
One of the only parameters you can't search for in the database, is the religion of the attacker. Pape has likewise not made it obvious for the reader what attacks are motivated by Islamic fundamentalism—or foreign occupation for that matter. We are just supposed to guess, how Pape exactly classified these attacks.
The worst part of it all, is that Pape has made it deliberately hard for people to download his data. To be fair, you can actually download the datasets, but they do not include any of the relevant details, such as group and religion, thus making it very difficult to do any form of analysis on these parameters. If you want to see which terror group was behind a particular attack, or what religion the attacker had, then you have to manually check the details of the attack. Meaning that in order to do any statistical analysis on these parameters, you would have to manually go through the details of all 4933 attacks—so that's what I did!8
So what does the data actually show? Let's start with Pape's own classification of religion. Meaning attacks where he identified one or more of the attackers religion. These 355 attacks constitute roughly 7% of all the suicide attacks. For sake of simplicity, I have aggregated all different Muslim denominations together into just "Muslim".
Pape's own data on religion reveal that 69.5% of all attackers are Muslim, roughly 26.7% are Hindu, 3.4% are secular, and then we have a single Christian attacker (whom you might have a hard time spotting in the pie chart, but there is actually a very thin blue line at the very top).
A clear Muslim majority, but still a surprisingly high number of Hindu attackers. The group of secular attackers is, as we saw above, a somewhat dubious category, but even if we add it to the Muslim attackers, we still don't get above 75% of all attacks.
But let's move beyond Pape's own classification and attempt to recreate his analysis of Islamic fundamentalism. To do this I have looked through all of the 102 terror groups in Pape's data, and attempted to identify whether or not the groups adhere to Islamist ideology. Using primarily Standford's Mapping Militant Organizations project and Tracking Terrorism I have considered a group to be Islamist, if these sites described its ideology as Islamist, or if the goal of the group is to implement an Islamic state based on Sharia law. If I have been unable to find any reliable information on the ideology of the group, I have classified it as unknown.
Just to avoid any accusations of somehow having tipped the balance in favor of Islamism, I have deliberately tried to skew my classification in favor of secularism. For instance, a group like the Islamic Army of Iraq, is ostensibly an Islamist group that have fought along side ISIS. But the group is described as advocating a "softer" version of Sharia law, which doesn't really compare with the version espoused by ISIS. Its goals are also distinctly nationalistic in nature, as opposed to the global caliphate that ISIS has in mind. Fair enough, I classify that group as secular then.
What is more, if just one secular group has claimed credit for an attack, then I classify that attack as secular, regardless of how many other rival Islamist groups might have taken credit for the same attack. Meaning that whenever there is the slightest doubt that this could be a secular attack, I classify it as secular.
With those rules for classifying groups, I end up with 75 Islamist groups, meaning roughly 3 out 4 terror groups are Islamist. One should keep in mind, however, that although Pape's data contain 102 distinct groups, the vast majority of these groups are obscure groups who have only carried out one or a handful of attacks. The groups that really matter in the big picture, all have names you will recognize immediately if you are at all familiar with the subject.
One should also keep in mind, that Pape has only managed to identify the responsible group in roughly half of the suicide attacks. It seems nonetheless reasonable to assume that the unidentified attacks would likely be distributed along the same ideological patterns.
As the chart above reveals, the vast majority of all attacks are committed by Islamist groups. A whopping 91%! If one looks at the number of people killed instead of the number of attacks, then Islamist groups are responsible for 93% of all the people killed by suicide attacks. This finding goes directly against everything Pape has been saying for the past 15 years and yet it comes directly from his own data.
If you look at how the distribution has evolved throughout the years, one can see a surge in Islamist suicide attacks, starting roughly around 2001, that quickly eclipses secular terrorism. One might forgive Pape for thinking Islamist suicide terrorism was unimportant before 2001, but today it is simply inexcusable. By now Islamist groups have close to a monopoly on suicide attacks.
Had Pape bothered to do an analysis of Islamist influence back when he wrote CtF, he would already then have discovered that Islamist groups perpetrated more than the lion's share—83%— of all suicide attacks. There was absolutely no excuse not to have done such an analysis.
Keep in mind that these results come not from some vague definition of what constitutes an Islamist attack—unlike Pape's definition of foreign occupation. This is what you get with the most conservative definition of what constitutes Islamic fundamentalism and deliberately having skewed the data against it. The actual proportion Islamist attacks is almost certainly even higher.
What if we attempt to look at islam per se? If we look at religion alone, we can reasonably—or so I would argue—infer the religion of not only the secular groups, but also of all those attacks where Pape could not identify the perpetrators.
For the 15 secular groups, 8 can be assumed to be Muslim groups. Some groups like PFLP, the Lebanese Ba'ath Party and Communist Party have all had Christian members, I have chosen to classify those as "Islam/Christianity". But groups like Al-Aqsa Martyrs Brigade, AMAL, the Free Syrian Army and the various Kurdish groups are distinctly Muslim groups, and therefore these can safely be assumed to be "Muslim" groups.
As for all the unknown attacks, we can roughly infer the religion of the perpetrators by looking at the location of the attack. Of all unknown attacks, almost 90% took place in just 6 countries: Afghanistan, Iraq, Pakistan, Nigeria, Syria and Yemen. With the exception of Nigeria, these are all Muslim majority countries with an almost exclusively Muslim population. I think it is reasonable to expect all attacks in these countries to be perpetrated by Muslims. While Nigeria is not a Muslim majority country, all of its known suicide attacks have been committed by an Islamist group, Boko Haram, plus Pape has classified these attacks as part of the same campaign that Boko Haram has been waging.
Overall, if the attack has taken place in a region where Muslim terror groups are the only ones known to ever have committed suicide terrorist attacks, then I would think it is reasonable to assume that the unknown attackers are also Muslims. That means that attacks in Lebanon and Israel for instance cannot be assumed to be Muslim; but attacks in Pakistan or Chechnya can.
At this point we are down to so few attacks that I also took the time to sift through a good chunk of the sources for the remaining ones, which often have clues to the motive behind the attack. The unknown attack in Spain for instance was clearly related to the Madrid bombings and can thus safely be assumed to be Muslim attack.
With this rough estimate of the attackers' religion, we can now get an idea of just how many suicide terrorist are Muslims. The answer is 96%. Both when you look at number of attacks and number of people killed. While you might say there is a lot of assumptions included in this estimate, a simple reality check would confirm that this is probably even still a conservative estimate.
The truth is that the LTTE—which was defeated in 2009—was the only non-Muslim terror group that has ever waged a large scale suicide terrorism campaign. Apart from a handful of sporadic, unrelated attacks committed by non-Muslims, all other attacks have been perpetrated by Muslims, and the vast majority of those directly in the name of Islam.
Considering the fact that Pape has done everything he could to hide or obfuscate the influence of Islam, he has nevertheless not been able to find any significant amount of non-Muslim suicide attacks; if we consider LTTE an outlier he would have practically nothing. Feel free to disprove me here. If you can find more than 20 non-Muslims suicide attacks, that were not perpetrated by the LTTE, I would be quite surprised. But even so, those numbers would still be completely dwarfed by the Muslim attacks.
As for Pape's theory of foreign occupation, a brief look at the graph above should be enough to cast serious doubt upon it. The graph shows that the number of terrorist attacks in the world, and as you can see, the level is higher than it has ever been before. Yet, by any reasonable metric the number of foreign occupations has gone down significantly—certainly the number and level of Western occupations in the Muslim world has drastically decreased.
If you look at Wikipedias List of military occupations, you will not only see that the far majority of these countries faced no suicide terrorism despite the occupation, but all the countries in the world that do experience major suicide terrorism campaigns are conspicuously absent from the list. Of course Wikipedia uses a different definition of occupation than Pape, but that is part of the whole problem.
Pape theory is untestable. It is simply unclear what fits in his definition of occupation and what doesn't. And he does not provide a clear explanation of how he himself classified countries as occupied or not, making it impossible for me to even replicate his own study. Nor does his theory make any useful predictions that could be retrospectively tested.
That said, I decided to plot the number of U.S. troops in Afghanistan and Iraq over time and compare that to the number of suicide attacks.9 If Pape's theory was to make any useful prediction, it should be that the suicide attacks should decrease as the troop count decreased. And keep in mind that Iraq and Afghanistan represents the two strongest cases for Pape's theory.
In Afghanistan the number of suicide attacks only seemed to spike after several years of occupation (and curiously enough around the same time that it started in Iraq), and has since remained at roughly the same level. The number of U.S. troops in the country has changed dramatically over time, but this does not seem to have had any significant impact on the terrorism.
In Iraq the correlation seems to be a lot more in Pape's favor—at least at first glance. Suicide terrorism and number of troops does indeed seem to fall together from 2007 on. But theory collapses completely in 2013 when terrorism surged once more, despite the fact that there were no U.S. troops in the country! In 2015 America decided to put a couple of thousand new troops back to help the Iraqi government against ISIS, but these numbers are not even close to previous levels, and they arrive too late to explain why the surge occurred in the first place.
I don't doubt that Pape will be able to conjure up with some ad hoc explanations for Iraq was still occupied in 2013 and how the U.S. is responsible, but do we really have to take his explanations seriously at this point?
Conclusion
If the evidence I have presented so far has not yet convinced you that Pape's theory on suicide terrorism is complete garbage, then I don't know what will. I could name more problems with his study—such as his many ad-hoc explanations for why the history conforms to his theory, the high number of "foreign fighters", or the many suicide attacks that Pape seems to have missed—but none of those problems substantially change the big picture (and you are at any rate probably already bored).
I suspect that most of the people who were taken in by Pape's theory, did so not because of the overwhelming evidence that Pape presented in support of it—I doubt most people even bothered to look at his evidence. Rather, they have been endorsing Pape's theory because they liked his conclusions; namely that Islam was blameless with respect to terrorism and that the terrorists are of our own making. These people are simply not willing to accept the fact that religion might have an influence.
Pape himself seems incapable of imagining that people will ever do anything for religious reasons; he even states this more or less explicitly in one of his recent lectures. He compares the religious language of ISIS suicide bombers, to the American soldiers at D-Day, who were not religious at all, but suddenly started going through the Rosary, when faced with certain death; "because what have I got to lose?". Pape says "suicide attackers are going to their death; they are not making any mistake about it", meaning nobody really believes in paradise or the afterlife.
You should really watch those two minutes of the lecture. That moment just perfectly captures the essence of Pape's confusion on this topic. The man is simply incapable of imagining that religious people ever believe what they say they believe.
UPDATE 23/10 2016:
Based on the response I received, I decided to rewrite the introduction and expand on the evidence against Pape's theory of foreign occuption at the very end of the article. I thought the mere fact that I had shown Pape's defintion of occupation to be utterly meaningless was enough to discredit it, but apparently a few people were not convinced. To be honest, I am not sure how one could even test Pape's theory. Because his definition of occupation is so vague, it is simply impossible to test.
Still, I felt like I could have said a few more words about it, and thus I have now included an investigation of the correlation between the number of U.S. troops and the number of suicide attacks in Iraq and Afghanistan.
Incidentally, Pape has still not responded to my question about the Christian suicide bombers.
I suspect that most of the people who were taken in by Pape's theory, did so not because of the overwhelming evidence that Pape presented in support of it—I doubt most people even bothered to look at his evidence. Rather, they have been endorsing Pape's theory because they liked his conclusions; namely that Islam was blameless with respect to terrorism and that the terrorists are of our own making. These people are simply not willing to accept the fact that religion might have an influence.
Pape himself seems incapable of imagining that people will ever do anything for religious reasons; he even states this more or less explicitly in one of his recent lectures. He compares the religious language of ISIS suicide bombers, to the American soldiers at D-Day, who were not religious at all, but suddenly started going through the Rosary, when faced with certain death; "because what have I got to lose?". Pape says "suicide attackers are going to their death; they are not making any mistake about it", meaning nobody really believes in paradise or the afterlife.
You should really watch those two minutes of the lecture. That moment just perfectly captures the essence of Pape's confusion on this topic. The man is simply incapable of imagining that religious people ever believe what they say they believe.
UPDATE 23/10 2016:
Based on the response I received, I decided to rewrite the introduction and expand on the evidence against Pape's theory of foreign occuption at the very end of the article. I thought the mere fact that I had shown Pape's defintion of occupation to be utterly meaningless was enough to discredit it, but apparently a few people were not convinced. To be honest, I am not sure how one could even test Pape's theory. Because his definition of occupation is so vague, it is simply impossible to test.
Still, I felt like I could have said a few more words about it, and thus I have now included an investigation of the correlation between the number of U.S. troops and the number of suicide attacks in Iraq and Afghanistan.
Incidentally, Pape has still not responded to my question about the Christian suicide bombers.
1 Info about the attackers are taken from Martin Kramer.
2 Cutting the Fuse (2010) p.196.
3 A point made much better in this paper, which criticizes Pape for this exact error.
4 Dying to Win (2005).
5 Cutting the Fuse (2010) p.140.
6 I cannot from the CPOST database alone, tell you exactly how Pape classified an attack as caused by foreign occupation. But I assume it must have something to do with his "campaign" classification. Since the vast majority of all attacks in Pakistan have been classified as the campaign "Pakistani Rebels vs. Pakistan & U.S. Allies" and since Pape has to believe that at least the majority of the Pakistani attacks are caused by foreign occupation (otherwise he would never reach the 95% of all suicide attacks, that he claims his theory explains), then I assume this is a campaign he classifies as pertaining to foreign occupation. Otherwise Pape's must have some classification that he does not share in his public database.
7 Cutting the Fuse (2010) p.150.
8 Obviously I did not literally do that. Instead I wrote a piece of software that could accomplish the task.
9 Number of U.S. troops is taken from the army's reports to Congress and can be found here and here.
Ingen kommentarer:
Send en kommentar