No terror impact.

The likelihood that anyone outside a war zone will be killed by an Islamist extremist terrorist is extremely small. In the United States, for example, some six people have perished each year since 9/11 at the hands of such terrorists—for an annual fatality rate of about one in 50 million for the period.

This might be taken to suggest, as one writer has characterized it, that “terrorism is such a minor threat to American life and limb that it’s simply bizarre—just stupefyingly irrational and intellectually unserious—to suppose that it could even begin to justify the abolition of privacy rights as they have been traditionally understood in favour of the installation of a panoptic surveillance state.” 1 And terrorism specialist Marc Sageman characterizes the threat terrorists present in the United States as “rather negligible.” 2 The vast majority of what is commonly tallied as terrorism has occurred in war zones, and this is especially true for fatalities.3 But even this has been exaggerated by conflating terrorism with war: civil war violence that would previously have been seen to be acts of insurgency are now often labeled terrorism.4

In order to put the numbers in some context, it has often been pointed out that far more Americans are killed each year not only by such highly destructive hazards as drug overdoses or automobile accidents, but even by such comparatively minor ones as lightning, accident-causing deer, peanut allergies, or drowning in bathtubs. Some comparisons are arrayed in Table 1.

In recent years, however, critics have attacked what they call “the bathtub fallacy.” 5

First, they stress that it is important to keep in mind that bathtubs are not out to kill you while terrorism is a willful act carried out by diabolical, dedicated, and clever human beings. Thus, although the number of people Islamist terrorists have been able to kill in the West since 9/11 has thus far been quite limited, those terrorists, as they plot and plan and learn from experience, may very well become far more destructive in the future.

Second, the critics charge that the comparison of terrorism with bathtub drownings is incomplete in that it doesn’t consider the possibility that the incidence of terrorist destruction is low precisely because counterterrorism measures are so effective.

Third, it is argued that, unlike bathtub drownings, terrorism exacts costs far beyond those entailed in the event itself. It damagingly sows terror, fear, and anxiety; disturbs our

psychological well-being; undermines trust and openness within the society; and reduces our sense of intrinsic moral worth even as it increases a sense of helplessness. They maintain, fourth, that the comparison is invalid because, unlike terrorism, bathtubs provide benefit.

And finally, they contend that terrorism costs are peculiarly high, particularly in a democratic society, because the fears it generates will necessarily need to be serviced by policy makers, and this pressure forces, or inspires, them to adopt countermeasures, both foreign and domestic, that are costly and sometimes even excessive.

In this article, we examine these five propositions and find all of them to be wanting. In the process, we conclude that terrorism is rare outside war zones because, to a substantial degree, terrorists don’t exist there. In general, as with rare diseases that kill few, it makes more policy sense to expend limited funds on hazards that inflict far more damage.

Terrorism is willed and may well become more destructive

Journalist Jeffrey Goldberg has suggested that “the fear of terrorism isn’t motivated solely by what terrorists have done, but what terrorists hope to do.” Bathtubs are simply not “engaged in a conspiracy with other bathtubs to murder ever-larger numbers of Americans.” However, terrorists “in the Islamist orbit,” he insists, “seek unconventional weapons that would allow them to kill a far-larger number of Americans than died on Sept. 11.” 6 Or as Janan Ganesh of the Financial Times puts it, “Bathroom deaths could multiply by 50 without a threat to civil order. The incidence of terror could not.” 7

Thus far, 9/11 stands out as an extreme outlier: scarcely any terrorist act, before or after, in war zones or outside them, has inflicted even one-tenth as much total destruction. That is, contrary to common expectations, the attack has thus far been an aberration, not a harbinger.8 And al-Qaeda central, the group responsible for the attack, has, in some respects at least, proved to resemble President John Kennedy’s assassin, Lee Harvey Oswald—an entity of almost trivial proportions that got horribly lucky once. The tiny group of perhaps 100 or so does appear to have served as something of an inspiration to some Muslim extremists. They may have done some training, may have contributed a bit to the Taliban’s far larger insurgency in Afghanistan, and may have participated in a few terrorist acts in Pakistan. In his examination of the major terrorist plots against the West since 9/11, Mitchell Silber finds only two—the shoe bomber attempt of 2001 and the effort to blow up transatlantic airliners with liquid bombs in 2006—that could be said to be under the “command and control” of al-Qaeda central (as opposed to ones suggested, endorsed, or inspired by the organization), and there are questions about how full its control was even in these two instances, both of which, as it happens, failed miserably.9 And, although some al-Qaeda affiliates have committed substantial damage in the Middle East, usually in the context of civil wars, their efforts to carry out terrorism in the West have been rare and completely ineffective.10 Even under siege, it is difficult to see why al-Qaeda could not have carried out attacks at least as costly and shocking as the shooting rampages (organized by other groups) that took place in Mumbai in 2008 or at a shopping center in Kenya in 2013. Neither took huge resources, presented major logistical challenges, required the organization of a large number of perpetrators, or needed extensive planning.

However, there is of course no guarantee that things will remain that way, and the 9/11 attacks inspired the remarkable extrapolation that, because the terrorists were successful with box cutters, they might soon be able to turn out weapons of mass destruction— particularly nuclear ones—and then detonate them in an American city. For example, in his influential 2004 book, Nuclear Terrorism, Harvard’s Graham Allison relayed his “considered judgment” that “on the current path, a nuclear terrorist attack on America in the decade ahead is more likely than not.” 11 Allison has had a great deal of company in his alarming pronouncements. In 2007, the distinguished physicist Richard Garwin put the likelihood of a nuclear explosion on an American or European city by terrorist or other means at 20 percent per year, which would work out to 91 percent over the elevenyear period to 2018.12

Allison’s time is up, and so is Garwin’s. These oft-repeated warnings have proven to be empty. And it is important to point out that not only have terrorists failed to go nuclear, but as William Langewiesche, who has assessed the process in detail, put it in 2007, “The best information is that no one has gotten anywhere near this. I mean, if you look carefully and practically at this process, you see that it is an enormous undertaking full of risks for the would-be terrorists.” 13 That process requires trusting corrupted foreign collaborators and other criminals, obtaining and transporting highly guarded material, setting up a machine shop staffed with top scientists and technicians, and rolling the heavy, cumbersome, and untested finished product into position to be detonated by a skilled crew, all the while attracting no attention from outsiders.

Nor have terrorist groups been able to steal existing nuclear weapons—characteristically burdened with multiple safety devices and often stored in pieces at separate secure locales—from existing arsenals as was once much feared. And they certainly have not been able to cajole leaders in nuclear states to palm one off to them—though a war inflicting more death than Hiroshima and Nagasaki combined was launched against Iraq in 2003 in major part under the spell of fantasies about such a handover.14

More generally, the actual terrorist “adversaries” in the West scarcely deserve accolades for either dedication or prowess. It is true, of course, that sometimes even incompetents can get lucky, but such instances, however tragic, are rare. For the most part, terrorists in the United States are a confused, inadequate, incompetent, blundering, and gullible bunch, only occasionally able to get their act together. Most seem to be far better at frenetic and often self-deluded scheming than at actual execution. A summary assessment by RAND’s Brian Jenkins is apt: “their numbers remain small, their determination limp, and their competence poor.” 15 And much the same holds for Europe and the rest of the developed world.16 Also working against terrorist success in the West is the fact that almost all are amateurs: they have never before tried to do something like this. Unlike criminals they have not been able to develop street smarts.

Except perhaps for the use of vehicles to deliver mayhem (though this idea is by no means new in the history of terrorism), there has been remarkably little innovation in terrorist weaponry or methodology since 9/11.17 Like their predecessors, they have continued to rely on bombs (many of which fail to detonate or do much damage) and bullets.18

There is another aspect to this argument. It is held that, whereas the number of bathtub deaths does not fluctuate much from year to year, terrorism deaths are not very evenly distributed over time and this quality somehow makes the phenomenon unpredictable and unstable. It is a “fat-tailed distribution” in which there are many small events and a few “outliers that are really important.” 19 Thus, we should give up, suggests Bloomberg’s Justin Fox: “Five or 10 or even 50 years of data isn’t necessarily enough to allow one to predict with confidence what is going to happen next year.” 20

The frequency and destructiveness of terrorism and terrorism cases is indeed anything but uniform. In 2016 there were some two dozen cases of Islamist terrorism in the United States and 49 deaths. In 2008 there was only one case and no deaths.21 However, many natural hazards show the same pattern as terrorism. For example, the frequency and destructiveness of tornados range widely: the death count can vary by up to twenty-fold from year to year. Moreover, they are also far more likely than terrorism to kill. However, the lumpiness doesn’t preclude sensible analysis.

Concern about this unevenness, as bathtub critics Justin Fox and Kenneth Anderson both note, stems from a book by Nassim Nicholas Taleb that stesses the importance of extreme events which he calls “Black Swans.” Taleb argues that “almost everything in social life is produced by rare but consequential shocks and jumps” and “our world is dominated by the extreme, the unknown, and the extremely improbable.” 22 However, Taleb’s account focuses on those unexpected and emotion-engaging events and phenomena (like 9/11) that became consequential (and therefore Black Swans), while ignoring ones that failed to do so. It accordingly suffers from what is called “selection bias.” Moreover, insofar as Black Swan events carry an “extreme impact,” this quality derives not so much from their unexpectedness or from the emotions they initially trigger as from the reaction or overreaction they generate. These reactions are sometimes as unexpected as the event itself, and often they do not correlate well with the event’s size or with its objective historical importance. Moreover, although some unexpected and emotion-engaging events do have considerable consequences, much consequential development in human history—probably most of it—stems not from such events, but from changes in thinking and behavior that are decidedly gradual and often little noticed as they occur.23