Does terrorism work?

Why "what would you do?" isn't a powerful argument.

Grievance as the cause of fundamental ills is overstated in many domains, but it is particularly overstated in the context of terrorism. “What would you do if you were XYZ oppressed group?” is a common refrain in this line of thinking, the implication being that terrorism is the only rational response. At its most despicable, it is used to justify (or “explain”) all manner of evil. Robert Fisk, the recently departed columnist, is a case in point. In the aftermath of being violently attacked in Afghanistan, he repulsively wrote:

I couldn’t blame them for what they were doing. In fact, if I were the Afghan refugees of Kila Abdullah, close to the Afghan-Pakistan border, I would have done just the same to Robert Fisk. Or any other westerner I could find.

The premise of the question “what would you do” has as its underlying premise that the relevant group is in fact oppressed, or that the force they are fighting against is somehow unjust. That premise is not usually true in the context of the claims often made, but assuming it was true, what is the correct answer? If I was an Effective Oppressee, I’d look at the evidence of what did and did not work.

Max Abrahms in his 2006 study looks at 28 terrorist groups designated by the U.S. Department of State as terrorist organisations between 2001 and 2006. In total, those groups have 42 strategic aims. How often were those strategic aims met? A mere 7%. But there’s more: the targets of these groups is predictive of whether these groups are successful or not. His findings are pretty stark:

This isn’t an odd finding either. Abrahms (2012) summarises the research following his initial study:

Jones and Libicki (2008) then examined a larger sample, the universe of known terrorist groups between 1968 and 2006. Of the 648 groups identified in the RAND-MIPT Terrorism Incident database, only 4 percent obtained their strategic demands… Cronin (2009) has reexamined the success rate of these groups, confirming that less than 5 percent prevailed.

[…]

Gaibulloev and Sandler (2009) analyze a dataset of international hostage crises from 1978 to 2005. They exploit variation in whether the hostage-takers escalate by killing the hostages instead of releasing them unscathed. The study finds that hostage-takers significantly lower the odds of achieving their demands by inflicting physical harm in the course of the standoff.

[…]

Chenoweth and Stephan (2008, 2011) provide additional empirical evidence that meting out pain hurts non-state actors at the bargaining table. Their studies compare the coercive effectiveness of 323 violent and nonviolent resistance campaigns from 1900 to 2006. Like Gaibulloev and Sandler (2009), the authors find that refraining from bloodshed significantly raises the odds of government compliance even after tactical confounds are held fixed.

So, “what would you do?” is not so powerful a question when the response attempting to be elicited is to do something woefully ineffective (leaving aside the obvious moral imperative). The truth is that grievance-based explanations would imply much greater levels of terrorism; they rarely if ever follow the variations of terrorism over time; and would in any event justify the targeting of terror groups themselves.

Indeed, this very point is why Charles Kurzman wrote his book The Missing Martyrs: Why Are There So Few Muslim Terrorists. As Kurzman explains ‘Islamist terrorists have managed to recruit fewer than fewer than 1 in 100,000 Muslims since 9/11’ (p.11). More recent data is that 0.0066% of Muslims are active Jihadists. That numbers are low is aptly contained in the fact that Al Qaeda planned to carry attacks on the West Coast of the U.S during 9/11 but they could not find anyone to carry it out. Any unifying attempt to explain the causes of such behaviour must start with the extreme abnormality of it.

A related point in all of this is that when these groups carry out terrorism, the apparent appropriate response is not force, but negotiation. Northern Ireland is held up as an example which, in tandem with the “what would you do?”, requires the apparent Oppressor to tie it hands, whilst the Opressee carries out terrorism to get the Oppressor to the table. This “Northern Ireland” model is used in various contexts, including trying to cajole Israel into negotiating with terrorist organisations like Hamas.

John Bew explains how such a narrative is fundamentally flawed:

The terms of the dialogue between the British government and the IRA were set by the war that preceded it. By the early 1990s, it had become increasingly clear that the IRA had been heavily infiltrated by informers and was subject to a strategy of containment by the British security services… While the IRA was far from beaten, there is incontrovertible evidence that counterterrorism operations were taking a heavy toll on the organisation. In military terms, it was a movement that was squeezed and weakened, and which had lost momentum. In political terms, it was a movement that had the potential for electoral expansion but which was being held back by its military actions

Contrary to the grievance-based explanations, the alleged “oppressive” actions in question (e.g. military interventions, security-based occupations and kinetic strikes) regularly lead to reductions in terrorism and violence: a finding that is replicated in Mali, Gaza, the West Bank, Iraq, Yemen, Afghanistan, the ongoing decimation of ISIS, drone strikes across the Durand line and even in Columbia.