DSBlog

Comments (0)

Can terrorism be analyzed statistically in a way that improves its predictability? According to a recent piece in Miller-McCune, a wonky bimonthly journal that pokes into all corners of the scientific research community, the answer is yes.

The Miller-McCune article cites the work of Aaron Clauset, a physicist and Assistant Professor of Computer Science at the University of Colorado at Boulder, who notes: ““It all comes back to this idea that you don’t have to know all the details of processes in order to understand how the interactions lead to patterns.”

Clauset found when studying terrorist acts over the past four decades (using the Oklahoma City-based Memorial Institute for the Prevention of Terrorism database) that, “When you start averaging over the differences, you see there are patterns in the way terrorists’ campaigns progress and the frequency and severity of the attacks. …This gives you hope that terrorism is understandable from a scientific perspective.”

The article continues: “After mapping tens of thousands of global terrorism incidents, [Clauset] and his collaborators have discovered that terrorism can be described by what mathematicians call a power law. Unlike the familiar bell curve — where most events tend to cluster around the average, with only a small number at the margins — a power law distribution produces a wide range of highly dissimilar events. ...Massive acts of violence, like 9/11 or the devastating 1995 bombing of the U.S. embassy in Nairobi, obey the same statistical rules as a small-scale IED attack that kills no one, Clauset’s work suggests. ‘The power law form gives you a very simple extrapolation rule for statistically connecting the two,’ he says.”

There are some intriguing aspects to Clauset’s work. But while he admits that a broad-brush approach looking at “big patterns” serves long-range planning better than more immediate security risk management priorities, we would question some of the underlying assumptions. First, there is an implication that because the scale of terrorist attacks follows a scaling law that small-scale and large-scale attacks are somehow similarly motivated. The evidence does not demonstrate that claim. There are several composite systems with independent components whose aggregate ‘score’ appears smoothly varying. Fitting a curve to the data does not prove the theory.

Second, Clauset looks mainly at the size and scale of attacks, and maintains that “there are a couple of things terrorist organizations can’t change.” But there is less focus on two additional variables – geographic scope (what targets, and where) and time (frequency of attacks and variances therein over time) – that can change quite dramatically. And if he looked at terrorist incidents going back 200 years (admittedly a much more difficult task given the lack of good data) he might have discerned variances not apparent in his narrower 40-year timeframe.

We also wonder about the overall utility of this exercise. Clauset’s fundamental claims are: 1) that the size distribution of terrorist attacks (measured in fatalities) follows a regular pattern (power law) over different terrorist organizations over time; and 2) that the frequency with which terrorist organizations carry out attacks is related to the size of the organizations.

Neither of these claims is particularly surprising, but we’re not convinced that noticing such patterns makes it any easier to predict future behavior. The goal of security professionals is to mitigate risk, a task that requires an understanding of the interplay between attack size, geographical scope and time, among other things. Some of those critical variables don’t readily lend themselves to statistical analysis. 

(HT: Slashdot for reposting the Miller-McCune piece.)

––––––––––––––––

Digital Sandbox is the leader in public safety risk management, providing analytic tools and information products to government agencies and large enterprises, for optimizing risk-based strategic, policy, and budgetary decisions.