Academia Denver Dialogues Terrorism

Evidence-Based Policy Meets the Politics of Terrorism

By Erica Chenoweth for Denver Dialogues.

Plakat_La_Terroriste

“La Terroriste” (ca. 1910) depicting the bombing of a Russian official’s car by the Polish Socialist Party’s armed wing. Photo: Polish Central Archives of Historical Records.

Today this tweet is making the rounds:

Screen Shot 2016-09-06 at 6.55.09 AM

Indeed. We know.

Over the past ten years, lots of folks have reiterated the triviality of the terrorism threat relative to other kinds of public safety or health hazards. When new attacks happen, some scholars take to the airwaves to reassure the public that the threat is not as severe as it might appear that day—and that any response ought to be proportional.

Do such claims and reassurances make a difference in guiding policy responses to terrorism? I’m doubtful.

One particular incident was fairly enlightening in this regard. Last fall after the alleged bombing of Metrojet 9268, I suggested that if IS had brought down the plane, it wasn’t as much of a game-changer as some might think. The piece was something of a response to those who suggested that attacks of this kind should precipitate major shifts in counterterrorism policy. A few days after writing the piece, I gave a radio interview in which I defended my view and offered further evidence that terrorism was a totally overblown threat—and that our misplaced obsession with it yielded all kinds of misguided, costly policies.

In the green room prior to the show, I chatted about this claim with the show’s other guest, a former FBI and homeland security official. He said he bought the statistics about the triviality of the terrorism threat, but that those statistics don’t matter. His basic point was that evidence does not drive policy responses to terrorism. Politics does. If politicians want to avoid accusations of falling asleep at the switch, they have to react to terrorism as if the media hype about it were true. I noted that this, of course, produces a vicious cycle, wherein politicians consistently reinforce rather than contest the narrative that terrorism is an existential threat to society (a process also theorized and documented in various scholarly works). Although he basically agreed with me, this fellow concluded that calling for proportionality based on such statistics was informative but essentially useless from a policy perspective.

This anecdote contains a powerful lesson for those who want to do engaged work in the traditional policy domain. There is increasing pressure for social scientists to do policy-relevant research, and there are increasingly more resources for them to do so. But fewer people pay attention to whether their work is “politics-relevant.” Many formal decisionmakers don’t give a damn whether research points toward (or against) a specific policy recommendation, unless it also points to the way in which that decisionmakers can advocate for and implement that evidence-based approach while inoculating themselves from potential political consequences. The terrorism domain provides an extreme example of this, but other hot-button policy issues have similar dynamics.

So for those scholars wanting to pursue effective policy engagement in their careers, it’s not just about interpreting a complex argument into digestible chunks and bringing attention to your work. It’s also about demonstrating how policymakers can use the information while navigating the political landscape. And it’s about informing the public–one’s community, activists, local media, and other changemakers–so  that they can pressure their elected officials to embrace and advocate for evidence-based policy.

In other words, to be an effective engaged scholar, “politics-relevance” is essential too.

9 Comments

  • Excellent argument Erica! I was so biased and misinformed. Now I can finally go to Molenbeek in Brussels without any fear whatsoever. I will make sure that me and my gay partner will wrap ourselves in the rainbow flag and enjoy a safe and pleasant stroll around this peaceful suburb. Surely, nothing will happen to us 😉 Cheers for enlightening the uneducated and ignorant masses !

  • Now these are some interesting statistics of two Americans per year when I compare some real numbers.
    Leaving September 11 out of the picture you can come to that conclusion. Factoring in September 11 which was also Jihadist Muslim immigrants that adds 2996 to the tally which brings the statistics up to 202 Americans killed domestically on average for the last 15 years.

    You can play the numbers to your advantage, but the facts are that it’s an average of 202 killed per year.

    • Brad, as the figure notes, the graph is basing this analysis on a 10-year average, which means that 9/11 is not included. There are good statistical reasons not to include 9/11 in any estimates of fatality averages, since it was an outlier (and a catalytic event that has affected both propensities to defend the US against further 9/11-like attacks and propensities to continue attacking the U.S.). Anyway, for various reasons it is often advisable to avoid counting outliers or high-severity terrorist attacks when they are incredibly infrequent. That’s why many political scientists using moving averages or weighting schemes rather than simple averages (see, for instance, http://jcr.sagepub.com/content/51/1/58.abstract). Saying the average is 208 per year is wrong. Saying that about 3,000 were killed in 2001 and that it’s been < 2 per year every year since is accurate. There are also good practical reasons for policymakers to know an outlier when they see one. This could help them better formulate proportional and cost-effective policy responses, rather than become totally obsessed with responding to / preventing something that is so rare it is unlikely to recur.

  • isn’t there a more potent argument against the use of these statistics? the infrequency of deaths from terrorism since 9/11 might be because of counter-terrorism policies. I’m not saying this is true (we can’t know counterfactual history) but an intelligent analysis would have to consider this entirely reasonable point.

Leave a Comment