How Biases Save Us From Giving in to Terrorism

Terrorists are hampered by biases as much as the rest of us. In a Wired commentary "The Evolutionary Brain Glitch That Makes Terrorism Fail" Bruce Schneier discusses the interesting findings of Max Abrams in his paper Why Terrorism Does Not Work (International Security, Vol. 31, No. 2 (Fall 2006), pp. 42–78).

Basically, terrorists run into trouble because people use correspondent inference theory to infer the intentions of others: the results of their actions are assumed to be concordant with their intentions. If a person sweeps the floor we assume he wants it clean (but he could just be working off excess energy). If somebody hits somebody else, we assume the intention was to harm (but it could just be a game). Similarly, people infer that the horrific deaths of innocents is the primary motivation of a terrorist – which likely leads to a misunderstanding of the real goals of the terrorist.

This is bad news for terrorism as an effective coercive means to political or social ends. Although the terrorist can state his demands and goals, people will tend to assume that he is just a sadist rationalising. Possibly a dangerous sadist one has to occasionally acquiesce to, but the goals are not seen as essential to him. His "real" goals are assumed to be the destruction of society, and this makes accepting demands less favorable. Abrams finds empirical support for this in that terrorists are much more likely to succeed with their demands if they focus their attacks on military goals rather than civilian ones, and if they have minimalist goals (evicting a foreign power, winning control of a piece of territory). Attacking civilians or wanting to change the world makes people assume the intention is something else.

This analysis assumes bias among the non-terrorists making them unwilling to play along, but clearly there are plenty of biases among the terrorists too. The correspondence makes them impute evil intentions to governments that behave clumsily or violently. The emotional salience of terror probably introduces a lot of availability bias, impact bias makes terrorists overestimate the emotional effect of their actions, groupthink is likely pretty big within terrorist grooming communities and so on.

It seems that one could probably analyse terrorism in terms of cognitive biases quite fruitfully. Whether that will lead to ways of reducing terrorism is another matter. Maybe unbiased terrorists will simply see that the Bayesian thing to do is simply to go home since terror doesn’t work efficiently – or they would start making non-hyperbolic long-term plans for surgical strikes that simply cannot be misunderstood. Conversely, maybe terrorists could be incited to bias themselves into inefficiency, but highly biased people can occasionally be dangerous. Maybe the real aim should be an unbiased anti-terror strategy – but as long as politicians and public are biased they will likely see the unbiased strategy as wrong.

GD Star Rating
a WordPress rating system
Tagged as:
Trackback URL:
  • Hopefully Anonymous

    Andres,
    I’m generally a big fan of your analysis, but I think this one misses the mark (both yours and the Wired article, and perhaps the actual study) in important ways.

    First, I don’t see acknowledgment by any parties of a non-transparent aim that’s very effectively served both by terrorists and governments (categories that are a bit arbitrarily distinguished from each other) when violent and emotionally charged conflict and attacks are engaged in: representational privilege.

    Here I think Straussian analysis is fruitful: my understanding is that Strauss claimed that governments engage in foreign conflicts as an effective way to succeed in domestic politics. Similarly, I think two organizational agents can engage in conflict with each other on the global stage in such a way that both benefit, to the detriment of organizational agents that don’t play leading roles in those conflicts. Thus I think there have been domestic politics benefits that have benefitted both conflicting parties in Bush’s conflict with Chavez, and in Bush’s conflict with Bin Laden. If one doesn’t analyze representational benefits for 2 parties to perform conflict with each other relative to 3rd, 4th, and Nth party competitors, then I don’t think one has fully captured the benefits of such conflicts, and thus I think one is missing important elements in analyzing them.

  • http://econstudentlog.wordpress.com US

    Yet another case where bias might not be such a bad thing on net.

    Even if the bias among non-terrorists does help a little, one should not underestimate the impact of terrorism either:

    http://www.creedexperiment.nl/pc2007/ocs/viewpaper.php?id=403

    http://www.creedexperiment.nl/pc2007/ocs/viewpaper.php?id=153

  • anon

    “my understanding is that Strauss claimed that governments engage in foreign conflicts as an effective way to succeed in domestic politics”

    That sounds like a conspiracy theory… I wonder if conspiracy theorists often feel like they know more about what is happening in the world than others.

  • Hopefully Anonymous

    Anon,
    I think it’s definitional either that conspiracy theorists do believe they know more about what is happening in the world than others, or that they think the entire world is in a second conspiracy against the conspiracy theorist in pretending that they’re unaware of the primary conspiracy that the theorist believes exists.

  • michael vassar

    I think that history provides plenty of datum suggesting that when political leaders (such as terrorists) accomplish their goals via violence they then formulate new goals and attempt to accomplish these through violence. Even worse, in the case of loosely organized groups, in the rare cases when they do turn to peace (Nasser for instance), they are prone to replacement by more violent followers.
    http://dilbertblog.typepad.com/the_dilbert_blog/2007/06/how_rational_ar.html
    discusses this fairly sensibly.
    Counter-examples are not readily obvious.

    Why is it automatically a bias to know or believe this? I would consider it to be a well documented fact.

  • http://www.aleph.se/ Anders Sandberg

    The papers US cite show that terrorism has an impact on government policy in the sense that it reduces the respect for human rights and increases the risk of cabinet failure. But unless the goals of terrorists are to reduce human rights or topple a particular cabinet (sounds just like correspondent inference), that is not success. I can imagine bin Laden sitting in a cave watching CNN, getting increasingly annoyed that instead of diminished support for Israel he just gets a lot of human rights abuse and badwill for Bush and Blair – *that* was not the point and does not help him much! On the other hand, correspondence theory explains that the public will think that that was indeed his goals, and furthermore than western human rights abuses are an expression of intention too (“Never ascribe to malice that which can be explained by incompetence”).

    There can certainly be benefits in engaging in a conflict that have little to do with the stated goals, for example increased domestic cohesion. But while these benefits might be a motivator to engage their role would plausibly be overstated by outside observers in analogy with the intention inference. This seems related to conspiracy theory or cynism: rather than believe stated goals one assumes the real goals to be hidden, domestic or directly self-serving. We have been seeing a ridiculous amount of this kind of explanations of recent US interventions in the Middle East. While oil, cohesion or Haliburton profits might be motivating, it seems excessive to assume (as many people I have talked with do) that one or more of these constitute the main motivation, at least without further evidence.

    I have noted that in general reasonably well-educated westerners these days have a hard time accepting a stated reason for *any* historical war. The sophisticated thing to do these days is to assume self-serving interest among the players, not that they actually believed in what they were claiming. But this may often be our bias in evaluating intentions combined with the illusion of transparency (as well as signalling being an unafraid and honest person attacking hypocricy) rather than an unbiased judgement. One sign of this is how we tend to not look for any hidden agendas when the war had an obvious and stated self-serving goal.

  • Stuart Armstrong

    Very interesting post Anders.

    Maybe unbiased terrorists will simply see that the Bayesian thing to do is simply to go home since terror doesn’t work efficiently – or they would start making non-hyperbolic long-term plans for surgical strikes that simply cannot be misunderstood.

    Or they will continue in their terrorist acts, but these acts will bear less and less connection with their political goals (as happened to the Anarchists and Communist movements in some countries). Al Queda in Mesompotamia might be the prototype of the unbiased terrorist – their long term goals are vague pipe-dreams. And in the short term, terrorism seems to be both their means and their goals.

    If our biases cause us to infer

    infer the short-term consequences of terrorism […](are) the objects of the terrorist groups,

    then we will end up with less terrorists, but those we do have will truly have terrorism as their short term goals.

  • Stuart Armstrong

    for surgical strikes that simply cannot be misunderstood.

    That sounds similar to the IRA’s tactic of planting a bomb and phoning the newspaper shortly before it exploded. This made clear that their objective was not mass casualties, and may explain why they were ultimately (semi-)successful.

    Surprising that other terrorist organisations have not copied this idea. They do seem willing to copy any effective way of killing their enemy, however, and any effective way of organising their internal structure.
    So I’d hazard a guess that terrorists are biased in a way that allows them to copy effective tactics, but not effective political strategies.

  • http://www.aleph.se/andart/archives/2007/07/biases_and_terrorism.html Andart

    Biases and Terrorism

    I blog on Overcoming Bias about How Biases Save Us From Giving in to Terrorism. The short of it is that people have a cognitive bias tthat makes them assume that actions and their outcome are the intention of the…

  • Stuart Armstrong

    The sophisticated thing to do these days is to assume self-serving interest among the players, not that they actually believed in what they were claiming.

    That’s the fashionable thing to do in all walks of life nowadays. It started in politics and economics, and now it’s spread.

  • anon

    I like the idea in some situations of using consequences to infer intentions, but in the case of terrorism, I don’t think that anyone who hears about a suicide bombing would conclude that their sole intention was to harm people. I seriously doubt that most people would think that the primary motivation of a terrorist would be to cause horrific deaths.

    Most people, I would think, recognize that there are other primary goals and that the terrorists attempt to use violent acts as a means to achieve their goals.

    From our perspective, it seems inefficient and irrational for them to continue, but clearly they must see things differently. Another possible explanation for the tendency of people to not see terrorist acts as related to their stated intetntions is that we simply have trouble comprehending how the acts are related to their stated intentions. From our perspective, it just doesn’t make sense. It seems unlikely that they share our conclusions regarding their methods.

    Attempting to see from their perspective is undeniable evidence that we rise above drawing conclusions from consequences alone and our thoughts are not consistent with this theory. In fact, I believe that from a young age we are taught to try to see from another person’s perspective before coming to conclusions.

    Tell me if it seems I misunerstood correspondent interference theory, but it seems to imply that we are not capable of incorporating prior/outside information and attempting to see from another’s perspective.

  • http://www.aleph.se/ Anders Sandberg

    I don’t think correspondent inference theory says we are incapable of seeing things from the other’s perspective, it just says that we tend to be biased or make mis-judgements about it. My initial reaction to a suicide bomber might be a horrified thought that he was an evil person seeking the deaths of innocents. A bit of reflection might suggest other goals, but that initial evaluation will likely bias my thinking anyway. If I am upset enough I might just discount all other goals, and state that given the evil perpetrated they don’t matter at all. In less emotionally charged situations seeing the other’s perspective may be much easier.

    Biases are troublesome because they are invisible to us. We discount or misunderstand the real goals of others, then act on this biased information – and the other person will then discount or misunderstand our already biased intentions. Even a small bias might set up a feedback that neither of us wants.

  • http://econstudentlog.wordpress.com US

    Anders, I was merely pointing to some of the effects of terrorism, I did not state whether or not these effects were intentional or not on part of the current set of terrorists (or their sponsors/political leaders/religious leaders/ect.). Whether or not these consequences are intentional or not, we do observe them, and that is my main concern.

    If one was in fact interested in undermining human rights and political stability in a Western country, judging by these papers, terrorism is one of means to achieve this end that seems to work. No matter what the current set of terrorists wants, this is not good news.

  • Hopefully Anonymous

    Anders,
    You wrote:
    “There can certainly be benefits in engaging in a conflict that have little to do with the stated goals, for example increased domestic cohesion. But while these benefits might be a motivator to engage their role would plausibly be overstated by outside observers in analogy with the intention inference. This seems related to conspiracy theory or cynism: rather than believe stated goals one assumes the real goals to be hidden, domestic or directly self-serving. We have been seeing a ridiculous amount of this kind of explanations of recent US interventions in the Middle East. While oil, cohesion or Haliburton profits might be motivating, it seems excessive to assume (as many people I have talked with do) that one or more of these constitute the main motivation, at least without further evidence.

    I have noted that in general reasonably well-educated westerners these days have a hard time accepting a stated reason for *any* historical war. The sophisticated thing to do these days is to assume self-serving interest among the players, not that they actually believed in what they were claiming. But this may often be our bias in evaluating intentions combined with the illusion of transparency (as well as signalling being an unafraid and honest person attacking hypocricy) rather than an unbiased judgement. One sign of this is how we tend to not look for any hidden agendas when the war had an obvious and stated self-serving goal.”

    Whether or not something is the sophisticated thing to do doesn’t bear on whether it’s an accurate explanation for a phenomenon. And a theory that a domestic agent makes foreign policy decisions (including performing conflict) to maximize domestic political outcomes is different in kind than a theory that a domestic agent makes these decisions to maximize on oil or haliburton profits. For one thing, domestic political agents are rewarded and punished with domestic political success and failure, so I think the useful analytical starting point may be to see the degree to which their primary motivations are domestic political success.

    To say “This seems related to conspiracy theory or cynism: rather than believe stated goals one assumes the real goals to be hidden, domestic or directly self-serving.” but I’m not sure the relevance of this point. I think the reasonable option is that a goal is either stated or that it’s unstated. Are your using the words “conspiracy theory”, “cynism”, and “hidden” pejoratively? If so, why?

    Probably sometimes, in some cases, an individual or organization will have stated goals which match actual goals. Probably sometimes, in some cases, an individual or organization will have unstated goals which don’t match stated goals. Sometimes these unstated goals may be internally transparent to the individual or organization (definitional real conspiracy), and sometimes these unstated goals may not be internally transparent to the individual or organization (incomplete self-knowledge of internal motivations). How actual vs. stated goals and internal transparency sort out seem to me to be investigative questions, and regardless of what is “sophisticated” to do, or what is the Anders Sandberg thing to do, I think we should retain some space from analytical flexibility rather than cast any one of these analytical reference points pejoratively.

    For example, you identified a potential bias at play in automatically assuming every agent is engaging in self-serving interest. I think there is another potential bias at play in assuming an agent’s externally (or even internally) stated goals are their actual goals, particularly if as the observer one may also be vested in believing that those are their goals.

    That you seem to uncritically sort organizational agents into the categories of “governments” and “terrorists” indicates to me that your analysis might be rooted more in conventional representational pageantry than a search for the best models of these social phenomena. The approach seems to be “I represent the western archetype, helping to protect the west from Bin Laden, a representative of the radical islamic terrorist archetype” -thus you may yourself be vested a bit in the representational privilege that this construct confers, and may have more difficulty analyzing it in terms of conferring such privilege to certain global subpopulations and not others. I’m bring this up not as an ad hominem, but as an indication of one reason why a preference to believe a party’s stated goals (internally transparent or not) match their actual goals may exist -one may consider onesself to be a stakeholder in representational privilege on the opposite end of a performed conflict with that other party, where the representational privilege would lose salience if it was transparent that both parties were performing the conflict primarily to benefit from such representational privilege (regardless of whether they where internally transparent about that being the primary motivator or not).

    Sorry if this isn’t clearly organized, but I trust you at least Anders are bright enough to parse through this morning jumble.

  • TGGP

    The “war for oil” argument never made sense to me because we were getting more oil from iraq before the invasion than after, and at much cheaper prices. I think think it gives too much credit for the actual architects of the war. Oddly enough, the one place I found that offers this sort of sensible (since I agree with it, of course!) view is this column by Michael Neumann at CounterPunch.

    The ultimate goal of most organizations with a military wing is rarely to kill a bunch of people, but I think that often does become a sort of operating goal. I think a lot of people joined the army after 9/11 because they were angry and wanted to kill a bunch of ragheads, and a lot of people join al Qaeda because they are angry and want to kill a bunch of Americans. The actual leaders of al Qaeda proper have tried to prevent the franchise in Iraq from turning it into a Sunni-Shia war, but that’s a sensible approach, and not want the enthusiastic recruits on the ground wanted. There is a satisfaction to be gotten from seeing your enemies get taken down a notch, even if it doesn’t accomplish anything, like when the football team you’re rooting for sacks the other quarterback even if a flag on the play makes it meaningless (I actually don’t really watch football much and don’t know how good of an example that is).

  • michael vassar

    I believe that late 19th century anarchists were among the earliest terrorists in the modern sense, and that human rights abuses and cabinet failures were high among their explicit goals.

  • http://www.aleph.se/andart/ Anders Sandberg

    It is noticeable that none of the terrorist groups discussed in Abrahms list have general anarchy as a goal. Overall, subverting society and ushering in an era of chaos (that will lead to utopia, of course) doesn’t seem to be very popular these days. Maybe because it was so obvious that this maximalist goal completely failed. Most terrorist groups probably have a hard time telling exactly how well they are doing.

    Hopefully Anonymous: I think you are reading far too much into my intentions. I think intentions can be viewed as a vector sum of goals, where some goals are publicly stated and some aren’t. Assuming that the main goals are usually unstated (or stated) would constitute an individual or cultural bias (I don’t think there is any human universal in this). Even trying to fix these priors with historical data will be contentious, leaving people disagreeing about who is really biased. I tend to think that in my community of western intellectuals people are biased towards overestimating the impact of hidden goals and underestimating the impact of overt goals, and no doubt they think I’m naive and biased in the opposite way.

  • michael vassar

    Evidence for self-serving war-making by US presidents

    http://www.marginalrevolution.com/marginalrevolution/2004/04/war_politics.html

    In general, the economic methodology is to assume self-interested behavior, where a person’s interests can be to some degree inferred by their accomplishments so far, especially if those accomplishments are difficult. This approach may not always match reality, but it seems to frequently be fruitful.

  • Hopefully Anonymous

    Anders,
    I understand that you think that you think (not a typing repitition error) that people are motivated by their stated goals more often than your academic peers do, and perhaps than the general population does.

    Beyond that, I hope you share an interest in me in what the best empirical data can tell us on this topic, even if it contradicts one or both of our intuitions.

  • Hopefully Anonymous

    “an interest in me” should read “an interest with me”.

  • TGGP

    Michael Vassar, the first to actually coin the term “terrorism” were French revolutionaries carrying out the “Reign of Terror” in reference to themselves. I agree that the non-state actor and “propaganda of the deed” stem from 19th century anarchists.

    Anders Sandberg, John Robb claims that today’s “global guerrillas” using “open source warfare” in a “bazaar of violence” intentionally try to hollow-out a state and cause system dysfunction without actually completely collapsing the state or replacing it. I think he greatly exaggerates all this. Oddly enough (given its name), the somewhat related blog Coming Anarchy has a more optimistic take on the ability of state actors (with the Victorian era empires as a model) to deal with this issue.

  • Keith Elis

    Anders, the odds of any one person being killed in a terrorist attack in their lifetime is very low. The bias literature indicates that most people would tend to *under*estimate even these very low odds. Add to this that people would tend to *over*estimate the odds of the government preventing future terrorist attacks, and I would predict that on average we individually feel our lives are quite safe from terrorism. (see, e.g., http://www.sciencedaily.com/upi/?feed=TopNews&article=UPI-1-20070712-17370700-bc-britain-terror-poll.xml)

    If so, terrorists just don’t kill enough civilians often enough to raise the odds of being killed to a worrisome prospect. There’s no need to change policy when on average we don’t feel threatened. This seems a simpler explanation for why we don’t bargain with terrorists.

  • http://www.aleph.se/ Anders Sandberg

    I don’t think people underestimate the odds of being killed by terrorism. As far as I know, the subjective risk literature is pretty unanimous in pointing out that subjective risk tends to be overestimated for improbable but dramatic events, especially if there is an element of moral outrage involved. Maybe some people or nations think they are at less risk than they are and trust their governments to be effective, but it doesn’t seem to fit the UK and the US at least.

    BTW, I found this interesting paper when responding:
    http://www.econ.ku.dk/Events_News/Zeuthen/Zeuthen_2005/Workshop%202005/Papers/Subjective%20risks%20Johansson-Stenman.PDF
    I havent gone into it much, but it seems to be pretty relevant for this blog: how should public policy reflect risk perceptions, when these may be biased?

  • http://zooko.com/log.html Zooko O’Whielacronx

    You wrote: “This is bad news for terrorism as an effective coercive means to political or social ends.”.

    But that is true only for the political or social end of persuading your enemies to change their policies. This is probably not the political or social end that most terrorists are trying to accomplish, and there is reason to believe that terrorism is a very effective means to other ends.

  • http://zooko.com/log.html Zooko O’Whielacronx

    Coincidentally I just happened upon this article, which is quite relevant and authoritative:

    http://www.edge.org/3rd_culture/atran06/atran06_index.html