9
$\begingroup$

Terrorism in general, and suicidal terrorism in particular, is popularly seen as “irrational,” but many economists and political scientists argue otherwise.

This quote is from Terrorism: The relevance of the rational choice model by Bryan Caplan. This paper has been published in 2006 in Public Choice. In relation with the recent terrorist attacks in Paris (Nov. 15), is there any new (economic) development on the question of the rationality of terrorism?

Edit: the title is intentionally short but by saying “is X rational?" I definitely mean "Can X be rationalized?"

$\endgroup$
  • 6
    $\begingroup$ Economics can't answer "is X rational?" That's the domain of psychology. Economics can answer "Can X be rationalised?" $\endgroup$ – EnergyNumbers Dec 9 '15 at 7:36
  • 3
    $\begingroup$ However, any good economist should be a decent psychologist :) $\endgroup$ – Thorst Dec 9 '15 at 7:39
  • 1
    $\begingroup$ @Lasse well that excludes just about everyone then. $\endgroup$ – EnergyNumbers Dec 9 '15 at 7:40
  • $\begingroup$ At least one guy got the Nobel Prize in economics "for having integrated insights from psychological research into economic science"... $\endgroup$ – emeryville Dec 9 '15 at 7:42
  • 1
    $\begingroup$ @denesp I guess there is a "and a good mathematician/logician" part to it to ;) $\endgroup$ – VicAche Dec 9 '15 at 18:16
10
$\begingroup$

I guess you might already know this, but I wanted to add a little detail to the other answers for the sake of any layman who comes here and gets the wrong end of the stick.

What is meant by rationality?

It is important to begin by saying that when economics use the term rational they have in mind a fairly precise definition that does not perfectly coincide with the way the word is sometimes used colloquially:

We say that a decision maker is rational if

  • s/he has preferences that are transitive and complete
  • s/he acts in a manner consistent with optimization with respect to those preferences.

On the first bullet: suppose some one has to make a decision. We say that they have transitive and complete preferences if (i) they are able to rank the alternatives among which they choose from "most preferred" to "least preferred", and (ii) that ranking is internally consistent.

On the second bullet: A person acts in an optimal manner given their preferences if they choose the alternative that is 'most preferred' among the set of all feasible alternatives.

So are terrorists rational?

As others have noted, the literal answer to the question is that we can't really (empirically) verify that terrorists are rational. This is because observing one rational decision does not rule-out the possibility that the individual concerned made an irrational decision at some point in the past (or that they will make an irrational decision in the future). The best we can manage is to look for behavior that violates the two conditions above and take observations of such violations to be evidence of irrationality.

But, to address the spirit of the question: just because voluntarilly committing suicide and murder in the name of your beliefs may seem irrational in the colloquial sense does not mean that it is necessarily so. Indeed, if a person has carefully considered the alternatives available to them and decided that the option they find most attractive is to conduct a suicide attack then this behaviour is entirely consistent with rationality.

Rational terrorism in context

It is important to stress that there is no value judgement implicit in an eocnomist describing behaviour as rational. Just because something is rational behaviour does not mean that it is good or desirable or can be condoned. Rather, it just means that we think we have a systematic way of understanding why people might choose to behave in a certain way. Building a systematic understanding of a phenomenon is an important first step in deciding upon the best way to respond to it.

$\endgroup$
4
$\begingroup$

I don't think we have enough data on that matter. They would be irrational, for example, if transitivity of their preferences does not hold. How do we get their preferences? Through the axiom of revealed preferences (WARP).

We can never say whether someone is rational, we can only say whether someone is irrational (read: his actions are not rationalizable). That is, if we observe 10 decisions from a person, and they are not conflicting, this is not sufficient to say that he is rational. However, if they do conflict, it is sufficient information to say that he is irrational.

You phrased the question about individual terrorists, and not about the collective. Hence, to judge any single one, we would need to observe enough actions to judge his preferences, and see whether we fail to rationalize his actions under a preference ordering. And then, to repeat myself, I don't think we have enough data on that matter.

$\endgroup$
  • $\begingroup$ Thanks @FooBar. If by saying that we don't have enough data you think about data on terrorist's priors and beliefs I definitely agree $\endgroup$ – emeryville Dec 9 '15 at 16:41
  • 2
    $\begingroup$ By sharing @RegressForward's point of view, I doubt we can say that someone is irrational (read his/her actions are not rationalizable) based on conflicting actions. It's "a failure to find a tractable utility function that adequately models their behavior." We definitely need more data on beliefs and priors. $\endgroup$ – emeryville Dec 9 '15 at 16:48
  • 2
    $\begingroup$ @emeryville Yes, your (RegressForward's) points are not that WARP does not hold, it's just that data necessary to find a violation of WARP is really hard, because even if someone apparently contradicts himself with two actions, maybe there's something about the background information (such as priors) that changed between those two events. Indeed, WARP is hard to apply outside of clean laboratory settings. $\endgroup$ – FooBar Dec 9 '15 at 17:51
  • $\begingroup$ @emeryville Interesting. May I add to the discussion? Theoretically, we have preference relations, definitions, and logical rules. In this abstract world, deductions can be made. However, are we not dealing with an empirical, or real world, matter? If so, this lends itself to casting the question into a Bayesian framework. For this reason, the question is probably, should I have degrees of confidence to believe that terrorists are rational? Uncertainty abound, it's conceivable that some terrorists are rational while others irrational. The answer may not be one of extremes (to excuse the pun). $\endgroup$ – Graeme Walsh Feb 16 '16 at 1:27
  • 1
    $\begingroup$ @emeryville Maybe that sounded more formal than I intended. I really just meant, "look at it from a Bayesian perspective". A decent reference is Leamer's Specification Searches book or Zellner's Intro to Bayesian Econometrics. Early chapters in both books have diagrams/flowcharts of a "Bayesian learning process". I basically figured, it would be interesting to ask/answer the question with reference to this particular kind of process in mind. $\endgroup$ – Graeme Walsh Feb 22 '16 at 18:32
3
$\begingroup$

We can not prove terrorists are not rational. We can only have a failure to find a tractable utility function that adequately models their behavior.

On the practical side, one cannot see enough decisions from a person ceteris paribus to confirm much about their rationality. Too much happens over time, without observation, and in a noncontinuous manner. Perhaps we can acquire evidence that their actions have a degree of stochastic noise about them. Or that they do they do not fit model X, or we are omitting a variable Y.

On the theoretical side, for any set of behaviors, Z, one can name an array of utility functions f(Z) which would explain such behavior. I say nothing optimistic about the simplicity, verification, or stability of these utility functions, but they can be modeled nonetheless. We can only eliminate a fraction of those f(Z)'s empirically. We are left with some f(Z)'s that could defend their rationality, it remains possible they are rational.

Absence of evidence is not evidence of absence.

$\endgroup$
3
$\begingroup$

Ron Wintrobe has a book on Rational Extremism, which explains how behavior of terrorists, in particular suicide bombers may be "rationalized". He theorizes that the act of blowing up oneself is a form of corner solution to an optimization problem faced by terrorists.

Here's a preview of one of the chapters of the book: http://economics.ca/2005/papers/0708.pdf

$\endgroup$
-1
$\begingroup$

Terrorists are definitely rational. Each one of them maximizes their utility by blowing themselves up. Their preferences and utilities are defined by their beliefs, of course, not the reality of the situation. Their rationality is bounded, because they may not know that their beliefs are false or misguided.

The open question is how bounded is the terrorists' rationality. There are plenty of works out there, which I admit I haven't read, but Inside Terrorism by Bruce Hoffman looks to be a very popular one.

$\endgroup$
  • 1
    $\begingroup$ As economists, I note that our hinges upon the idea that our subjects are rational... just with perverse utility functions. Note that you cannot prove otherwise, there is always a more complex utility function to model a behavior. $\endgroup$ – RegressForward Dec 9 '15 at 13:06
  • 2
    $\begingroup$ "Terrorists are definitely rational". What does that even mean: Is the concept of rationality defined based on properties of terrorists? $\endgroup$ – FooBar Dec 10 '15 at 15:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.