Terrorism and the media in a game-theoretic setting

Are terrorists rational?

Economics focuses on individual choice, assumes individual rationality and employs mathematical models to analyze human behavior. So, what can economics contribute to the study of terrorism – and how realistic is the assumption of rationality for terrorists?

First of all, definitions of rationality can vary substantially. In a thick sense, rationality implies that an individual is fully informed and perfectly calculates her optimal course of action. In a thinner sense, a rational individual is only required to have stable preferences (Caplan, 2006).

Perhaps, nobody can put it better than Schelling (1960) who argued in defense of the rationality assumption:

[W]e seriously restrict ourselves by the assumption of rational behavior – not just intelligent behavior, but of behavior motivated by conscious calculation of advantages, a calculation that in turn is based on an explicit and internally consistent value system. […] The advantage […] for theoretical development is not that, of all possible approaches, it is the one that evidently stays closest to the truth, but that the assumption of rational behavior is a productive one. It gives a grip on the subject that is peculiarly conductive to the development of theory. It permits us to identify our own analytical processes with those of the hypothetical participants in a conflict; and by demanding certain kinds of consistency in the behavior of our hypothetical participants, we can examine alternative courses of behavior according to whether or not they meet those standards of consistency. The premise of ‘rational behavior’ is a potent one for the production of theory. (p. 4)

However, terrorism seems to be strongly associated with irrationality. A Google search for the two terms “terrorist insanity” yields approximately 4 million results versus only 742,000 for “terrorist rationality”. Nonetheless, studies have shown numerous examples of terrorists behaving in accordance with the rationality hypothesis:

• When security measures for one mode of attack increase, terrorists switch their tactics. After the introduction of metal detectors in airports, skyjacking dropped, but other kinds of attacks increased (Enders & Sandler, 1993).

• Terrorists adjust their tactics according to how well a target is protected. For example suicide attacks are usually only used for hardened targets (Berman & Laitin, 2006).

• Terrorists time their attacks to have a maximum political impact. Attacks just before an election are more likely (Aksoy, 2011).

Gerry Becker (1968) was the first to look at crime from an economic perspective. He argued that criminals are rational utility maximizers that engage in crime when they expect the net benefit of their action to be positive. Hence, they can be deterred when the severity of punishment or the probability of detection is sufficiently high.

A game theoretic model

Game-theory as a branch of applied mathematics is frequently employed in economics and political sciences. It can be used to analyze situations in which payoffs mutually depend on the other players’ action. Since the distribution of payoffs is highly interdependent in terrorism, game-theory seems to be well suited to analyze terrorists’ interaction with other players such as targeted countries, other terrorist groups, or the media.

The aim of terrorism is to spread fear beyond that of its immediate victims. So, clearly, the media is a critical component of terrorism. However, opinions on the nature of this relationship differ:

• Margaret Thatcher described the media as the “oxygen” on which terrorism depends (Apple, 1985).

• The historian Walter Laqueur (1967) said: ‘‘The media are a terrorist’s best friend. The terrorist’s act by itself is nothing. Publicity is all.’’

• Others have raised the concern that the media and terrorism will move in a vicious circle that moves to every higher levels of terrorism (Frey & Rohner, 2007).

In the game theoretic model presented here, (potential) terrorists can choose between non-violent political acts and terrorism (T). The media determine the probability (p) that a terrorist attack is covered in the news. Thus, there are two players, whose choice of strategy influences the other’s payoff.


Figure 1: The reaction functions of both the media (dashed, black) and the terrorist organization (red) are shown. The two lines intersect at the Nash-equilibrium.

By maximizing the utility functions of both players the reaction curves can be obtained (see figure 1). A reaction curve indicates the best response to a parameter which is determined by the other player. The media set their reporting propensity according to the overall amount of terrorism. Terrorists set their level of terrorism according to the expected reporting probability. Both curves intersect at the Nash-equilibrium. At this point, neither player has an incentive to unilaterally alter her choice of parameter.


The Nash-equilibrium is intermediate for most parameter choices. The terrorism-media interaction, thus, seems to have some self-regulating properties: High levels of terrorism lead to lower attention per incident and this lowers the incentives for terrorism. A high attention/incident ratio is only sustainable at low levels of terrorism.

It is useful to see the reaction curves in a local context. For example, a terrorist attack in the United States is a very rare event and will therefore lead to a high attention/incident ratio. On the other hand, a bombing in Bagdad is much higher on the Y-axis (see figure 1) and therefore will have more difficulty making it to the news.

A vicious circle of media attention and ever more terrorism thus seems unlikely, simply because a high attention/incident ratio becomes unsustainable at high levels of terrorism. So, the system is best characterized as self-stabilizing. On the other hand, this means that very low levels of terrorism are equally hard to achieve.

The model briefly outlined here can be extended to cover other settings such as multiple terrorist groups (endogenously and exogenously determined), adaptive expectations, and changes in external parameters.

For a fuller discussion see: https://www.tandfonline.com/doi/abs/10.1080/19434472.2011.594629


Aksoy, D. (2011). Elections and the timing of terrorist attacks in democracies. Mimeo. Retrieved from: htp://www.princeton.edu/~daksoy/Deniz_Aksoy_Homepage/RESEARCH_files/Aksoy%20elections%20and%20terror.pdf

Apple, R.W. (1985, July 16). Thatcher urges the press to help ‘starve’ terrorists. The New York Times, A3. Retrieved from https://www.nytimes.com

Berman, E., & Laitin, D.D. (2005). Hard targets: Theory and evidence on suicide attacks, NBER Working Paper, no. 11740. Retrieved from https://www.nber.org/papers/w11740

Caplan, B. (2006). Terrorism: The relevance of the rational choice model. Public Choice, 128(1), 91–107; doi: 10.1007/s11127-006-9046-8

Enders,W., & Sandler, T. (1993). The effectiveness of antiterrorism policies: Vector-autoregression-intervention analysis. American Political Science Review, 87(4), 829–844. Retrieved from https://www.jstor.org/stable/2938817

Frey, B.S., & Rohner, D. (2007). Blood and ink! The common-interest game between terrorists and the media. Public Choice, 133(1), 129–145; doi: 10.1007/s11127-007-9182-9

Laqueur, W. (1976). The futility of terrorism. Harper’s, 252 (1510), 99–105. Retrieved from https://www.harpers.org/archive/1976/03/0022427

Schelling, T.C. (1980). The strategy of conflict. Cambridge, MA: Harvard University Press, (Original work published 1960.)

This entry was posted in Analysis and tagged , , , . : . Bookmark the permalink. Both comments and trackbacks are currently closed.
  • Navigate to regional portals

    North America Middle East South Asia Europe
  • Latest analysis from our experts

    Paul Kamolnick
    Paul Kamolnick · United States
  • Get our Extremism Tracker Email

    * indicates required