Exploring the structure of terrorists' WMD decisions: A game theory approach

Defense and Security Analysis

Volumn 20, Issue 4, 2004, Pages 355-372

  Franck, Raymond E ^a   Melese, Francois ^a  

^a NAVAL POSTGRADUATE SCHOOL   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

GAME THEORY; TERRORISM;

EID: 23344438565     PISSN: 14751798     EISSN: None     Source Type: Journal    
DOI: 10.1080/1475179042000305813     Document Type: Article

References (40)

1. A. Lake, 6 Nightmares, Boston: Little and Brown, 2000.
2. It is fair to say there is a strong degree of consensus, but not necessarily unanimity. Lake, op. cit.;
3. W. Enders and T. Sandler, "Is Transnational Terrorism Becoming More Threatening?: A Time-Series Investigation", Journal of Conflict Resolution, Vol. 44 No. 3, 2000;
4. W. Enders and T. Sandler, "Patterns of Transnational Terrorism, 1970-1999: Alternative Time Series Estimates", International Studies Quarterly, Vol. 46, 2002.
5. The term "macro-technology of conflict" comes from Hirshleifer. J. Hirshleifer, The Dark Side of the Force: Economic Foundations of Conflict Theory, Cambridge, UK: Cambridge University Press. An example of a "disruptive" measure is a pre-emptive attack in the event terrorists acquire WMD. The threat of a pre-emptive attack as a "disruptive" measure is discussed by Melese and Angelis in this issue, in a paper entitled "Deterring Terrorist from Using WMD: A Brinkmanship Strategy for the United Nations".
6. T. Sandler and D. Arce M., "Terrorism and Game Theory", Simulation and Gaming, Vol. 34 No. 3, 2003.
7. This representation is similar to Davis and Jenkins'. P. Davis and B. Jenkins, "A System Approach to Deterring and Influencing Terrorists", Conflict Management and Peace Science, Vol. 21 No. 1, 2004.
8. S. Blomberg, et al., "An Economic Model of Terrorism", Conflict Management and Peace Science, Vol. 21 No. 1, 2004.
9. R. Hudson, et al., Who Becomes a Terrorist and Why, Washington DC: Federal Research Division, Library of Congress, 1999, undated reprint, Guilford, CT: Lyons. There are other standard definitions. One compendium is provided by Terrorism Research Center Inc., http://www.terrorism.com/.
10. Hudson, op. cit.
11. T. Friedman, "The Search for P.M.D.s", New York Times, 23 May 2004.
12. Quoting Osama bin Laden: ". . . we do not worry about what America says. We look at ourselves as worshippers of Allah who created us to worship him and follow his books and prophets. I am one of Allah's worshippers. I worship Allah, which includes carrying out the jihad to raise Allah's word and evict the Americans from all Muslim land". A. Robinson, Bin Laden: Behind the Mask of theTerrorist, New York: Arcade, 2002;
13. Friedman, op. cit.
14. R. Lifton, Destroying the World to Save It: Aum Shinrikyo, Apocalypic Violence and the New Global Terrorism, New York: Holt, 1999.
15. This is similar to Davis and Jenkins, op. cit. Our "political" terrorists resemble their "Type B", and "fanatic" is similar to "Type A". We think our terms are more descriptive.
16. Hudson, op. cit.
17. T. Sandler and K. Siqueira, "Global Terrorism: Deterrence Versus Preemption", unpublished manuscript, October 2003.
18. Sandler and Enders, op. cit., 2002.
19. Hudson, op. cit.;
20. J. Stern, Terror in the Name of God: Why Militants Kill, New York: Harper-Collins, 2003.
21. Stern, op. cit.;
22. Hudson, op. cit.
23. There is a distinction for which we do not account. An operation may have the intended physical effect, but not damage the target as intended. Attacks on a housing compound in Saudi Arabia in May 2003 allegedly achieved the desired level of physical damage. However, the area's foreign residents had recently relocated. Hence, this operation may not have had the desired effect on the target government(s). However, it would be considered "effective" in our scheme.
24. The mass transit bombings in Spain, March 2004, were "effective" in achieving much of the intended damage. According to many observers, they also significantly influenced the results of the Spanish elections. If so, the operation was also highly "impressive".
25. The Achille Lauro operation of 1985 failed to achieve the intended physical effects, and those directly involved appeared to be both brutal and incompetent. If those perceptions were widely shared, then that operation was both ineffective and highly unimpressive.
26. Sandler and Siqueira, op. cit., classify target country measures as "deterrence" or "pre-emption". Williams classifies these measures as "defensive", "offensive" and "preventive". We use terminology from Franck and Hildebrandt - which identifies two forms of "offsetting" measures to opposing capabilities: "defense" and "disruption". Defensive measures are intended to negate the effects of opposing operations, while disruptive measures are intended to dislocate those operations.
27. C. Williams, "Paying for the War on Terrorism: US Security Choices since 9/11", unpublished manuscript, January 2004;
28. R. Franck and G. Hildebrandt, "Competitive Aspects of the Contemporary Military-Technical Revolution", Defense Analysis, Vol. 12 No. 2, 1996.
29. Williams, op. cit.;
30. Sandler and Sequeira, op. cit.
31. Hirshleifer, op. cit.
32. Sandler and Arce, op. cit., point out that terrorists respond to defensive measures against one class of operational objectives by increasing operations against other classes, i.e., they seek out "soft" (less defended) targets. We assume this effect is translated into a lower probability of effective operations as a result of defensive measures.
33. It seems highly unlikely that a terrorist organization with the capability would plan a campaign solely with WMD. Therefore we represent the terrorists' choice at Nodes K and L (in Figure 4) as plans which involve conventional weapons only, and plans which feature some WMD. The terrorist organization can, of course, choose to not conduct any operations. It is more likely that terrorists will vary the number of operations over a planning period rather than completely stop. Stern, 2003.
34. The extensive form of a game presents players' decisions in the form of a decision tree. A number of game theory texts, e.g., Rasmussen, provide precise definitions of the extensive form. E. Rasmussen, Games and Information: an Introduction to Game Theory, New York: Blackwell, 1989.
35. We realize that symmetry of changes in probability is arbitrary. However, it simplifies the discussion without changing the general nature of its conclusions.
36. In making these comparisons, we assume that conventional operations against defensive countermeasures have a positive expected payoff; i.e., PE[S+I(PI-PA)] - (1-PE)C > 0. While this is a reasonable assumption, it precludes our exploring the possibility that risk-taking terrorists might resort to WMD if conventional operations are a losing proposition.
37. It is natural from a zero-sum game theory perspective to suppose that the target government would choose ΔP (at Node H) so as to make the terrorists indifferent between conventional (Node A) and WMD (Node B) operations. This implies ΔP = {(w-1)[PE(S+I (PI-PA)(1-PE)C] }/{(w-1) [S+I(PI-PA)+(w+1)C}. Since it must also be true that 0 ≤ ΔP ≤ PE, it is not clear that we can satisfy both conditions simultaneously. It is also not clear that there is a set of countermeasures available that would achieve the desired value of ΔP, even if both conditions could be satisfied. In this discussion, we assume that the terrorists' payoffs at Nodes A and B are not necessarily equal.
38. In this model, we assume the results of WMD operations are a large multiple (w) of conventional operations. Increasing (PI-PA), all other things equal, increases the expected payoff to the terrorists in both types of operations, making WMD operations accordingly more attractive, relative to conventional.
39. This seeming paradox is not far-fetched. In American football (as one example), offensive schemes include the capability to use the forward pass. Some teams are pass-oriented, and intend to use passing plays as the centerpiece of their offense. However, some teams are strongly run-oriented, and use the threat of a forward pass to force opponents into defensive schemes that are more conducive to successful running plays.
40. This is true except for the special case of a terrorist organization that is purely fanatic.