Does Terrorism Work?: A Literature Review

What follows is part of an honors thesis submitted by Ryan McCarthy to the University of North Carolina at Chapel Hill. Those interested in this article might also be interested in the second portion of his thesis on the history of the IRA

Political violence is an incredibly rare occurrence — and understandably so. The vast majority of individuals, if the frequency of politically motivated violent acts can be taken as any sort of reliable metric, do not believe that violence in pursuit of abstract political goals is justifiable or even pragmatic. Why, then, are certain individuals so willing to resort to killing, bombing, and burning as a means of achieving their political objectives? A preliminary look at this question might lead one to believe that the cause is weak political systems that do not allow individuals to air their grievances in an otherwise legitimate manner. However, instances of political violence are found within states that have what are considered to be strong political institutions. In fact, the so-called “paradox of terrorism” begs the question of why “terrorist groups tend to target societies with the greatest number of political alternatives, not the fewest.”[1]

The questions of how terrorists operate, why they decide to resort to violent tactics, and what they want are often addressed. Equally important is the question of whether or not terrorism works. Answers to this question in the literature are varied. For instance, Andrew H. Kydd and Barbara F. Walter begin their article “The Strategies of Terrorism” with the declarative statement, “Terrorism often works.”[2] They are not setting out to prove that political violence works; they are operating on the assumption. On the other hand, Max Abrahms wrote an article entitled “Why Terrorism Does Not Work” arguing that the prevailing view in the political science community that politically motivated violence is an effective strategy is not only unfounded, but also harmful to the policy-making community. Seeing as no consensus has been reached on the question of whether or not terrorism is a useful method of political change, it must be concluded that the question has not been decisively settled. In fact, a more prudent question might be “Does terrorism work for the group in question?” Terrorism is a notoriously fickle beast, in that there is no consensus on how the academic community should define the term. The problem is not only that terrorism is a politically and emotionally charged term. Even in a more dispassionate sense, terrorism is difficult to define because it is carried out by a variety of actors, in a variety of ways, against a variety of targets. The nebulous quality of this collection of actions makes it difficult to study, and even more difficult to draw any significant conclusions about. However, detailed analyses of terrorist groups themselves, including their political goals, methods, and levels of activity, can help scholars highlight commonalities across terrorist groups in addition to providing important information about groups both active and inactive. 

Scholars disagree on the clearest political indicators or interactions that seem to predict the occurrence of incidents of terrorism. The papers in this review share a focus on terrorism as a tool for bringing about political change and an interest in determining what drives groups to use political violence as a vehicle for pursuing their goals. 

A particularly prominent paper on the topic of political violence is “The Strategies of Terrorism” by Andrew H. Kydd and Barbara F. Walter. Their paper revolves around the assumption that terrorism works. Building on this assumption, they attempt to explain what terrorists want, what methods they use to pursue these goals, why these methods work, and what the best responses of governments are given the methods that terrorist groups employ.[3] Their central argument for explaining the existence of terrorist groups “is that terrorist violence is a form of costly signaling.”[4] They contend that terrorist organizations are often too weak to engage in conventional warfare with their adversaries. As a result, they must resort to dramatic acts of public violence that demonstrate their ability to raise the costs of defying their wishes and indicate their commitment to continuing violent conflict if their demands are not met. 

Kydd and Walter delineate five distinct strategic logics that may explain the behavior of terrorist groups. The various explanations are: attrition, intimidation, provocation, spoiling, and outbidding. In short, attrition is the strategy that is adopted when a group wants to inflict heavy costs on a target group. Intimidation is when a group uses violence in a bid to control a society through fear and punishment. Provocation is an attempt to induce an opponent to retaliate with indiscriminate force, which should engender support for the group. Spoiling is an effort to upset bargaining processes that the group finds unfavorable so that they may continue to better their bargaining position or extract concessions. Finally, outbidding is a strategy undertaken by terrorists when they wish to show that they are the most committed and extreme faction representing a specific issue or population, which helps them garner support amongst those populations who desire hardline groups to represent them in bargaining situations. 

Kydd and Walter conclude that closer analyses of each of the strategies that they outline will yield a more helpful conclusion as to the best response to each. They go on to say that the two most important variables in future studies of terrorist strategies are information and regime type. As far as information is concerned, they contend, “terrorism is not a problem of applying force per se, but one of acquiring intelligence and affecting beliefs.”[5] This assertion underlies their argument that terrorist groups resort to terror tactics in an attempt to communicate their position and readiness to fight. The second variable, regime type, is important in their analysis because they note that democracies are more often the target of attrition campaigns. They hypothesize that democracies have certain features that may encourage terrorist attacks, whereas authoritarian regimes dissuade terrorist attacks due to their ability to respond with unconstrained force. This paper is important because it offers a comprehensive framework of the motivations behind terrorist behavior, but it does so in a generalized way such that these categories could apply to any instance of political violence. 

Another important paper in the analysis of terrorist behavior is “What Terrorists Really Want” by Max Abrahms. This paper presents an analysis of terrorism within a rationalist strategic framework. Abrahms explains that the strategic model views terrorists as “political utility maximizers; people use terrorism when the expected political gains minus the expected costs outweigh the net expected benefits of alternative forms of protest.”[6] If this is truly the case, Abrahms argues that policymakers can use this information to their advantage in crafting counterterrorism policy; rational agents are, after all, easier to predict. He goes on to detail the three central assumptions of the strategic model that will determine whether or not terrorist activity can be predicted:

(1) terrorists are motivated by relatively stable and consistent political preferences; (2) terrorists evaluate the expected political payoffs of their available options, or at least the most obvious ones; and (3) terrorism is adopted when the expected political return is superior to those of alternative options.[7]

These assumptions ensure that terrorists are dealt with as rational agents who know what they want and wish to pursue their goals in the most effective way available to them, even if their goals are antisocial or disruptive. 

However, Abrahms does not buy into the strategic model’s predictions. He believes that terrorists often behave in ways that contradict the predictions of the model. For instance, he observes, “terrorist organizations do not achieve their stated political goals by attacking civilians.” [8] His assertion suggests that terrorism is, for clear reasons, a highly unpopular action that is sure to turn both public opinion and the opinion of policymakers against any cause associated with terrorism. Abrahms notes that it seems unlikely that politically motivated actors would knowingly engage in behavior that hinders the pursuit of their objectives. Another of the alleged contradictions Abrahms highlights is the fact that “terrorist organizations reflexively reject compromise proposals offering significant policy concessions by the target government.”[9] In short, Abrahms believes that there are enough instances of terrorist groups behaving irrationally or not in accordance with their stated goals that applying the strategic model to terrorism is a foolish endeavor. Instead, he concludes that terrorists are socially alienated individuals who seek to “develop strong affective ties with fellow terrorists.”[10]

However, Robert A. Pape wrote a paper entitled “The Strategic Logic of Suicide Terrorism” in which he discusses just that: whether or not the use of suicide attacks as a terrorist tactic is logical and strategically beneficial. He reaches the opposite conclusion of Abrahms. Through an analysis of numerous incidents of suicide bombing, Pape concludes that suicide tactics are indeed strategically logical because they are designed for use in the pursuit of a political end, rather than being driven by pure hatred and fanaticism.[11] His analysis yields him five conclusions: first, the reaffirmation that suicide terrorist attacks are strategic; second, the observation that “suicide terrorism is specifically devised to coerce modern democracies to make significant concessions to national self-determination”; third, suicide terrorism is on the rise because terrorists have seen that it is indeed effective; fourth, Pape observes that while moderate terrorist campaigns are effective at eliciting moderate concessions, more ambitious and pervasive campaigns are less likely to gain greater concessions; and finally, Pape believes that the most effective way to reduce the likeliness of suicide attacks is to make terrorists believe that they cannot carry out these attacks, through investment in border and homeland security.[12]

Pape’s paper is important because it reaffirms the notion that terrorism is a strategic behavior. Indeed, Pape’s argument seems to support the notion that terrorism is by definition rational, since most scholars hold that terrorism is politically motivated violence. Pape’s paper is also important because it analyzes a specific type of terrorism. He uses a database of suicide terrorist attacks to support his conclusions and outlines a fairly extensive theoretical logic for the use of suicide attacks. This sort of in-depth knowledge about types of terrorism is useful to policymakers and academics so that they can inform themselves about why groups prefer certain types of tactics to others, and more effectively understand the logic of these differing tactics, thus allowing them to craft better policy in response. Indeed, his conclusion suggests that governments would be greatly benefited by trying to remove the incentives to terrorism that certain groups may encounter — namely the ability of groups to carry out attacks that have repercussions far exceeding the amount of planning and materiel that went into their execution. By investing in border and homeland security, governments signal to terrorists that future attacks will be even more costly to carry out due to the additional security measures that these groups must now surmount. However, his paper obviously does not address the best way for governments to combat terrorist action, only the way that he believes is best for preempting. 

While an analysis of why certain groups resort to political violence and what tactics they prefer is important, it is equally important to examine the reasons that nonviolence may also be a viable strategy. Maria J. Stephan and Erica Chenoweth undertook an analysis from this perspective in their paper “Why Civil Resistance Works.” As the title makes clear, Stephan and Chenoweth do not buy into the argument that “opposition movements select violent methods because such means are more effective than nonviolent strategies at achieving policy goals.”[13] Instead they argue that many groups have adopted nonviolent strategies such as “boycotts, strikes, protests, and organized noncooperation”[14] to great success. If this is true, then it is entirely understandable why their analysis is necessary amidst the literature extolling the effectiveness of violent methods of resistance, not only because it contradicts the traditional knowledge, but also because it indicates that terrorist groups and individuals considering joining these groups need not view violence as the only logical choice available to them.

Stephan and Chenoweth suggest that nonviolent resistance is not only an effective alternative to violent resistance, but that nonviolent methods are actually more effective than their violent alternatives, with nonviolent campaigns ending successfully 53% of the time and violent campaigns ending successfully 26% of the time.[15] They attribute this success to two traits that nonviolent campaigns possess but violent ones lack. First, they note that nonviolent methods of resistance are able to more easily garner international and domestic support. Since violence tends to fall outside the spectrum of acceptable behavior for most people, nonviolent campaigns are able to attract support from individuals who may agree with the objectives of a campaign but may not necessarily agree with the use of violent tactics. Second, they point out that violent resistance on the government’s part is justified in response to domestic violence in terms of public opinion, whereas violence in response to a nonviolent campaign is widely unfavorable and may actually hinder the government’s campaign. These two observations form the foundation that undergirds the authors’ argument that violence need not be the norm in opposing unwanted social structures.

After running a sample of 323 violent and nonviolent campaigns through a statistical model, Stephan and Chenoweth conclude that resistance campaigns that manage to induce loyalty shifts among civilians and security forces are more likely to succeed. These loyalty shifts seem to be an important part of a successful campaign, and they note that none of their violent case studies managed to bring about a loyalty shift, whereas two nonviolent campaigns (East Timor and the Philippines) managed to sway the loyalty of the populace and security forces. Additionally, they conclude, “violent and nonviolent campaigns that fail to achieve widespread, cross-cutting, and decentralized mobilization are unlikely to compel defection or evoke international sanctions in the first place.”[16] Essentially, the ability of a campaign to accrue domestic and international support is crucial to its success, and nonviolent campaigns seem to be more adept at garnering support; therefore, it seems that nonviolent campaigns will succeed more often than violent ones. Their theoretical predictions and dataset seem to support their hypothesis that nonviolent resistance is a viable alternative to violence in civil campaigns. This paper is relevant for our purposes because we are looking for an answer to the question “what motivates political violence?” If evidence suggests that political violence is inefficient in comparison to nonviolent alternatives, then analyzing what causes it takes on a new context, one in which the actions of terrorists must be deemed ineffective and thus irrational. 

A paper by Todd Sandler and Walter Enders, entitled “Transnational Terrorism: An Economic Analysis,” looks at terrorism as a phenomenon that can be explained using economic theory and game theory.[17] This paper has subsections that deal with game theory and hostage taking, game theory and governmental response, terrorism in the context of rational choice theory, and a benefit-cost analysis of terrorist-thwarting policies, among other sections. In each section, Sandler and Enders briefly recap the previous literature on each of these topics. What this paper lacks is a section on the game theoretic or economic logic of resorting to political violence in the initial stages of a conflict. It is only natural to concern ourselves with the potential targets of terrorism, the goals that terrorists seek, and other post hoc considerations after the initiation of a terrorist campaign. However, game theory can also yield helpful insights about the decision to instigate a terrorist campaign against a target government. 

Many of these models take the form of a decision tree.[18] The merit of this particular approach is that it considers the uncertainty inherent in interactions between states and terrorists, and it considers the alternatives available to each side in a conflict. At the end of “Terrorism and Game Theory,” Sandler and Arce design a decision tree for two different games between terrorists and states: first, the choice terrorists face regarding whether they should attack tourists or businesses after a state has chosen to try to deter the terrorists; and second, the state’s choice to offer concessions to either moderate or hardline factions of a terrorist group, and the resultant choice that these factions are faced with regarding whether or not they should accept or reject these concessions.[19] These models implicitly assume that the actors are indeed faced with choices, and that they deliberate based on probability and reasoning. Accepting these premises leads to the conclusion that decision trees are effective models for simplifying the decision making process and modeling games between terrorists and states.

Harvey E. Lapan and Todd Sandler also use game theory to model the decision of states to pre-commit to a non-negotiation position, in which they publicly declare their unwillingness to grant terrorists any concessions. At the same time, terrorists are faced with the decision of whether or not to initiate a campaign of violence or kidnapping. It is apparent that these two decisions interact with the likelihood of one another. For example, a state would be more likely to improve its internal security if the likelihood of a terrorist attack was higher. On the other hand, a terrorist group is less likely to attack if they see that their target state has been beefing up security, which may lead them to then switch targets. In the current stage, which in this case operates on the assumption that the terrorists have succeeded in kidnapping one or more individuals, the government must decide whether or not to negotiate with the terrorists. Obviously this decision will be affected by whether or not the state has publicly committed itself to a non-negotiation position ahead of time. Lapan and Sandler’s model suggests that terrorists will conclude that governments whose declarations are shrouded in uncertainty are more willing to negotiate than they initially seem, as never negotiating becomes a problematic strategy for governments to adopt in the face of a successful terrorist campaign. In this sense, the decision to commit to a strategy of never negotiating is more an attempt to improve the relative bargaining position of the state vis-à-vis the terrorists rather than a genuine political position.[20] This paper is noteworthy because it parallels the decision terrorists face between escalation and de-escalation in response to some government behavior.

Despite the lack of a definitive answer to the question of what causes political violence to be an attractive option for groups, the literature on the interplay of campaigns between governments and insurgent groups is rich. For instance, “The Repression of Dissent: A Substitution Model of Government Coercion” by Will H. Moore uses a statistical model to analyze the response of states to the behaviors of dissident groups. Indeed, his approach holds similar assumptions to game theory, as he expects that the behavior of the state can be predicted based on the most recent behavior undertaken by the dissident group. That is, the state chooses between repression and accommodation when deciding how to respond to dissident groups by first looking at whether or not these groups have cooperated or protested most recently. Since the players have two choices available to them, this scenario could easily be modeled using a 2×2 game

Moore implies that dissidents and governments are both actors who are responsive to costs imposed on them from the outside. Additionally, he says that his model views the state as an actor who takes a retrospective rather than a prospective approach; the state takes its behavior cues from the past behavior of dissident groups rather than looking to the future to predict what they may do.[21] In a game theoretic model, we would call this behavior a “tit-for-tat” strategy; this is a strategy predicated on the notion that rounds of a game will reoccur so that both actors will have the knowledge that they must interact again in the future hanging over them. This strategy proposes that an actor should take note of the most recent behavior of an opponent actor and essentially “repay the favor” by responding in kind. So if during the last round of negotiations a dissident group defected from the bargaining process, tit-for-tat would suggest that the state should defect this time. The logic behind this strategy holds that actors should use their knowledge of the shadow of the future (the knowledge that they must continue to interact indefinitely) in order to commit themselves to credible bargaining positions. 

After running the data through his model, Moore concludes that the “states tended to substitute accommodation for repression and repression for accommodation whenever either tactic was met with dissent.”[22] This outcome indicates that his hypothesis was correct: states will switch strategies whenever the ones they are currently employing are met with a strategy of dissent by outside groups. He concludes that states and dissenting groups are both responsive to costs being imposed upon them and are willing to inflict costs in order to try to bring policy in line with their vision of how it should be. Furthermore, he argues that his findings support the conception of states and dissenting groups as rational agents who know what they want and undertake actions that they believe will further the pursuit of their goals. 

The articles I’ve discussed above indicate a level of disagreement about the effects of terrorist activity. What is unquestionable is the conclusion that terrorists are political, and therefore rational, entities. Identifying them as rational is not to lend any moral justification to their choice of tactics, but to argue that they consciously formulate goals and choose those actions which they feel are most likely to bring them closer to achieving those goals. Terrorism may appear to be counterproductive or irrational, but it is important to bear in mind that these groups are choosing their actions carefully in an effort to elicit the desired response from the target of their campaign. Of course, much more goes into the decision to defer to the demands of a terrorist group than the simplified logic here would indicate. This includes the willingness of the target government to offer concessions or concede outright, the willingness of a state’s population to continue to take casualties in their fight against the terrorist group, and the general political climate and culture of that state. So, does terrorism work? In a word, yes, but only sometimes. 


[1] Max Abrahms, “What Terrorists Really Want: Terrorist Motives and Counterterrorism Strategy,” International Security 32, no. 4 (April 1, 2008): 84

[2] Andrew H. Kydd & Barbara F. Walter, “The Strategies of Terrorism,” International Security 31, no. 1 (July 1, 2006): 49. 

[3] Kydd & Walter, “The Strategies of Terrorism,” 50. 

[4] Ibid.

[5] Ibid., 79.

[6] Abrahms, “What Terrorists Really Want,” 78.

[7] Ibid., 79.

[8] Ibid., 82.

[9] Ibid.

[10] Ibid., 96.

[11] Robert A. Pape, “The Strategic Logic Of Suicide Terrorism,” American Political Science Review 97, no. 3 (August 2003), 2.

[12] Ibid.

[13] Maria J. Stephan & Erica Chenoweth, “Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict,” International Security 33, no. 1 (June 26, 2008), 7.

[14] Ibid., 8.

[15] Ibid.

[16] Ibid., 42.

[17] Todd Sandler & Walter Enders, “Transnational Terrorism: An Economic Analysis,” European Journal of Political Economy 20, no. 2 (September 2004): 301-316.

[18] See Sandler & Enders (2004) and Sandler & Arce (2003).

[19] Sandler & Arce, “Terrorism and Game Theory,” 326 & 332.

[20] Harvey E. Lapan & Todd Sandler, “To Bargain or Not To Bargain: That is the Question,” The American Economic Review 78, no. 2 (1988): 16-17.

[21] W. H. Moore, “The Repression of Dissent: A Substitution Model of Government Coercion,” Journal of Conflict Resolution 44, no. 1 (2000),108.

[22] Ibid., 120.

1 Comment

Leave a Reply