Introduction

This analysis is based on a game theory model which assumes the premise that the behaviour of Jihadist groups (taken generally, and al-Qaeda in particular) and of States is rational and responds to strategic choices.[1] The analysis begins with an introduction to the context which situates Spain and global Jihadist groups on opposing sides of the game board of strategic interests. There follows an analysis of each of the scenarios and their implications in the strategies of Jihadist organisations, as well as their influence on Spanish foreign policy and international security policy decisions. Finally, conclusions are presented and the scenarios are classified in accordance with their significance and their capacity to destabilise Spain’s position in its various theatres of operations abroad, most notably those linked to peace missions and, in particular, Afghanistan.

Assessment of the Jihadist Risk to Spain

The assessments of the risks posed by terrorism hinge on the threats to objectives which present a particular set of weaknesses. As an expression of violence, terrorism taps any weaknesses it detects among its objectives to increase the impact, and therefore the damage, from its attacks, and to thereby secure tactical or strategic advantages. It is therefore clear that, the more vulnerable States are, the more havoc can be wreaked upon them by groups intending to cause harm. At the same time, to suitably conceptualise the threats, it is necessary to gauge intentions and capacities. In other words, groups with hostile intentions which also have operating capacities to back these intentions, will be more likely to pose a risk.

The perception that Spain is not exempt from the risk of a future Jihadist attack on its territory has steadily gained weight in recent times. This perception of risk is fuelled by both objective and subjective elements of insecurity. Among the objective elements, the attacks on Spanish interests in Casablanca in May 2003 but, above all, the bombings in March 2004 in Madrid, both of which were attributed to Jihadist groups linked one way or another to al-Qaeda, constitute the most obvious empirical references that gauge the capacities and intentions of Jihadist violence in the risk equations concerning terrorist violence against Spain. All of these elements constitute representations of capacities of Jihadism convergent with their intentions, declared or otherwise, to attack and harm Spanish interests, resources, presences or citizens. This also applies to the attacks (here the concepts of insurgency and terrorism blur, and these theoretical definitions are outside the scope of this work) suffered by Spanish troops or agents in theatres of war, most notably Afghanistan and Iraq. Similarly indicative of the intentions and capacity of the Jihadist threat to Spain, although not yet an operating focus of eventual terrorist actions, are the intelligence-based police and legal operations to dismantle cells providing (at least) logistical and recruitment support to Jihadist terrorism, which have been conducted in Spain in the last few years.

For its part, among the most significant subjective elements when it comes to shaping the perception of risk to Spain are the expressions of the intentions by various individuals linked in one way or another to Jihadism and to al-Qaeda as a reference, in particular by the Egyptian Ayman al-Zawahiri, broadcast on the Internet and through social communications media. In this connection, Bermejo and Reinares analysed the evolution of Spaniards’ subjective security with respect to Jihadist activity.[2] What we are interested in highlighting from the subjective standpoint of this specific component in risk assessment is that, although the hostile messages may be partly linked to the intention of threat, it seems more reasonable to think that the threat itself is using them as a way of creating a certain perception, both among the State’s institutions and its citizens. Accordingly, communication designed by al-Qaeda transcends its traditional place in the risk equation (intentions) to become a resource in itself (capacity) in committing terrorism. We place this under the subjective heading because it is precisely its capacity to spread the feeling of fear, of danger, to the population and the authorities (and therefore dependence on our perception) that should define this component in our analysis.

Based on these premises, in terms of both intentions and capacities of a Salafist Jihadist threat, and according to the weaknesses deriving from Spain’s position in the sphere of international relations with respect to other countries which, in Salafist ideology, fall within the so-called ‘lands of Islam’,[3] Jihadism is considered to constitute a risk that is international in scope and nature. Along these lines, any analysis aimed at unravelling and examining in scientific terms the strategic behaviour of the Jihadists may shed some light (although it will almost always be partial until we make headway in understanding the phenomenon in a comprehensive and interdisciplinary fashion) on the question which repeatedly emerges every time the threat is evidenced in the form of an attack: why Spain?

Of the various methodological approaches to the strategic analysis of the Jihadist threat, we have opted to classify the available observations in an analytical game theory model. Game theory is an area of applied mathematics aimed at studying decision-making processes in which two or more rational players compete to achieve their interests in a common decision-making sphere. Originally formulated by John von Neumann and Oskar Morgenstern[4] in the second half of the 20^th century, it was initially applied precisely to strategic models, at that time in the military sphere. It has subsequently been applied to a variety of social scenarios and, more recently, to terrorism.[5]

The application to terrorism of models based on the game theory for analysis of strategic decisions by Jihadist groups implies accepting that this paradigm assumes:

(1) That there is a conflict between the players involved and that they are pursuing opposing interests.

(2) That there is an interaction based on the concept of the other as an adversary.

(3) That the players are rational when choosing their decision strategies.

(4) That the result of analysing the decision-making sphere is limited to the variables included in the formal model to which the statistical formulation is applied, leaving out, therefore, other variables which might influence –and indeed which probably do influence– the decisions which the players make.

When we propose an analysis of Jihadist terrorism based on game theory we are facing a number of limitations since, in the first place, we are looking at two groups of players, States and Jihadist groups, and assuming that the latter have a more or less convergent decision-making core, while States (such as Spain) make decisions focusing more or less solely on the observation of the threat (when we know that other factors enter into play, such as, for a start, the influence on Spain from other players, not always States, in the international community).

Another major restriction in game theory models, closely linked to assuming that players are rational, is that the actions of those making decisions within the framework of the model will adjust to strategies defined explicitly or implicitly to achieve certain tacit or declared objectives. This condition renders the model fragile at a practical level, especially in terms of assessing the risk in counterterrorism policy where very often opportunist attacks shape post-hoc argumentations, not necessarily coherent with previous discourses. As we have already highlighted here and pointed out elsewhere,[6] for now it is not possible to calibrate (and this impediment affects the theoretical core of a game model head-on) whether the discourse or planned ideologies which apparently underpin the actions of al-Qaeda and Jihadism in general are aimed at acting as a central pillar for a strategy applied in coherence with a discourse, or whether they are a tactical communications resource which, in their own right, seek a specific response from the victim or to create a certain perception in order to, based thereon, continue to pursue violence.

Lastly, in regard to the weaknesses of the analysis which we present, is the ethical dimension of the assessment. The game theory model assumes certain causal relations where the actions of some players are determined by those of others in an interactive system, of competition or cooperation, depending on the individual cases. It is worth pointing out here that, like all models, the game theory in this case dispenses with ethical viewpoints when considering the rationality of those making decisions, thereby overlooking the fact that terrorists are instigators of violence aimed at changing situations and that, for this purpose, they build tailor-made arguments to fit their own self-serving ‘reasoning’. Similarly, the ethical disaffection of the formal models which place all players on an equal footing in terms of rationality do not place into perspective the fact that terrorism is, essentially, a totalitarian instrument based on the attempt to impose a specific vision of society by force.

However, even bearing in mind these constraints, an attempt to quantify terrorist risk does bring added value to the state of the art, if only to visualise the limitations and make a firm choice, firstly for models of analysis of terrorist behaviour, and secondly for making these models integrated.

Our analysis poses two hypothetical situations to model possible attacks on Spain’s interests by Jihadist groups. Subject to the general limitations of formal models, an interpretation is given of the contextual situations in which Spain participates internationally based solely on two causal links (hypotheses which explain a cause and effect process) which are mutually exclusive, but which are designed so that as to cover most of the logic in respect of a hypothetical situation in which an attack is perpetrated against Spain, as well as the possible response by the State. This model is performed, as we have already indicated, without prejudice to the fact that there are alternative hypotheses and/or additional variables on which said hypothetical situations may be based.

Causal links of Jihadist actions

Based on the usual discourse of global Jihadism, with al-Qaeda as its maximum exponent, the strategic behaviour of this terrorist group in relation to the areas in the world where there is already a multinational presence of troops suggests the perception, at least for the Jihadist ideology, of a causal link whereby their attacks are aimed at destabilising the status quo to force governments involved in international deployments to rethink their positions. Still according to al-Qaeda’s rationale, these withdrawals would give way in these countries to Salafist organisations which would take up positions in order to build an Islamic State, as part of an Ummah presently deviated from the First Community of the Believers. In accordance with this neo-Salafist ideology, the status quo of these countries linked to the outside world and led by ‘impious’ governments is the current reason for the deviation of the Ummah and the impediment preventing Muslims from returning to the sources of Islam. Accordingly, the actions of Salafist Jihadist organisations are aimed at the main standard-bearer of the status quo, in other words, the government institutions which constitute the State, and/or any other presence and/or interest of States in the international political scenario.

Based on this strategic rationale, and considering that from the position of States there are three major categories of response to a Jihadist action: (1) withdraw the positions of government institutions of the State and/or in the lands of Islam (w); (2) maintain the status quo exactly as it was previous to the attacks which sought to destabilise (sq); and (3) deploy additional resources to offset the violent actions of the Jihadists (d); it is possible to assert that Jihadist organisations aim, first and foremost, is to destabilise the status quo in such a way as to force a withdrawal of the positions of government institutions and/or so that all resources that do not belong to Islamic lands be withdrawn from them. Their second choice would be a status quo. And, lastly, in terms of rationality adjusted to the discourse and announced intentions, Jihadism rejects any international policy decision which would imply an additional deployment of resources.[7]

From the standpoint of Muslim States where terrorist attacks are perpetrated in order to destabilise, and particularly where al-Qaeda and its supporters aspire to control the State, a situation of competition arises between Jihadist organisations and government institutions vying for control of the State, since those States that compete with Jihadist organisations for control of the institutions would rationally and as a first choice opt to deploy troops to offset the Jihadist actions and prevent them from acquiring any kind of power. After ‘rationality’, they would choose to sustain the status quo as presented prior to the destabilising event or attack. Lastly, they would choose the option of withdrawing their own positions, which would signal a setback for government institutions and control of the State for Jihadist organisations.[8]

As regards the institutions of secular States present in the lands of Islam which would be targeted by violent actions, but which are not in a situation of non-cooperative competition with Jihadist organisations for control of the State in the lands of Islam (such as Spain), they are expected to act in accordance with two possible rival interpretations of causality which explain Jihadist action (see Chart 1). According to the game theory model applied, if secular States assess that a possible Jihadist action takes place without the intervention of any contextual situation which might be justified by the ideology of, for example, al-Qaeda, these foreign States would behave strategically in line with the same order of priorities as the States where the majority of the population is Muslim (Hypothesis A); in other words, they would opt to deploy resources, rather than sustaining the status quo or withdrawing from their positions (d > sq > w). However, if they understand that there is a contextual situation, inscribed in the Jihadist ideology, providing a causal link (Hypothesis B), then they would opt to withdraw those resources which intervene in the lands of Islam to break the causal chain. Secondly in order of ‘rationality’, they would choose to maintain the status quo prior to the Jihadist action. And only thirdly would they make the decision to deploy resources to offset the destabilisation, since they understand that it is the deployment of their positions which gave rise to the causal link in the first place.[9]

Chart 1. Causal chain of Jihadist action (rival hypotheses)

Accordingly, we can assert that, based on the neo-Salafist ideology and applying the game theory as a rational model, the basic premises (Chart 1) are that al-Qaeda has defined a situation of non-cooperative competition between Jihadists and Muslim States in the lands of Islam vying for control of the territory and the public powers. Secondly, there is a situation of non-cooperative competition between Jihadists and foreign or secular States which have some kind of presence in the lands of Islam and which share Hypothesis A regarding the causality of Jihadist action. And finally, there is a situation of possible cooperative competition between the Jihadist action and foreign States present in the lands of Islam which share Hypothesis B concerning causality.

Context

Based on the Salafist Jihadist discourse justifying violence, Spain’s position in regard to the commitments undertaken as a member of NATO and the United Nations, two of the main standard-bearers of international security, place it in two scenarios of considerable importance in terms of the decision-making spheres regarding the terrorist threat: Afghanistan and Lebanon. Both of these scenarios, especially Afghanistan, have sufficient elements to become the ‘justification’ for possible future attacks against Spanish interests, all from a decision-making model that is based on the assumption that Jihadism, essentially materialised and configured via al-Qaeda, considers any foreign presence in ‘Islamic territory’ to be against its interests.

Furthermore, Spain has an influence on the political and institutional stability of North Africa, and its geopolitical position makes it a natural strategic ally of African countries in a number of spheres of intergovernmental cooperation concerning matters of home affairs, the fight against terrorism and border control, among others. From this standpoint, Spain may also be considered to be in a position of competition in North Africa on the geo-strategic game board which al-Qaeda has defined for those territories as a strategic player in the model.

Jihadist deployment and its inter-relation with States (analysis of scenarios)

The map of Jihadist activity is charted according to continued terrorism activities[10] based on neo-Salafist Jihadist ideology (y), the deployment of organisations in Islamic lands (o) and the commonly used strategy to destabilise the status quo, namely whether their actions are aimed at undermining the international or the local status quo (isq vs. lsq). The first two variables determine the intensity of the Jihadist activity in a particular State, while the second indicator signals whether the events are international or local in dimension.

The analysis of each of the scenarios that comprise Spain’s front against the Jihadists, as set forth in this section, responds to the rationale constructed from the two hypotheses (causal links) presented above. One of the hypotheses considers that there is in fact no causal link between the presence of Spain in international Islamic scenarios and possible terrorist actions by Jihadist groups (mainly linked to al-Qaeda) against Spain (Hypothesis A). The other hypothesis, for the purposes of contrasting the model’s goodness, understands that Spain participates contextually in the causal link and that said situation therefore would serve the Jihadist ideology as ‘justification’ for a terrorist action against Spain (Hypothesis B).

If there is a terrorist attack within the framework of the first hypothesis, Spain would be expected to react (using its own capacities and those integrated in standard-bearing international security bodies such as the ones we have already mentioned) to pursue those responsible both ideologically and at operating level. However, if there is a terrorist attack within the framework of the second hypothesis, Spain would be expected to react mainly by reconsidering its contextual position in the lands of Islam, without prejudice to the same mechanisms being triggered as for the first hypothesis. From the logical and strategic standpoint of the Jihadists, it would be more advisable to perpetrate an attack fitting into the second hypothesis, since there is a remote possibility that Spain might be persuaded to withdraw from its positions in what al-Qaeda considers ‘the lands of Islam’.

Iraq
In a rational decision model based on competing interests, the link between new attacks and the situation relating Spain to Iraq, particularly when the withdrawal of Spanish troops was instrumented by al-Qaeda as being a consequence of its violent acts, loosely fits a system for justifying further Salafist terrorism against Spain under the argumentational umbrella of the Iraqi case file. In the hypothetical model proposed, the only link which might be established with Spain are the investigations led by the security forces and the judiciary to dismantle the Jihadist networks of recruitment and ideological training in this country.

Furthermore, international-linked Jihadist activity, in other words, seeking the constitution of an Islamic Ummah and/or aiming their attacks against the international status quo (except for the local insurgency), has been considerably reduced in Iraq. Accordingly, the game theory model forecasts that Iraq does not currently constitute a sustainable front of attack on Spain.

Afghanistan
In Afghanistan, however, Spain does have troops deployed on the ground. And, furthermore, based on the indicator of the intensity and dimension of Jihadist actions[11] there was a considerable increase in 2006 which, as Reinares analysed,[12] continued in 2007 (Table 1).

Table 1. Jihadist deployment in Afghanistan: a profile

In this case, the interpretation of Hypothesis B (w > sq > d) in which there is a causal relationship, and therefore Spain is clearly a target for Jihadists, may be possible if it is understood that Spain’s position in respect of Afghanistan is explained via Hypothesis B of causality instead of Hypothesis A. The simple position of Spain in respect of one or other hypothesis does not precisely determine that an attack will be perpetrated in its territory, but does establish that if there is one a process of destabilisation will be unleashed in which Jihadist groups will focus on a withdrawal of Spanish positions in Afghanistan. If the theoretical positions of the government and society in respect of the causal link behind Jihadist actions are measured,^[13] it emerges that Spain shares Hypothesis B (w > sq > d) with regard to a possible destabilising event linked to Afghanistan.^[14] Consequently, Spain’s position with respect to this conflict is unstable. If this indicator is compared with the survey in the 15^th Wave of the Barometer of the Elcano Royal Institute, 60% of the population surveyed is in favour of a withdrawal of troops from Afghanistan.^[15]

The result of the game model theory pointing towards an increase in the threat against Spanish interests in Afghanistan coincides with the analysis of risk performed by the Elcano Royal Institute’s Global Terrorism Programme.[16] Fernando Reinares conceptualises two groups of players perpetrating violent actions on the ground: the Taliban and al-Qaeda; for both, he has defined not only converging interests but an association in terms of planning, strategy and execution. Accordingly, based on the rational decision model which is an axiom in the theory applied in this analysis, it is assumed that Taliban terrorism and al-Qaeda terrorism respond to a unified decision-making front, with harmonised criteria.

Pakistan
Institutional instability in Pakistan is generating considerable concern internationally, since whoever controls the government also controls the country’s nuclear weapons store. Furthermore, Pakistan is a neighbour of Afghanistan and, just as in the latter country, it has seen a substantial increase in Jihadist activity in 2006. Although there was a considerable reduction in 2007 (see Table 2), the dangers of a possible destabilisation of Pakistan should not be overlooked.

Table 2. Jihadist deployment in Pakistan: a profile

For all of these reasons, while there is not a scenario that might be instrumented directly by the current discourse of al-Qaeda to justify an attack on Spanish territory, this should not be ruled out as a possible future scenario of instability which would affect not only Spain but other members of the international community linked in our model with causal Hypothesis A (d > sq > w), whereby violent actions by Jihadism against Spain would be perpetrated regardless of its presence in international conflict zones related to Muslim countries.

Furthermore, Pakistan borders the areas of Afghanistan where the Jihadist actions are most numerous and it is used by al-Qaeda and the Taliban as an operational base. Accordingly, the destabilisation of Pakistan could add to destabilisation in Afghanistan in order to foment Jihadist actions against the armed forces of countries operating under the NATO umbrella.

Lebanon
Both the current instability in Lebanon and the deployment of Spanish troops there constitute a possible source of destabilisation in which, as in Afghanistan, causal Hypothesis B (w > sq > d) may be applied to explain a terrorist action against Spain’s interests or territory. Assessments of the positions of the government and society in regard to this scenario[17] yield similar results to those obtained in regard to Afghanistan, and the results of the 15^th Wave of the Elcano Royal Institute Barometer reveal that 60% of Spaniards are in favour of a withdrawal of troops from Lebanon.[18]

Despite these observations, in Lebanon there are currently no Jihadist actions comparable to those of Afghanistan or Pakistan (see Table 3). Nevertheless, the presence of foreign troops makes Lebanon an ideal scenario for Jihadist actions aimed at destabilising the international status quo via attacks on troops deployed there.

Table 3. Jihadist deployment in Lebanon: a profile

Algeria
Jihadist activity in Algeria is similar to that of Lebanon. It is smaller in scale than in Afghanistan and Pakistan, although its activity has increased significantly in the last two years (see Table 5). However, unlike Lebanon, in Algeria there are no Spanish troops. Accordingly, Jihadist activities in this territory can only be explained using Hypothesis A (d > sq > w), since there is no situation fitting the Jihadist ideology intervening in the causal chain, at least for now.

Table 4. Jihadist deployment in Algeria: a profile

This means that a possible attack on Spain could not be based on its positions in Algeria, but must be justified by the other two scenarios in which Spain has deployed troops. Actions in Algeria may be strategically aimed, due to its geographical proximity, at destabilising Spain in the rest of Muslim countries.

Proportionally, Algeria has suffered a substantial increase in Jihadist activity in the last few years. This increase coincides with the recent merger between al-Qaeda and the Salafist Group for Preaching and Combat (GSPC).[19] This change in the Jihadist configuration in North Africa is a concern for Europe mainly because its actions in destabilising local institutions will also have repercussions on international security, particularly in Europe due to its geographical and political proximity.

Spain is one, though not the only, of the European countries with most influence on political institutions in North African countries. Consequently, it would not be strange for Jihadists to seek to justify an action based on Spain’s institutional involvement in this region. At all events, this kind of indirect involvement is less tangible in the Jihadist ideology than the presence of troops in the lands of Islam, so that an attack based solely on this scenario would not have the same destabilising force in the international community as in Afghanistan or Lebanon, where there are multinational mandates in place. The lack of sufficient motives for internal destabilisation at international level makes this scenario strategically less appealing to the destructive logic of Jihadist actions aimed at fostering the withdrawal of troops from outside the lands of Islam, all by virtue of the rationale (which is not the only one available, but the one we have modelled here) chosen for this analysis based on game theory.

Conclusions

In accordance with the modelled parameters, we can conclude that Jihadists would find it appealing to oppose Spain via its military presence in Afghanistan and in Lebanon. The first of these two scenarios is strategically ideal in rational terms since, from the Jihadist strategic standpoint, there is a possibility, albeit tiny, that the attacks might be interpreted in line with the hypothesis which understands that the presence of Spanish troops in the lands of Islam ‘justifies’ Jihadist actions in response. Although this hypothesis is not shared in the same way, and nor does it take place in the same political and social contexts as Spain’s involvement in Iraq, it should not be ruled out from the Jihadists’ standpoint, because although unlike in Iraq Spain’s positions, within the framework of NATO (Afghanistan) and the United Nations (Lebanon), are sustained by international institutional strength, Spanish society would prefer not to be involved in these conflicts either.

Pakistan may be seen as a long-term threat linked to Spain via its involvement in NATO’s mission in Afghanistan, since a loss of control in Afghanistan would pave the way for Jihadists to seize power in Pakistan and gain control of its weapons store. Consequently, involvement in Afghanistan may have greater repercussions in the short term as a front for Jihadist attacks on Spain, because any change in favour of the Jihadists there would afford them greater projection in terms of logistical and operating expansion in the region.

Lastly, another front for institutional and/or ideological destabilisation, quite feasible from the Jihadist standpoint is Algeria. Jihadist actions will likely not trigger major destabilisation internally, but will be broadly covered by the media and will sound alarms in countries that are geographically and politically close by, in other words, in Europe, most notably in Spain and France.

Accordingly, Spain faces the Jihadist interest in operating and territorial expansion in Afghanistan and ideological expansion in Algeria. To combat and cancel out the destabilising effect of Jihadist actions linked to any of these scenarios, Spain must tackle them in unity and stability, while at the same time it must prevent any of its involvement in Islamic lands, pursuant to its membership of international organisations, from being interpreted in line with the hypothesis that says that there are contextual situations which would justify a possible terrorist attack.

The game theory model applied in this analysis predicts that Spain is still a potential target in the Jihadist strategy of destabilisation (see Appendix). The absence of troops deployed in Iraq does not cancel out Spain’s presence in Lebanon and Afghanistan (where Jihadist actions have increased notably of late). However, the effectiveness of attacks aimed at destabilising Spain’s position in the international system would not be valid if there were no position of instability in respect of the causality of a Jihadist action. If the position is common with Hypothesis A, which interprets that there is not a situation that intervenes in the causal link, then there is no instability or reason for Spain to alter its positions in the international system, so that the strategic value of a Jihadist action would be undermined. However, if there are reasons whereby a Jihadist action might be explained in line with Hypothesis B, such as in Afghanistan and Lebanon, where there is a situation that intervenes in the causal link, then there are indeed strategic reasons for a Jihadist action to attain its objectives of destabilisation and eventual withdrawal of positions from the lands of Islam. Notwithstanding this possibility, Spain’s position in respect of both positions is intermediate, so that despite its involvement in Afghanistan and Lebanon, it is in a much more stable position that in the context of 11 March 2004.[20] Consequently, it should not be ruled out that Spain may be targeted again by Jihadists, although due to internal stability it is not a clearly favourable objective for the strategy of destabilisation which Jihadist organisations pursue.

Miguel Luparelli Mathieu
PhD in Economics and International Relations (Universidad Autónoma de Madrid) and associate researcher at the Sociedad Española de Psicología de la Violencia

José A. Mathés
Analyst-methodologist at the Bureau for Internal Security Studies, Office of the Secretary of State for Security

Andrés Montero Gómez
Director of Intelligence at Interligare, Professor of Security at the Universidad Nacional de Educación a Distancia and Professor of Security Management at the Universidad Rey Juan Carlos

Appendix: formal interpretation of the estimates

Chart 2. Interpretation of the variable hdsq

Chart 3. Game theory model for Afghanistan

In the profile [EICI] there is a situation of possible cooperation because for both Jihadists and Spain: w > sq > d. Accordingly, in the profile [ECI] it can be observed that for Jihadists w > d > sq and for Spain w = sq > d. Here, to, there is a situation of cooperation. Therefore, from the perspective of Jihadist ideology, to attack Spain could unleash a reaction that would be favourable for its objectives.

Bibliographical References

Atkinson, S.E., Sandler, T. and Tschirhart, J.T. (1987), ‘Terrorism in a Bargaining Framework’, Journal of Law and Economics, No. 30, pp. 1-21.

Bermejo, R., and Reinares, F. (2007), ‘International Terrorism and Public Opinion in Spain’, ARI 32/2007, Elcano Royal Institute.

Luparelli Mathieu, M.S. (2007), La acción yihadista – Estrategia de las organizaciones yihadistas Salafistas y respuesta de la Comunidad internacional, doctoral thesis, Universidad Autónoma de Madrid.

A. Montero, ‘Crítica de la razón bélica contraterrorista’, Sistema, No. 193, July 2006, pp. 121-129.

Morgenstern, O., and Neumann, J. (1947), The Theory of Games and Economic Behavior, Princeton University Press.

Reinares, F. (2007a), ‘What Threat Does Jihadist Terrorism Currently Pose to Spain?’, ARI 33/2007, Elcano Royal Institute.

Reinares, F. (2007b), ‘Terrorismo Talibán’, Política Exterior, No. 120, pp. 57-66.

Reinares, F. (2007c), ‘Bajo sombras de espadas: al-Qaeda y la amenaza terrorista en el Magreb’, Claves de Razón Práctica, No. 175, pp. 16-22.

Sandler, T., and Arce, D.G. (2003), ‘Terrorism and Game Theory’, Simulation and Gaming, No. 34, pp. 319-337.

Sandler, T. and Lapan, H.E. (1968), ‘The Calculus of Dissent: An Analysis of Terrorists’ Choice of Targets’, Synthese, No. 76, pp. 245-261.