All Deterrence Is Local: The Utility and Application of Localised Deterrence in Counterinsurgency

Abstract

The deterrence of non-state actors is a relatively understudied and not particularly well-understood aspect of deterrence studies. This chapter contributes to the emerging body of knowledge on this matter by coining the idea of localised deterrence. Based on a discussion of counterinsurgency theory it is argued that tailored measures can be effectively employed for deterring violent non-state actors by targeting their relations with the local populace. Subsequently, this chapter explores theoretical as well as practical aspects of localised deterrence in order to explain how this concept can be conceptualised and operationalised to effectively deter insurgents and their supporters amongst the local populace. Ultimately, this allows us to reflect upon the concept and set an agenda for embedding localised deterrence within the wider body of deterrence studies by identifying new avenues of research.

1 Introduction

Deterrence has always been a crucial element in the fight against irregular opponents.¹ Due to the nature of such adversaries this application not only required deterring the actual fighters and other members of irregular groups, but also their supporters among the local populace. In colonial warfare, for example, this led to the wide-spread use of exemplary force as a means for punishing locales in which irregular adversaries thrived, while also sending a message to the opponent as well as the wider population. Gradually, however, an awareness emerged that in the long run the brutalities in which this approach typically resulted outweighed the success on the short term as the government’s authority was almost continuously challenged. Moreover, especially in case of liberal democracies, the shifting of ethical, judicial and political standards rendered the use of force against the local population no longer acceptable. While all of this altered the application of deterrence in irregular warfare, practices like starvation and forced migration could be observed well into the second half of the twentieth century. In today’s Western interventions such methods have been completely replaced as the emphasis lies on the use of persuasive methods for winning the support of the local population. But, how is it then possible to deter irregular opponents such as violent non-state actors that interact with and hide among the local population?

Deterrence studies first emerged during the Cold War in which the concept was crucial for maintaining a strategic balance between the two rivalling superpowers. From this background, it is fully understandable that the field has traditionally focused on interstate competition at the strategic level. More recently, however, violent non-state actors have gained an increasingly important role in the international security environment. It is not surprising, therefore, that there is a growing interest in the utility of deterrence vis-à-vis such opponents. In this regard, Chap. 14 by Shamir and Chap. 19 by Jakobsen offer valuable and much-needed insights for forwarding the understanding of the concept’s application outside its traditional context. Yet, overall deterring violent non-state actors remains a relatively understudied aspect of deterrence studies. Partially, this can be ascribed to the tactical nature of this specific form of deterrence. Contrary to the classical notion, which emphasizes strategic level efforts to change the behaviour of states in the international system, the utility of deterrence against non-state actors typically involves ‘the immediate behaviour of individual people or small groups’ with potentially strategic implications.² Wilner, among others, has repeatedly pointed at the far-reaching ramifications of this different focus for deterring extremist violence.³ In such cases punishment and denial—the mainstays of classical deterrence—do not suffice as it is also crucial to target what the opponent believes through delegitimisation. Consequently, the cost-benefit analysis underlying the application of deterrence should take the perception of the target audience into account. Influencing the opponent requires a thorough understanding of cultural values and circumstances. A profound insight into the local context, thus, is key for successfully deterring violent non-state actors. Or in other words, in such cases all deterrence is local.

This chapter seeks to contribute to the expanding body of knowledge on the deterrence of non-state actors by focusing more closely on what we label ‘localised deterrence’; i.e. the tactical application of deterrence based on the specifics of the local situation. Thus far, research in this field has mainly focused on the fight against terrorist organisations.⁴ We aim to broaden this scope by discussing the localised nature of this particular form of deterrence from the perspective of counterinsurgency. The Global War on Terror and the fight against the Islamic State, after all, have demonstrated that dealing with extremist groups might not only require the implementation of a counterterrorism strategy, but also the conduct of counterinsurgency in unstable countries. Moreover, as today’s violent non-state actors have become increasingly capable of interacting with populations and challenging weak governments, fighting insurgencies will remain a relevant theme in future warfare. The renewed competition for global influence among big powers, in this regard, is expected to augment this as ‘grey zone tactics’ might encompass the exploitation of non-state groups as proxies in order to destabilize states.⁵ Our focus on counterinsurgency, thus, not only serves to widen the scope of the field, but is also prompted by the reality of modern warfare. Yet, this choice might not be obvious to all as counterinsurgency is mostly associated with ‘winning the hearts and minds’, which seemingly does leave little space for deterrence. As we will explain below, however, counterinsurgency as a form of irregular warfare in the first place is a highly dynamic contest between adversaries that inherently—and inevitably—involves coercion and therefore also deterrence.⁶

While we here coin the term localised deterrence, this chapter does not intend to build a theory or provide a comprehensive overview of its application. Our contribution should be considered a first attempt to introduce the concept as a distinctive object within the field and to identify new avenues for advancing the state of knowledge in the area of deterring non-state actors. To do so, we will first discuss the utility of localised deterrence in counterinsurgency from a theoretical perspective. This allows us to subsequently expand on this basis to obtain a fundamental understanding of the concept and identify germane issues with regard to its practice as well as the adjoining intricacies and challenges. Ultimately, the combined theoretical and practical insights serve to reflect upon the concept and set an agenda for expanding the body of knowledge by introducing localised deterrence as a specific subject of deterrence studies.

2 Explaining the Local Turn

Why does localised deterrence matter in counterinsurgency? Answering this question first requires us to clarify how the concept fits in with counterinsurgency theory. Therefore, our exploration starts with a discussion of the logic underlying counterinsurgency which will allow us to understand how localised deterrence might be used for fighting insurgencies. This section thus explains the rationale of localised deterrence as a tool of counterinsurgency.

In counterinsurgency state and non-state actors vie for control over and support of the population. As such, obtaining legitimacy is a key objective for all parties involved in this violent struggle.⁷ Legitimacy is essentially understood as the acceptance and justification of authority based on social constructs, and is influenced by local culture, norms, values and beliefs.⁸ Hence the importance of gaining a local understanding, as, among others, reiterated in NATO’s AJP-3.4.4. Counterinsurgency (2016), which states that ‘an effective counter-insurgent force needs to have a(n) (often hard-earned) cultural understanding as well as a more general understanding of the societal, economic and political landscape of the affected country’. Or in other words, in counterinsurgency ‘societal context is king’.⁹ A profound insight into the local social landscape and culture is not only pivotal for obtaining legitimacy, but also for denying support to the adversary, to influence relevant audiences on the basis of their motivations and drivers, and to obtain intelligence about the insurgents and their activities.¹⁰ Ultimately this boils down to the fact that in counterinsurgency, a thorough understanding is imperative in order to obtain control over the target society while simultaneously preventing the opponent from doing so. Kilcullen, in this regard, has conceptualised this struggle as ‘competitive control’ which typically is won by the actor that manages to impose a predictable, normative, ordered system ‘that tells them [the people] exactly what they need to do, and not to do, in order to be safe’.¹¹

Thus, both insurgents and counterinsurgents, compete to establish, enhance and maintain control over the population, denominated contested control.¹² Control, in this context, is defined as the capability of one societal agent to influence the circumstances of action of others.¹³ Insurgents use various violent methods to obtain such control, including terrorism, which is characterised by the threat or use of violence to obtain publicity and spread a message of fear.¹⁴ While terrorism is a tactic that can be employed by insurgents, insurgencies typically rely more heavily on political subversion and guerrilla warfare to exploit and address political grievances for achieving local change. Whereas terrorists do not necessarily need popular support, insurgents require the support (or at least acquiescence) of the population to win the struggle for control. Consequently, insurgencies employ hybrid strategies that not only target the state and its incumbents, but also seek to influence the relevant populace by either reaching out to the people or by imposing fear through the use of systematic terrorist attacks.¹⁵ It is important to reiterate here that work in the field of deterring non-state actors has typically focused on terrorism and extremist organisations who seek to influence their target audiences through sustained (symbolic) violence. Whereas the labels for different types of violent non-state actors are often used interchangeably—especially by politicians, insurgencies are more complicated due to their hybrid character and their intricate interactions with the local population. Contrary to terrorists, insurgents might prefer political subversion and only employ terrorist tactics if necessary. Consequently, this chapter’s adoption of the counterinsurgency perspective also serves to widen the scope of deterrence theory with regard to its application against violent non-state actors.

Counterinsurgents, on their turn, seek to obtain control through either collaboration with the population or by establishing dominance through the use or threat of force. The Western model of population-centric counterinsurgency focuses on the former and is often characterized as the ‘hearts-and-minds’ approach. Its emphasis lies on gaining popular support by enhancing state legitimacy through the use and creation of persuasive programmes and incentives as well as the provision of security.¹⁶ The second approach, also known as the authoritarian model, seeks to establish unquestioned dominance by coercing both the enemy and the population.¹⁷ In reality the divide between these two models is more nuanced, since they both use a mix of coercive and persuasive strategies—though the West typically focusses on persuasion, and authoritarian regimes more on coercion.¹⁸ Additionally, the exact approach is not only influenced by regime type, but also by other factors like, for instance military culture.¹⁹ While both models use different methods and means, they share the goal of enhancing and spreading state control.²⁰

On the other hand, the authoritarian approach is problematic from a Western perspective due to its reliance upon force. The Western model itself overemphasises persuasive hearts-and-minds methods, which might be insufficient for establishing control. This is supported by the historical track record of Western counterinsurgency which proves that success is notoriously hard to achieve, as has yet again been illustrated in Iraq and Afghanistan. Moreover, today’s conflicts have become even more complicated as they typically involve several types of actors with different motivations and interests. Consequently, depending on the local situation, coercion—compellence, and/or deterrence—might prove to be more effective.²¹

This inability to incorporate a broader range of methods for increasing the effectiveness of the counterinsurgency approach at the grass roots level is further enhanced by the fact that strategic decision-making almost exclusively takes place from a top-down perspective.²² Contrastingly to counterinsurgency’s localised nature, the crucial bottom-up view needed for understanding the way the local context informs people’s choices, or ‘strategies of survival’, in a situation of violent contention is lacking.²³ This further hampers the ability to determine which tool(s) on the continuum ranging from persuasion to force (which will be dealt with below) should be employed for attaining the desired effect of enhanced control over the populace.²⁴ Indeed, instead of taking a top-down, outsider or etic perspective, effective counterinsurgency first and foremost requires a profound insight into the perception of the local population; the so-called emic perspective.²⁵ The resulting understanding serves as an underpinning for deciding upon the most effective course of action at the grass roots level. This might include deterring people from taking an active or supportive role in the insurgency as well as other measures to deter non-compliance with the government. But how does exactly such deterrence materialise?

3 Localised Deterrence

The local character of counterinsurgency implies that whenever deterrence is applied, this should be tailored to the specifics of the locale in which the struggle for control takes place. Hence, we have labelled this highly-specialized form localised deterrence as it can be applied against violent non-state actors that interact—either through force or persuasion—with a particular society. But how does this approach work? What exactly should be deterred and who should be deterred for that sake? Why do these actors exactly behave in an unintended way? And which methods can be employed for obtaining the desired change in behaviour? In order to conceptualise localised deterrence as a distinctive topic within deterrence studies it is pivotal that we answer these questions from a theoretical perspective as well as discuss salient facets of its operational application.

As mentioned afore, adopting an emic perspective is instrumental in understanding the local context through the eyes of the target audience. Decision-making on the basis of this information allows for customizing the cost-benefit analysis underlying the application of deterrence and thereby overcoming traditional coercion theory’s universal bias which assumes that all actors will act in the same way, be it individuals or nations.²⁶ Equally important, a thorough emic understanding also serves to design effective measures of deterrence. In this regard, deterring an opponent or sympathiser among the population is typically achieved through a range of methods that are part of the broader coercive framework and which can be depicted in what we call the influence continuum (Fig. 15.1). As mentioned afore this continuum has force on one extreme and persuasion on the other, and typically all parties in a conflict seek to effectively influence other relevant actors through (a mix of) coercive and/or persuasive methods.²⁷ Whereas the former focuses on altering behaviour by depriving freedom or autonomy through (the threat of) force or other measures, the latter, for this same purpose, exploits an actor’s perception of free choice through intentional messaging.²⁸ As the term continuum implies, effectively influencing an actor might require to shift between coercive and persuasive methods or to employ a mix of measures in order to attain the desired effect.

Fig. 15.1

The influence continuum (Fig. 15.1 is based upon Kitzen 2016, p. 101.)

Before deciding upon a method, however, effective localised deterrence first presupposes a clear purpose of what to achieve. When seeking to influence an opponent or social group, it is pivotal to determine which behaviour needs to be deterred. After all, this determines who exactly should be targeted and what specific method(s) should be employed to achieve the intended behavioural change. All of this should be based on a thorough situational awareness that results from an emic understanding of the local conflict dynamics. Such an operational approach to deterrence is described by, among others, Mackay, Rowland, Tatham and Van Kuijck who all recognise that exerting influence must be tailored to the local context.²⁹ For that purpose, it is pivotal to adopt a multidisciplinary method that allows for identifying triggers and circumstances of behaviour, as well as the different relevant target audiences. Of course, this serves the aim of effectively changing undesired behaviour and attaining the deterrer’s goals. In order to obtain a more profound understanding of an operational approach of localised deterrence, we will now deal with its different facets, starting with the rationale of targeting specific behaviour.

4 What to Deter?

As aforementioned, effective deterrence requires an actor to determine which non-desired behaviour it wants to stop or which desired behaviour it wants to maintain. Often, however, goals remain vague and are not specified into relevant behaviours. In counterinsurgency, for instance, the main aim is to establish control over the population. This, typically, is translated in sub-goals like securing the people, isolating the insurgents, neutralizing the insurgent’s subversive effort (while simultaneously enhancing the government’s authority) and armed organization, as well as creating unity of effort among all friendly actors involved in the campaign.³⁰ Although all of these objectives are relevant, they should be further specified into relevant behaviours that can be targeted. But why is it so important to focus on such specific behaviours?

Whereas traditionally the hearts-and-minds approach emphasizes attitude—‘minds’—, this is a particular poor focal point for changing behaviour in circumstances of violent contention. Whilst different attitudes are a factor of consideration, they are only predictive for behaviour under specific conditions that mostly are absent in a conflict zone or do not apply to relevant behaviour.³¹ Such specific behaviour itself, however, is a useful indicator that can be measured.³² An increase in intelligence about the adversary’s intents and actions received from the population, for instance, reveals much about a desired behaviour. A decrease in friendly interactions between counterinsurgents and local population, on the other hand, gives a clear indication of the rise of an undesired behaviour. Specific behaviour, therefore, offers a measurable and palpable focal point on which to act.

In the context of deterrence, the main goal would be to identify and forestall non-desired behaviour. This, consequently, should be operationalized as a SMART objective, i.e. being specific, measurable, achievable, relevant, and time-bound.³³ In a struggle for control over the population, it is crucial to determine which different behaviours from both insurgents as well as the local populace should be deterred. Whereas counterinsurgency has traditionally focused on attitudinal approaches such as bringing about a shift from impartiality towards support for the government, this has remained very much an intangible and unmeasurable objective. Focusing on behaviour overcomes this problem by first seeking to understand why the population matters. As people, among others, provide food, funding, supplies, resources, sanctuary, information and recruits to the party they support, various relevant behaviours can be distinguished.³⁴ Examples of non-desired behaviour include collaborating with the insurgents by taking up arms, placing IEDs (Improvised Explosive Devices), delivering food, money, or other specific resources, and withholding crucial information from the counterinsurgents. Similarly, insurgent behaviours such as murdering, maiming or raping civilians, spreading propaganda and night-letters, kidnapping of important key figures, armed robbery, theft, and destruction of property can be targeted.³⁵ For instance, the Taliban still relies on opium, hashish trafficking, and levying taxes for their funding.³⁶ This would entail different behaviours under the concept of “Taliban Funding”, such as poppy cultivation, the transportation of opium, and the use of smuggling routes. Looking at poppy cultivation, one needs to understand which different behaviours form this specific non-desired behaviour, which can include exchanging money for protection of the crops, buying relevant materials for growing poppy, and planting the seeds. Campaigns were conducted between 2006 and 2013 (with exaggerated reports of crop destruction and allegations of corruption) to deter the population from growing poppy and the funding of the Taliban, but the opposite happened. Instead of understanding that poor households prefer low-risk activities in the high-risk environment of Afghanistan and addressing this issue by taking into account the various behaviours, the campaign resulted in the rise of antipathic behaviour towards the government.³⁷

Wilner has reiterated the importance of identifying such measurable factors like behaviour for deterring violent non-state actors.³⁸ Moreover, he has pointed out that with regard to the use of force, measuring the amount of violence is insufficient. Instead, assessing the type, nature, time-period and changes of violence will offer a better understanding of the behaviour involving the use of force and the way to address it. Thus, effective deterrence first and foremost requires identifying the target behaviour, which subsequently allows for identifying the relevant audience.

5 Whom to Deter?

Elsewhere in this volume Jakobsen states that an increasing number of actors can be potentially deterred in modern conflicts. In counterinsurgency, a myriad of actors is present. Typically, this is perceived in terms of a distinction between friendly, neutral, and hostile audiences.³⁹ Furthermore, active and passive supporters of both sides play a role in the violent struggle, and different categories of membership of insurgencies can be distinguished (leaders, armed elements, cadres, auxiliaries, underground, and mass base). Effective deterrence, however, requires a focus on the group(s) most relevant to the defined undesired behaviour. Hence, identification and segmentation of involved actors is pivotal and should be as specifically as possible.⁴⁰ Again, it should be stressed that this depends on the target behaviour. For example, if a state seeks to deter foreign fighters from joining insurgencies, it not only needs to determine who travels to the conflict zone, but it also should identify the relevant intermediaries. The more specific the behaviour is, the more accurate the relevant target audience can be identified in order to design and implement an effective intervention strategy.

Depending on the local context relevant target audiences can be segmented in categories such as gender, country of origin, age, occupation, and socio-cultural or ethnical background. In case of the aforementioned example the target audience, for instance, consists of male and female foreign fighters aged 18–25 from Europe that travel to another country to join an insurgency, and smugglers instrumental in providing travel tickets and false passports.⁴¹ Another illustration is provided by the identification of vulnerable groups that might be potentially tempted to actively support an insurgency with the state they live in. Boko Haram’s target audience in Nigeria e.g. can be divided into, among others, illiterate minors with difficult upbringings, and unemployed young males.⁴² A second real-life example of the diversity of the relevant audience is provided by of Al-Shabaab, which receives funds from different sources and groups, either voluntary or involuntary. This, among others, is illustrated by the extremist group’s tactic of trading and taxing of charcoal. When seeking to deter Al-Shabaab’s funding of insurgent groups, this requires not only influencing behaviour of its members, but also of truck drivers and smugglers who transport charcoal from different locations, as well as the manufacturers and labourers who produce that charcoal.⁴³ Identifying the relevant audiences, however, can even become more complicated in case of violent non-state actors consisting of an intricate web of interconnected networks as is the case with Al-Qaeda in the Islamic Maghreb (AQIM), which receives funds through various sources. One such source concerns a minor criminal activity involving sales of luxury vehicles through family networks and individuals in Guinea-Bissau and Mauritania. In this case, relevant audiences would include sales persons and companies of luxury vehicles, the transporters of the luxury vehicles, family members that provided transfer services in order to trace the transactions, and the high-end buyers that can afford and actually buy these luxury vehicles.⁴⁴

Since it is often impossible to target all identified audiences, it is necessary to make sense of the different categories and prioritize them for the purpose of deterring non-desired behaviour. Therefore, relevant groups should be categorized in different spheres such as the actual target audience, those reacting positively to the application of deterrence to the main audience, those reacting negatively, and those ambivalent.⁴⁵ Using the example of AQIM, the target audience consist of persons and groups that are not members of AQIM and knowingly engage in the luxury vehicles sales process to provide money to AQIM. The audience that reacts positively to the deterring ‘message’ is made up of persons and groups (family members, local leaders, and high-end buyers) unknowingly supporting AQIM through the luxury vehicles sales process, and who can make a difference in the behaviour of the actual target audience. AQIM members as well as those involved in transfer of funds form the audience that can react negatively to the measures applied to the main audience. The audience that is ambivalent and might even be best left alone consists of transporters of the luxury vehicles. It should be mentioned here, that this categorization remains hypothetical as any accurate classification of different audiences should be based on research from a local perspective (which is outside the scope of this contribution). This, however, does not affect the pivotal point that a thorough and as specific as possible identification of the relevant audiences is instrumental in considering how the target behaviour can be deterred.

6 Driving Factors for (Un)desired Behaviour

To determine how to change undesired behaviour, it is a prerequisite to learn the triggers and factor underlying that specific behaviour. As touched upon in the previous paragraph, once relevant behaviours and audiences are identified, a more profound analysis should be conducted at the local level. This encompasses in-country vetting by use of the local language(s) in order to consider as many salient issues as possible. As repeatedly echoed throughout this chapter, it is crucial to know who exactly the audiences are and how these people understand the world from their lens or perspective; behaviour can only be changed by adopting such an emic perspective. Even a seemingly relatively simple or straightforward matter like the term ‘evil’ or ‘wrong’, might be perceived completely different from the local perspective.⁴⁶

This insight has been observed time and again in the counterinsurgency campaigns of the last two decades, but it seems very hard to learn this lesson and incorporate an emic perspective in decision-making at even the tactical level. A striking example is provided by an attempt to reduce the number of IED-strikes and American deaths through an information operations (IO)-campaign in Afghanistan.⁴⁷ Portraying US soldiers as friendly people devoted to their family succeeded in positively changing the local attitude towards them. Yet a change of undesired behaviour—the placing of IEDs—did not occur. An evaluation brought to light that there were two reasons for this failure. First, those placing the IEDs turned out to be extremist, whose attitude and behaviour could not be changed in this way. More important was the second finding that the IO’s campaign use of images of soldiers and their families also conveyed the message that life in the US was good. Unintendedly, this prompted the people that were producing and transporting IEDs to step up their effort. Since financial reasons were a primary motivation, it turned out to be that the portrayed richness of live in the US made them want to make more money in order to immigrate to that country. Thus, contrary to the desired effect, the locals involved boosted production and transportation which resulted in an increase of IED strikes. The Americans, from their perspective, did everything right in their attempt to influence the population, the audience’s perception, and specifically those involved in producing and transporting IEDs. Yet, due to a poor insight into the local perspective. it did not lead to the expected desired outcome, but far more to the opposite.

Employing effective methods of localised deterrence, thus, requires an equally localised understanding of the undesired behaviour and target audience. For that purpose, it is pivotal to adopt a multidisciplinary approach that combines insights from various academic fields (e.g. communication, anthropology, social psychology, and neuroscience) for conducting bottom-up research in order to answer the crucial questions of what, who, why, and how.⁴⁸ After all, the actual application of localised deterrence first requires a more detailed knowledge of what exactly is the relevant behaviour and who makes up the target audience, why people engage in the relevant (non-)desired behaviours, and how this can be changed in an holistic manner that fits the local context. These different aspects, thus, not only enhance respectively descriptive and prognostic information about the audience and targeted behaviour, but also provide an insight into which transformative and tangible method is most effective for attaining the desired behavioural change.

7 How to Deter?

Although that it is impossible to deal in this chapter with all different methods that can be used in localised deterrence, the influence continuum (Fig. 15.1) clearly demonstrates that methods may vary from the use of force to pure (verbal) persuasion. Moreover, depending on the local situation and more specifically the nature of the involved actors, shifting between different coercive and persuasive methods or adopting a mix might be required.⁴⁹ While ultimately, a profound understanding of the local social landscape and conflict dynamics should inform the preferred method(s), this conceptual exploration benefits from distinguishing between the three main categories of audiences that shape the human terrain in which the struggle for control is fought. In this regard it is important to notice that local people’s strategies of survival typically boil down to weighing sanctions and incentives in order to limit the damage suffered from the conflict and, when possible, to strengthen their position.⁵⁰ Furthermore, counterinsurgency is mostly conducted in developing countries with a so-called web-like society and hampered by a scarcity of resources. Due to this combination, effective engagement of different groups focuses at leaders and the way they seek to address the people’s strategies of survival. Although, yet again, this heavily depends on the local situation, we have opted here to follow this logic since it is impossible to discuss all methods of localised deterrence. Moreover, it allows us to provide a structural oversight of methods, and to discuss the practical application of these methods.

The first category consists of those friendly to the counterinsurgency. These people to a greater or lesser extent collaborate with the government. The undesired behaviour, therefore, is that they diminish or stop their collaboration, or even switch sides and turn towards the insurgency. According to the logic of Western, population-centric counterinsurgency all of this should not be deterred by use of force, but rather by persuasive methods that reward cooperation. If necessary, this might be accompanied by the limited use of non-violent, so-called soft coercion such as the (threat of) withdrawal of military, political or economic support.⁵¹ Deterrence, thus, is provided by the denial or loss of benefits. This can be addressed through, for instance, a strategy of co-option with local power-holders.⁵² The counterinsurgency campaigns in Iraq and Afghanistan have demonstrated that such a strategy might be particularly effective in gaining increased support for the government, as epitomized by the Sunni awakening movement in Iraq. Yet, while it sometimes took several years before counterinsurgents gained sufficient understanding in order to determine who should be engaged and what methods were most effective, the most challenging aspect was to remain in control of co-opted local power-holders. As such actors typically act to maximise their self-interest, the threat that these allies either diminished their collaboration or even switched side was always very real. Localised deterrence, therefore, was instrumental in implementing an effective system of checks and balances that kept the counterinsurgency in control and mitigated undesired behaviour.

In the second category are those neutral to the conflict, the so-called fence-sitters. In this case, the undesired behaviour not only concerns (the start of) collaboration with the insurgency, but also all other actions with a spoiling influence. Again, the Western emphasis on ‘hearts and minds’ prescribes that such an audience is preferably engaged through persuasive methods backed up by soft coercion if necessary. The difference with the previous category is that in the case of neutrals soft coercion is more accepted as a method to enforce compliance. One particular powerful illustration of the effectiveness of localised deterrence in such a case concerns the targeting of an actor’s power basis through manipulation of his supportive network and the empowering of rivals. The thorough vetting of the local human landscape will help to identify local political players as well as the pattern and structure of political authority in the target society. Based on this advanced understanding potential spoilers or opponents can be singled out and their power base might be attacked by (the threat of) reaching out to rivals willing to cooperate.⁵³ This, for instance, has been practiced repeatedly in Afghanistan to curtail the power of warlords who threatened the international effort and whose spoiling behaviour could be effectively deterred in this way.

The third and last category is made up of those hostiles towards the counterinsurgency, i.e. sympathisers among the population and active members of the insurgency. Here the undesired behaviour that should be deterred is a continuation of collaboration with the insurgents or further participation in the insurgency. For the sake of this, the use of different coercive methods, varying from the use of force to soft coercion, is widely accepted—although in Western counterinsurgency compellence might also include persuasive methods. For the application of localised deterrence, it is useful to distinguish between irreconcilable extremists and more moderate reconcilable individuals. The latter can be engaged either through soft methods or by (the threat of) force in order to make them comply or even to convince them to switch sides. Yet, as an example from Afghanistan demonstrates, this is difficult to accomplish as military organizations are used to deal with adversaries by use of force.⁵⁴ In this particular case a Taliban leader’s local power-base was effectively eroded which led him to offer a switch towards the government side. Since this individual was already listed for either being captured or killed, it proved impossible to accommodate his proposal. Thus, a huge chance to turn the local situation was missed due to a lack of flexibility from the side of the counterinsurgents. This also points to the prominent place of targeting in Western military operations.⁵⁵ While this might not always be the best option for deterring behaviour of the more moderate individuals, it is highly effective for use against the more extremist, irreconcilable senior and mid-level leaders. In most cases this specific method does not so much function to deter the involved individual, but far more aims to enforce compliance from the wider (relevant segments of) insurgency and its supporters. A so-called strategy of decapitation, indeed, can have a profound impact on the organisation as a whole.⁵⁶

These insights into the application of localised deterrence for influencing behaviour in different groups allow us to conclude this section by drawing up a taxonomy of localised deterrence in counterinsurgency. Conceptually, three different approaches can be discerned. Each approach is tailored to a specific audience and relevant behaviour, and therefore relies on a specific mix of methods. Yet, it should be emphasized again that implementing the concept first and foremost remains a local matter. Only when the salient questions of what, who, why, and how are answered on the basis of a profound understanding of the local, emic perspective, localised deterrence can be effectively implemented in the practice of counterinsurgency warfare.

8 Conclusion

This chapter has explained why understanding the local context is essential for effectively deterring undesired behaviour. Such localised deterrence is crucial since not all actors make the same cost-benefit analysis and the population at the grass roots level, as a consequence of the increasing role of violent non-state actors, has become far more important in modern conflicts. The re-emergence of global power competition is expected to further contribute to this trend as non-state groups will be increasingly exploited as proxies in grey zone warfare. Counterinsurgency, in this regard, provides some answers for fighting violent non-state actors that intricately interact with the local population. As the struggle for control over the population essentially is a local fight, counterinsurgency prescribes understanding the local culture and perspective. Yet, Western counterinsurgency does not provide a sufficient answer since its focus on attitudes and persuasive methods does hamper the adoption of an effective mix of methods for deterring non-desired behaviour. Rather, deterrence and compellence are both sides of the same coin that can be achieved through a combination of coercive and persuasive tactics, depending on both the objective and the local context. Moreover, the top-down character of strategic and military decision-making often prevents developing and incorporating a true emic understanding of the local social landscape. Hence, we urge for a change in mind-set and approaches towards both counterinsurgency and deterrence.

Localised deterrence offers an approach for stopping non-desired behaviour or maintaining a (desired) status quo on the basis of a profound emic understanding of the local context and conflict dynamics. This not only allows for properly identifying a specific behaviour, but also for selecting the right target audience, the triggers and motives of relevant behaviour as well as the tactics that are most fitting in a certain environment. The concept, thus, has triggered the development of novel approaches that operationalize localised deterrence by answering the pivotal questions of what, who, why, and how. Although first and foremost a local approach, our exploration of the practical application of the concept in counterinsurgency has allowed us to draw up a conceptual taxonomy. According to this categorisation three different types of audiences (friendly, neutral, enemy) with each a specific undesired behaviour and a unique mix of methods can be distinguished. Furthermore, it was also found that challenges in the practical application of localised deterrence not only concern the development of a local understanding and the subsequent design of an effective approach, but also involve the adoption of a flexible mind-set that allows for effectively shifting between various methods. Yet, this chapter has provided some examples in which localised deterrence was effectively incorporated in the struggle for control in recent counterinsurgency campaigns.

Newly available tools offer a far more structured and methodological strong approach towards the concept. Hence, we would like to conclude by suggesting to adopt localised deterrence as a distinct topic of deterrence studies and explore new avenues of research in order to contribute to the emerging body of knowledge on the deterrence of violent non-state actors. This is especially important as much of the empirical evidence has not yet been published. By setting a research agenda the field can optimally benefit from the data of recent cases in, among others, Iraq, Afghanistan, and the Sahel, which will become available over the next years. We, in this regard, would like to suggest to study the fundamental value of the concept outside the context of counterinsurgency, to further explore the way the questions of what, who, why, and how can be answered, and how the concept can be implemented most effectively. This will not only contribute to the further conceptualisation and theorization of localised deterrence, but also will shed a light on the wider practical utility of the concept against violent non-state actors who are increasingly exploiting their networked ties with local populaces throughout the world—either for their own purposes or as proxies in the global power competition. The strongest contribution to the field, in this regard, is expected to come from localised deterrence’s departure from the one-fits-all approach as originally entrenched in classical coercion theory. Since all behaviours are influenced by the specific local context, culture, and perceptions of the audience, it is pivotal to further explore this new strand in deterrence theory.