Abstract
Introduction
The concept of deterrence is widely understood and has been used throughout history at all levels of human interaction. In displays of their military prowess to warn off challengers, ‘…the Romans understood all the subtleties of deterrence…’ (Luttwak, 1999, p.3). Public and brutal punishments for crimes in the period before the 20th century used a ‘…shrewd sense of theater and timing…’ to deliver a wider deterrent effect (Ignatieff, 1978, p.21).
Whilst deterrent strategies may be employed across a wide landscape of human relations, the concept regularly features in discussions amongst researchers and practitioners within two fields: Security Studies and criminology. Deterrence is of interest to scholars in the first field due to its potential to prevent conflict among states, societies or other groups. In criminology, deterrence relates primarily to the use of the criminal justice system to reduce offending through the prospect of punishment.
In the realm of criminology, the traditional model of deterrence is rooted in the belief that ‘… threats can reduce crime by causing a change of heart, induced by the unpleasantness of the specific consequences threatened’ (Zimring and Hawkins 1973, p.75). General deterrence is intended to deter by making examples of offenders to deter the wider community. Specific deterrence is intended to deter further offending by those already known to the criminal justice system or those who are deemed likely to offend (Bjørgo, 2016, p. 21). Along with debates about the effects of certainty, celerity and severity of punishments, the concepts of general and specific deterrence have engaged criminologists for many years. A more recent development in the field attracting great interest is Focused Deterrence whereby a sub-group of putative offenders is targeted with communication of likely consequences should they commit crime.
Most people could provide a reasonable description of what interests criminologists and the role of deterrence in the criminal justice system. The general public would probably not display the same level of awareness in relation to Security Studies and the relevance of deterrence within that discipline. Security Studies is a sub-field of the academic discipline of International Relations. It was not until after the failure of deterrent alliances and coercive diplomacy that resulted in First World War that the first university department dedicated to the study of International Relations was established at Aberystwyth in 1919. Unsurprisingly, most research at that time revolved around interstate rivalries, the use of military power and the use of deterrence to preserve the status quo between nations. The events of Second World War and the subsequent Cold War did little to change this focus (Sheehan, 2022, p. 204).
Two types of deterrence were seen as important in the context of the large power/power bloc rivalry of the Cold War. Deterrence by punishment aimed to make a would-be opponent view a conflict as potentially too painful to endure. Deterrence by denial sought to deny an aggressor the benefits of any aggression. In a cost-benefit calculation, punishment adds to aggressor costs and denial reduces aggressor benefits making the aggressor believe that its aims will not be achieved (Mallory, 2018, p.3).
Following the end of the Cold War in 1989 and a reduction in the threat of major conflict, the discipline of Security Studies went through a process of change. Whereas security had previously been studied with a focus on interstate conflict, some chose to expand the field of study to encompass ‘non-traditional’ threats to security. Depending upon one’s stance within the discipline, threats to security are not always seen as being related to state use or threatened use of military force, though some retain that narrow, traditional view. Others see ‘non-traditional’ threats such as disease, famine, environmental change, cyber-conflict, organised crime and terrorism as security issues which should sit within the ambit of Security Studies 1 (for useful overviews of the evolution of Security Studies see: Balzacq and Cavelty,2017; Collins,2022; Hough,2018; Malik, 2021).
In the immediate post-Cold War period, the reduced threat of major conflict caused some to question whether there was a future for the study of deterrence within the field of Security Studies. The attacks on the USA on 11th September, 2001 and the subsequent ‘War on Terror’ provided an answer. There was a marked increase in research activity linked to the deterrence of terrorism.
Aims of the Paper
In 2004, Freedman, commenting upon debates on deterrence within the fields of Security Studies and criminology, noted ‘… how little they draw upon each others’ work’ (Freedman, 2004, p. 60). Whether consciously or not, some have responded to Freedman’s observation and sought to meld findings from the study of deterrence across both disciplines (see e.g. Stein and Levi, 2014; Rascoff, 2014; Rid, 2012). In a statement of particular relevance to this paper, Braithwaite (2005) noted that the concept of deterrence in the realm of security is ‘…more dynamic compared to the static models that dominate criminology (p. 98)’.
This paper is rooted in two categories of deterrence study from across the disciplines of Security Studies and criminology: 1. The deterrence of terrorism, increasingly studied following the 9/11 attacks. A rich terminology has been developed to describe the theory base of terrorism deterrence. 2. The use of Focused Deterrence as a crime reduction measure. Whilst studies of Focused Deterrence have found it to be effective, there is a lack of theorising the mechanisms in play.
The paper seeks to take the terminology that has developed in Security Studies around deterrence in general and around measures intended to control terrorism in particular and to put that lexicon to use in developing a descriptive framework suitable for application to Focused Deterrence. It is hoped that the descriptive framework advanced will help interested parties hypothesise and test potential causal mechanisms thereby improving the use of Focused Deterrence. It is not suggested that this paper will deliver the ultimate framework for discussing, devising, assessing and evaluating focused approaches to reducing crime. It is a conceptual piece attempting to establish a mindset. This paper is intended to light some lamps on a path that others might take, further illuminate, widen and improve.
The paper has four parts. The first briefly reviews the literature relating to Focused Deterrence and the deterrence of terrorism. The second sets out the language employed in approaches to reducing terrorism and considers its potential for use in relation to Focused Deterrence. The third uses this language to describe activities within the Focused Deterrence model. The fourth consists of a discussion with potential implications for practice.
Literature
To identify key references relating to Focused Deterrence and the deterrence of terrorism, the authors utilised ‘snowballing’ and ‘reverse snowballing’ techniques (Sayers, 2007). Simple internet searches were conducted using the keywords ‘Focused Deterrence’ and ‘Deterrence Terrorism’. Prominent articles/books were accessed and works cited in those items were then accessed and the process repeated in a ‘snowballing’ approach. The product of this approach was then assessed for relevance to the task in hand and analysed. As ‘snowballing’ only reveals material pre-dating the articles used to initiate the ‘snowball’, a ‘reverse snowballing’ process was then undertaken by the authors to find more recent material. The citation tracking facilities in online databases such as ISI Web of Knowledge and Google Scholar allow a researcher to identify more contemporary materials that have cited prominent articles/books revealed by the initial ‘snowballing’ exercise. For example, checking the citation history of a highly influential article published in say 2000 will reveal any articles citing that work since 2000 up to the present day. The results of the ‘reverse snowballing’ exercise were accessed by the authors, assessed for relevance and the product analysed.
Focused Deterrence
Along with debates about the effects of certainty, celerity and severity of punishments, the concepts of general and specific deterrence have engaged criminologists for many years.
A more recent addition to deterrence studies in criminology is Focused Deterrence. Focused Deterrence was first deployed in Boston, USA, in the mid-1990s and was operationalised in that city under the name of ‘Operation Ceasefire’. It was a response to the horrifically high levels of firearm-enabled injuries and deaths amongst young men. Focused Deterrence is a problem-oriented approach (Goldstein, 1979, 1990; Kennedy, 2019). As with other problem-oriented approaches, Focused Deterrence strategies are crafted following identification and detailed analysis of the presenting problem. Analyses in Boston prior to the first use of Focused Deterrence revealed, amongst other things, that the problem was highly concentrated both geographically and demographically. For young male members of the gangs identified in Boston at the time, there was a one in seven chance of dying from gun violence over a 9 year period (Kennedy 2011).
The Focused Deterrence approach was also referred to as a ‘pulling levers’ strategy. This term was used to describe a strategy where all ‘levers’ available to the police and partner agencies were ‘pulled’ in attempts to change offending behaviour. Police, partner agencies and community members told gang members that support in the form of offers of education, training and employment would be given to those who wanted to change their life-styles. However, gang members were also told that every ‘lever’ would be ‘pulled’ to target and disrupt those gangs that continued to use firearms. Moreover, rather than targeting only those doing the shooting, the attack would be on all members of the group. To do this, there would be punishment for all offences committed and al rules broken, including those normally ignored as minor transgressions such as public consumption of alcohol (Kennedy, 1997, 2009; Kennedy et al., 1996). Gang members were particularly vulnerable to the Focused Deterrence approach due to their typical ‘cafeteria’ style of offending, a term which recognises the versatility of offending displayed by gang members who often exploit the menu of offending opportunities presented to them (Klein, 1995, p. 68). There was no licence to commit other types of offences for offenders who eschewed gun violence. Such offending received a traditional enforcement approach. However, for those linked to gun violence the threatened special response was consistently applied.
The Focused Deterrence strategy appeared to be stunningly successful (Kennedy, 1997; Kennedy et al., 1996, 1997, 2001). In the period immediately prior to the use of Focused Deterrence, Boston suffered approximately 100 homicides per year. Operation Ceasefire was implemented in 1996. In 1999, 31 homicides were recorded, an impressive reduction (Kennedy, 2011, p.75). The Boston strategy has been replicated in numerous locations. By 2018, the approach had been implemented in 84 U.S. cities and other countries including Scotland, Turkey and Brazil. (Braga et al. 2018). Though best known for its use against firearm-enabled offending involving groups, the Focused Deterrence approach has also been used in relation to other crime types such as drug-dealing (Saunders et al., 2016) and domestic violence (Sechrist and Weil, 2018).
Braga and Weisburd performed a systematic review of the effects of Focused Deterrence strategies upon crime and reported ‘The basic findings of our review are very positive’ (Braga and Weisburd, 2012, p. 347). However, in common with other commentators on the subject of Focused Deterrence (e.g. Corsaro et al., 2013) the authors highlighted the importance of identifying potential mechanisms at play that could contribute to observed effects (pp. 349–351). Further calls for studies of theoretical mechanisms underpinning Focused Deterrence were made by the same authors in a later study (Braga and Weisburd, 2014). A further systematic review by Braga et al. in 2018, whilst finding that Focused Deterrence appeared to be effective also noted the lack of understanding of the operant mechanisms in successful initiatives The authors stated, ‘Nearly all of the Focused Deterrence program evaluations included in this review could be described as “black box” evaluations where it is uncertain which program elements were most important in generating observed crime reduction effects’ (Braga et al., 2018, pp. 35–36). The need to address the ‘black box’ of Focused Deterrence has been echoed by other authors. (e.g. Schnobrich-Davis and Gardner, 2018; Sullivan and Lugo, 2018; Trinkner, 2019). In short, evaluators were of the opinion that Focused Deterrence is effective, but nobody knows why due to a failure to identify causal mechanisms. The authors of this paper were reminded of a saying attributed to various sources: ‘It works in practice, but does it work in theory?’
Deterrence of terrorism
During the Cold War, nuclear deterrence was a binary situation. Deterrence either worked or it did not and if it had failed the failure would have been all-too obvious. In considering potential responses to terrorism, a major question faced by policymakers and researchers alike lies in whether terrorist acts can be deterred, given the fanaticism of some of the actors. Most experts in the field believe that they can, though acceptable results are more likely to represent marginal gains. Imperfect deterrence is better than no deterrence at all (see Lieberman, 2020; Trager and Zagorcheva, 2005; Wenger and Wilner, 2012; for useful overviews). In this respect, reducing the threat posed by terrorism is more akin to the reduction of crime where a world with less crime is preferable to a world with more crime, but the inevitability of some level of crime is accepted (Rid, 2012; Vinson, 2015).
The nature of measures required to reduce terrorist incidents will differ from those used to manage Cold War-era threats. Recognising that fanatical actors are likely to be difficult to deter directly, some have highlighted the importance of disaggregating the components of terrorist groups and their supporters to identify players who may be more ‘deterrable’ (Davis and Jenkins, 2002). Abrahms makes the point: ‘Terrorism may be deterrable even though most of its practitioners are not’ (2014, p.5). By way of example, finding ways to deter outwardly respectable figures from providing covert financial or other logistical assistance to terrorist groups could lead to a reduction in incidents through starving fanatics of support.
The Concept of Preventive Influence
During the post-Cold War period, the potential for broadening the concept of deterrence in Security Studies has been subject of debate, much of which has been informed by the emergence of the new threats previously described. The debate is neatly summarised by Mazarr et al. (2018). The narrowest concept holds that military means are the way to achieve deterrence. A broader concept accepts that means beyond the military may be necessary to deter an adversary. Both the narrow and broader concepts are based in threats that aggression will result in punishment and/or denial of the benefits sought. A third and even broader approach to deterrence allows for combining threats with positive inducements to forestall aggression. In relation to this much-broadened approach, Mazarr et al. acknowledge that ‘… the threat-plus-inducement understanding of deterrence can obscure the boundaries of the concept’ (Mazarr et al. 2018, p. 4).
Arguments in favour of mixing inducements and threats to affect an adversary’s cost-benefit calculation are attractive; whether such an approach should be viewed as deterrence is not. Writing some years before Mazarr et al. (2018), others raised concerns about the conceptual boundaries of deterrence. Davis and Jenkins in seeking to develop potential responses to terrorism found the concept of deterrence to be ‘…too narrow to use as an organising principle…’ (2002, p. 9). Recognising the potential value of blending inducements with threats they promoted the use of influence measures to reduce the threat from terrorism. In this, they shied away from stretching the concept of deterrence beyond the commonly accepted understanding that it involves the threat of negative consequences. In similar vein, Jeffrey Knopf writing later held that deterrence ‘…arises mainly because an actor expects a course of action to lead to a negative outcome’ (Knopf, 2009, pp. 38–39). In asserting this, Knopf drew upon the everyday use of language, believing that nearly every lay person could define deterrence and that those definitions would be likely to revolve around threats and not positive inducements (p. 34). In common with Davis and Jenkins (2002), Knopf chose to view deterrence as ‘…a form of influence…’ (2009, p. 34) within an overarching concept of ‘preventive influence’. Preventive influence ‘… might also be pursued through the promise of rewards or offer of reassurances’, but for Knopf, such measures represented positive inducements as opposed to deterrent measures (p. 38). Davis, Jenkins and Knopf all took the view that deterrence is but one part of a range of possible measures to exert influence, a range which includes dialogue and inducements as well as threats. But what is influence? Mazarr et al. (2021) found the term to be ‘…ill-defined and often employed in vague and imprecise ways’, where ‘Many definitions very explicitly note that influence involves achieving effects without force and often indirectly’. Other definitions are silent on whether force may be used to exert influence (p. 12). Though widely used in everyday language, researchers have tussled with the seemingly simple term (see e.g. Santa Maria, 2013).
The authors of this paper offer a way out of the definitional thicket surrounding the term ‘influence’ by asserting that force
A further example of the use of limited force with a view to influencing the decision making of adversaries is the concept of Cumulative Deterrence. Cumulative Deterrence may be defined as ‘…the simultaneous use of threats and … force over the course of an extended conflict’ (Almog, 2004, p.8), where the intention is to produce a deterrent effect in the future due to responding consistently as previously threatened. Consistent responses with attendant messaging that force has been used as a response to provocation and will only occur if certain lines are crossed may accumulate to produce cumulative deterrence over time. The distinction between compellence and cumulative deterrence is fine but still exists. Limited force that influences an adversary to stop doing something at the time is compellence. Cumulative deterrence is the result of the repeated use of limited force over time and building a reputation for consistently following through on threats. Both compellence and cumulative deterrence use force not to remove freedom of action but to target decision making in an adversary. Under the argument advanced to this point, they are influence measures.
Reflections after consideration of the focused deterrence and Security Studies literature
The term ‘Focused Deterrence’ does not accurately describe the approach first used in Boston in the mid-1990s to reduce gun violence. This has been acknowledged by some of the earliest champions of the approach (Kennedy et al., 2017, p. 160, p. 160)
But what of the ‘brand’? Focused Deterrence might be a misnomer, but it is an established misnomer as evidenced by the wide use of the term by both practitioners and researchers. However, there is a solution which both preserves the ‘brand’ and avoids convoluted approaches such as ‘the approach formerly known as Focused Deterrence’. From its inception, Focused Deterrence has also been referred to as a ‘Pulling Levers’ approach (Kennedy, 1997, 2006) though to date use of the term has been eclipsed by the more popular ‘Focused Deterrence’. ‘Pulling Levers’ as a term perfectly captures the spirit of ‘preventive influence’. It could provide a more accurate description of the approach than the term ‘Focused Deterrence’ and also act as catchy shorthand for those seeking to deploy multi-faceted influence strategies. The authors suggest that ‘Pulling Levers’ becomes the preferred term in common parlance when referring to what are in fact preventive influence approaches. Fortunately, the term ‘Pulling Levers’ already has some currency which may be amplified. The term ‘Focused Deterrence’ should be allowed to quietly slip away. As a start, from this point ‘Pulling Levers’ will be used in the rest of this paper to refer to the ‘Pulling Levers/Preventive Influence’ approach ‘formerly known as Focused Deterrence’.
Improving understanding and use of the Pulling Levers approach
The terminology of Preventive Influence
Effective use of pulling levers strategies to reduce crime and identify causal mechanisms for any effects observed requires access to a richer terminology than is currently used in criminology. Aims, methods and potential mechanisms of change to be activated in relation to any intervention must be accurately described. Tests of efficacy for hypothesised mechanisms must be devised if mechanistic explanation of pulling levers approaches is to be achieved within an assessment and evaluation regime. 2 It was stated earlier in this paper that terminology developed within the discipline of Security Studies in relation to reducing terrorism could be exploited with a view to achieving these ends.
Figures 1 and 2 draw upon relevant Security Studies literature and represent efforts to summarise terminology relevant to the concept of preventive influence. Figure 1 consists of a lexicon of methods for exerting influence. Figure 2 is a lexicon of deterrence. Conceptually, Figure 2 is an expansion of the deterrence section of Figure 1. Whilst deterrence is shown as only one tool within the influence toolkit (Figure 1), it has many variants which are briefly ‘unpacked’ in Figure 2. It is not contended that the terminology presented by way of the two lexicons is exhaustive. It is contended that the two interlinked lexicons represent a step towards developing an analytical framework to provide support to those seeking to identify, activate, assess and evaluate causal mechanisms to improve the use of preventive influence, in the guise of Pulling Levers, in the field of crime reduction. A lexicon of preventive influence. A lexicon of deterrence.
Figure 1 has been devised drawing upon the work of several authors (Davis, 2012a, 2012b; Davis and Jenkins 2002; Davis et al. 2021; Knopf, 2009; Schelling, 2008). The measures within Figure 1 should not be viewed as operating in isolation from each other. The intention of a well-formulated strategy of influence would be to blend measures as required to achieve the desired result.
As can be seen from Figure 1, deterrence is one way to exert preventive influence. Deterrence, however, comes in many forms. Figure 2 consists of a lexicon drawn from the deterrence literature in Security Studies, with an emphasis on the literature surrounding approaches to deterring terrorism. Some of the terms within the lexicon are listed in ‘family’ groups (A to I) to highlight where similar deterrent mechanisms might be intended to be activated even though different terms may be used to describe the deterrence type. Illustrative references are provided within the table. These references by no means represent the sum total of the scholarship surrounding the items listed. They merely seek to raise awareness in the reader of the types of deterrence subject of discussion in the Security Studies realm.
Describing the ‘Pulling Levers’ approach using the terminology advanced
The core ‘Pulling Levers’ model is summarised below (Kennedy, 1997, 2006, 2009, 2019; Kennedy et al., 1996). Though having been used against other crime problems as mentioned above (e.g. open-air drug markets and domestic violence), the ‘Pulling Levers’ approach in criminology is best known as a response to gun crime. In recognition of this, commentary has been added to stages of the model to illustrate how the terminology advanced earlier can be used to accurately describe measures intended to reduce firearm-enabled violence. Where a deterrence type can be described by more than one term, only the first term in each ‘family’ group within Figure 2 is used.
The core ‘Pulling Levers’ model as used against gun violence
(a) Develop understanding of the problem in the local context by way of a multi-agency partnership.
During this stage of the model, contextual intelligence is gathered to develop Tailored Deterrence approaches based upon identification of contextual ‘centres of gravity’ where the application of pressure is likely to generate behavioural change. For example, if violence is linked to the local criminal economy, that economy may be viewed as a point of focus for deployment of influence measures to change behaviour. (b) Use communities, victims and those close to potential offenders to describe the harms caused by the problem behaviour and to communicate that things will change.
Prior warning of introduction of the new regime represents attempts at Dissuasion of potential offenders. Engagement with key groups within affected communities seeks to undermine any belief amongst offenders that their behaviour is acceptable to the community. This seeks to establish Deterrence via Delegitimisation. (c) Make offers of support in relation to employment, training, housing etc., to those involved in the proscribed activity should they wish to change their ways.
Offers to provide support to those directly involved in proscribed behaviour may be viewed as Inducements (providing rewards for behavioural change) and/or Co-Option (bringing an offender into the law-abiding community). The existence of schemes and arguments for taking up offers of support may convince those on the periphery to avoid becoming involved in offending, thereby representing a successful use of Persuasion. The existence of such offers supports the concepts of Deterrence via Delegitimisation.The ground is taken from below the feet of the offender who justifies their behaviour by stating ‘I do this because there is nothing else for me’. (d) Directly and repeatedly communicate with the identified players that a failure in them to change their behaviour will attract a premium multi-agency enforcement response. Those giving up the prohibited behaviour will attract a standard enforcement response for any other types of offending behaviour.
Such communications seek to influence offender behaviour by way of Deterrence by Punishment. In addition, through differentiating responses between types of offending in a harm-reduction approach, such approaches seek to achieve Narrow Deterrence. Some types of offending might still occur and be addressed by traditional responses, but proscribed behaviours will not be tolerated and will receive a premium enforcement response. The long-term goal is the establishment of Deterrence via Norms, whereby the proscribed behaviour is not even considered as an option by offenders. (e) Punish any instances where the prohibited behaviour is exhibited through taking action for any offences committed or rules broken, exploiting the versatility of offending displayed by those involved.
This element of the model relies upon Cross-Domain Deterrence. It recognises that whilst it may not be possible to prosecute offenders for committing the offences within the proscribed offending ‘domain’ (i.e. gun crimes), behaviour in other offending ‘domains’ that are easier to detect makes offenders vulnerable to influence via law enforcement. Action may be taken for drug offences, parole violations, motoring offences etc. If violence is being pursued in relation to drug supply, disruption of that market could be undertaken to deny the violent the financial benefits they seek: Deterrence by Denial.
Swift and consistent application of the threatened punishments for every transgression represents Compellence, which could be expected to contribute to the generation of Cumulative Deterrence over time. (f) Target punishment action on those posing the greatest risk and offenders associated with them, with a view to coercing offenders into controlling the behaviour of their more problematic associates.
Threatening to punish key players is classic Deterrence by Punishment. Extending that threat to those associated with the most dangerous offenders invokes Indirect Deterrence: applying deterrent pressures to those who might not pose a direct threat but who can be coerced into applying deterrent pressure to those who do pose a threat.
Discussion and potential implications for practice
There are several advantages in describing ‘Pulling Levers’ activity using the terminology as demonstrated above: 1. In discussions, all interested parties speak a common language. 2. The terminology can stand as a checklist of measures to be considered when practitioners are considering what to do in a situation. 3. Every measure invoked using the terminology (as highlighted in bold in the example above) is a putative mechanism. The identification of each putative mechanism should naturally generate the question ‘How will you know if it is working/has worked?’. The terminology provides a means to clearly articulate which mechanisms it is hoped will be triggered to influence behaviour and allows tests for the efficacy of each mechanism to be devised. Devising tests of efficacy for each identified putative mechanism goes some way towards tackling the mechanistic explanation issue faced by the ‘Pulling Levers’ approach.
The quest for mechanistic explanation of ‘Pulling Levers’ has required what Davis called an increase in ‘… the battle space’ to go beyond deterrence and capture the wider concept of influence (2012a, p. 69). The inaccurately named Focused Deterrence is more than deterrent approaches. It consists of pulling numerous levers, from co-option to compellence within a wide preventive influence approach. Agreeing with others’ observations that Focused Deterrence is more than deterrence (Kennedy et al. 2017; Braga and Kennedy, 2021), the authors advocate the use of the terms ‘Pulling Levers’/preventive influence strategies to more accurately describe the activities undertaken.
The ‘Pulling Levers’ concept is a way of thinking, and its use requires the adoption of a creative mindset. The concept is not prescriptive in the use of methods. It merely requires that the problem to be addressed is carefully considered and that any response should look to maximising the influence exerted by way of playing to the intervener’s strengths and targeting the adversary’s weaknesses. The relevant ‘levers’ may be found along a spectrum ranging from softer measures like persuasion and inducement to harder options involving threats.
It is one thing to adopt a ‘Pulling Levers’ mindset, but its practical use lies in establishing a common language to allow researchers and practitioners to discuss, devise, implement and review ‘Pulling Levers’ strategies. As Socrates is alleged to have said, ‘The beginning of wisdom is the definition of terms’. Desired end states can be described along with the levers to be pulled to achieve those end states. Detailed descriptions of the influence process and the disaggregation of its elements allow for identification of hypothesised mechanisms to be tested for their individual contribution, or lack of it, to observed outcomes.
The formulation of a ‘Pulling Levers’ strategy consists of answering a number of key questions: 1. What is the behaviour to be addressed? 2. What would be viewed as an acceptable outcome of any intervention? 3. What is the context surrounding the behaviour? 4. Within the context, what ‘levers’ of preventive influence can be identified for use, drawing upon the terminology advanced? 5. What mechanisms of effect may be hypothesised and clearly articulated for those influence measures identified? 6. How can progress towards the desired outcome be measured and the hypothesised mechanisms tested? (Ongoing assessment) 7. What lessons for the future may be learned? (Final evaluation)
The use of ‘Pulling Levers’ strategies to reduce gun violence, drug supply and domestic violence has already been mentioned. However, there is no reason why the framework and terminology advanced to this point could not be used to good effect to address other crime problems. For example, ‘County Lines’ drug-supply groups in the UK use vulnerable adults and children to transport illegal drugs from metropolitan areas to rural locations where the drugs are then sold. Preventing all drug supply between such locations would be impossible while demand exists. However, a harm-reduction approach could be taken to reduce exploitation of the young and vulnerable using the ‘Pulling Levers’ mindset. Communicating to crime groups that the use of children and/or vulnerable adults will lead to focused enforcement activity could produce good results. Profits could also be threatened by communicating to customers of targeted groups that they are complicit in exploitation and they too will receive extra police attention. The key to success would be to make drug-supply groups compete to avoid attention from the ‘Pulling Levers’ approach. The way for them to do that would be to stop exploiting children and the vulnerable.
Conclusion
Consideration of the literature surrounding the control of terrorism has provided the basis of a descriptive framework capable of being used as an aid to discuss, devise, implement, assess and evaluate ‘Pulling Levers’ strategies. Deterrence is the use of threats to prevent the occurrence of undesirable activity. However, prevention of such actions may be achieved through the use of other forms of influence, and the term ‘preventive influence’ is a useful overarching term for such a mindset. ‘Pulling Levers’ is a catchy, shorthand term, which captures the essence of the concept and already has some currency. It is hoped that the language of preventive influence developed in this paper will aid future efforts to understand and employ ‘Pulling Levers’ strategies. If practitioners do no more than use the terminology developed within this paper as a framework to aid considerations around developing, implementing and reviewing future ‘Pulling Levers’ interventions, the authors will think that their time has been well spent.