Introduction

Since the 1990s, demobilisation, disarmament, and re-integration (DDR) programmes have emerged as central components of post-conflict processes, to the extent that they have become ‘a sine qua non of contemporary peace support operations and state-building’ (Colletta and Muggah 2009: 427). At their most basic, DDR programmes aim to re-introduce former combatants into civilian life following peace settlements, in order to ‘reduce the prospects for war recurrence, reduc[e] military expenditures and re-assert… the state’s monopoly over the legitimate means of coercion’, ‘neutralis[e] spoilers, break… command and control of previously armed factions and promot[e] sustainable livelihoods’ (Colletta and Muggah 2009: 428). Scores of DDR programmes have been implemented around the world since 1989, but as Robert Muggah (2010: 4) has pointed out, their balance sheet is extremely mixed, with few – if any – unmitigated success stories. He identifies two major reasons for this state of affairs: on the one hand, the fact that DDR programmes are ‘erroneously conceived as a kind of magic bullet that automatically and simultaneously cares for a wide range of development and security challenges’, and on the other, that they tend to consider reducing violence as a relatively straightforward ‘technical’ issue.

Leaving aside for now the thorny question of the relationship between security and development (see Bates 2001; Jones and Rodgers 2016; Jensen 2010), and focusing instead on the violence-reducing function of DDR programmes, this can generally be said to be underpinned by certain basic assumptions. These include, in particular, the notion that ‘former combatants face unique and specific challenges’ – in that they have a lack of usable skills and may be suffering from trauma – and that ‘targeted reintegration programmes are necessary’, which need to be specifically adapted to permit beneficiaries to overcome barriers such as stigma and discrimination linked to their combatant role (Barron 2009: 255). As a result, there exists a widespread perception that DDR programmes are ‘usually straight-forward’, and essentially involve addressing ‘a checklist of issues’ that although variable in detail, will work across different contexts (Clark 1996: i). To a certain extent, this vision of things is due to the fact that DDR programmes are a form of emergency intervention, for which having established templates is a practical necessity. However, as Jensen and Stepputat (2014) and Munive (2014) have pointed out, it is also a consequence of the academic and policy literature about DDR programmes being extremely parochial. Certainly, the reference points for most debates about any given DDR experience tend to be other DDR programmes, something that inevitably gives rise to a doctrinal ‘pensée unique’.

Indeed, it is striking how rarely comparisons with other types of violence-reduction initiatives are made within this body of literature when one considers that DDR programmes often have much in common with them. Gang interventions are a case in point in this respect, with the current ‘global war on gangs’ very much underpinned by similar basic assumptions to DDR programmes about template interventions or the nature of those being targeted, for example. At the same time, however, the logic of these underlying suppositions is often laid bare much more visibly in the context of gang interventions compared to DDR programmes due to the hotly disputed and heterodox nature of debates concerning the former. There is, furthermore, also scope for DDR programmes to learn not only from the failures of various wars on gangs, but also from gang research more broadly, with investigations into the reasons that push individuals to leave gangs being particularly useful to consider. Many of these more organic processes often challenge the basic assumptions underpinning both gang interventions and DDR programmes, but in doing so arguably also provide concrete elements for potentially re-thinking both, similar to the trajectories of gang members after they leave the gang.

Drawing on long-term ethnographic research on gangs in Nicaragua and South Africa, this article aims to explore what DDR programmes might learn from research on gangs and the problems associated with government-instituted so-called ‘wars on gangs’, putatively aimed at reducing or fighting gang-related violence. It begins by considering interventions associated with the global war on gangs, comparing their underlying premises and practices with those of DDR programmes, exploring their similarities, differences and ultimate failures. It then moves on to explore why individuals leave gangs, focusing on the more organic processes that deplete gangs of their members, and particularly on cases where they do so of their own volition rather than as a consequence of an external intervention – as globally this constitutes the majority trajectory out of a gang (see Rodgers and Hazen 2014). The article considers the insights that research on this topic might have for DDR programmes, more specifically in relation to different possible occupational trajectories of ex-gangs, and whether they increase or decrease the likelihood of violence. The conclusion then aims to distil some of the potential lessons that gang research can offer DDR programmes. While we do not offer explicit recommendations or elicit specific lessons, our analysis suggests several possible alternative avenues and more sustainable ways of thinking about violence reduction for DDR programmes. These include focusing on the management of violence rather than trying to do away with it altogether.

Comparing DDR Programmes and the Global War on Gangs – Lessons in Failure

To compare DDR and so-called ‘wars on gangs’ might at first appear slightly counter-intuitive. Certainly, there are without question significant differences between ex-combatants and gangsters. For example, with regard to motivation, ex-combatants often fight for an ideological cause, while gangsters are generally thought to be economically-driven. As we have argued elsewhere (Rodgers and Jensen 2009), however, these differences are not necessarily clear-cut, and must be ascertained empirically rather than simply assumed. Moreover, even if there are differences between ex-combatants and gangsters, the policy interventions that they generate often exhibit significant similarities. In particular, there are at least four elements that warrant direct comparison: (1) the ambition of interventions; (2) the preponderance of templates; (3) certain central assumptions, and (4) the necessity of outside action. Firstly, in terms of their ambition, both forms of interventions are crucially concerned with ‘violence reduction’. They seek to reduce violence in situations characterised by high levels of conflict through a targeted intervention aimed at the ‘perpetrators’ of violence – combatants or gangsters – who are depicted as a clearly identifiable group generally distinguished from ‘victims’. Secondly, DDR projects and wars on gangs tend to rely heavily on templates. Thirdly, and also similarly to the war on gangs, DDR programmes are based on a number of key binary assumptions about what the conflict entails – who the soldiers are and what the community into which they are to be integrated looks like. Finally, both DDR programmes and wars on gangs involve external agents at the centre of the violence reduction – international organisations in the case of the former, and the police or the army in the case of the latter (while the police and army are generally not external to particular national contexts, those deployed in relation to gangs are frequently not local, and therefore do not have an inside perspective which makes them highly analogous to international organisations).

Jensen and Stepputat (2014) deconstruct each of the assumptions in relation to DDR programmes, concluding that the challenge for these programmes is to transcend institutional and legal categories of ‘victims’ and ‘perpetrators’, and to look at ‘practices’ on the ground instead. They argue that questions about who is involved in violence, how, and for what reasons, should all take centre stage, rather than the preconceived assumptions concerning bounded categorisations to be acted upon on the basis of blueprints and templates. A similar challenge clearly exists in relation to the wars on gangs. However, one of the obvious differences between DDR programmes and the wars on gangs is that the latter tend to take place not only in developing but also developed countries. In this regard, wars on gangs are politically much more important to more than just the international development community. Violence in the form of gangs matters to national political constituencies across the world in a way that DDR in Sierra Leone does not matter for the majority, for example.

Crises centring on gangs and gang violence are therefore a constant element of public discourse in many countries, even if concerns about them arguably only develop episodically in relation to particular shocking events, such as street shootings or spectacular criminality. This results in immediate calls for urgent action. Thus, rather than discussing the emergence of templates, we should arguably be talking about the ‘emergency of templates’ in the context of wars on gangs. A crisis – for instance the outbreak of gang war – will lead to politicians and others immediately demanding action in order to contain it. To do so quickly, state institutions turn to tried and tested methods. The recent so-called ‘gang wars’ in Denmark provide an apt illustration. When gangs began shooting in 2009, media, experts and working groups within the municipality and the police began to look for ‘answers’ as to how to stop the gang war.1 Anti-gang initiatives deployed in Manchester were adopted as a model to follow.2 During a subsequent gang-related conflict, a different solution was found in Holland, which involved removing hash markets through legalization.3 The next time a gang-related event occurs, another solution will undoubtedly be found elsewhere. What this highlights, however, is that templates to reduce gang wars often proceed in spurts, galvanized by an emergency, after which there will be a period of laxity that will last until the next crisis (see Johansen and Jensen n.d.). In this respect, DDR programmes are often better planned, and build on a less disparate set of ‘best practices’. Certainly, the extent to which the specific nature of the different templates really matters in the context of the wars on gangs is open to debate, particularly compared to DDR programmes.

However, the nature of the wars on gangs also depends on where in the world they occur.4 Responses will vary in terms of the resources governments have at their disposal, and the responses will often be varied. Having said this, at the core of wars on gangs is a particular trope that can be traced to anti-gang initiatives which emerged in the US during the 1980s that partly featured a state withdrawal from the country’s inner cities, and an increase in police authoritarianism (Wacquant 2008; Venkatesh 2000). This was the so-called ‘Zero Tolerance’ policing model, based on the notion that smaller crimes lead to bigger crimes. Thus, to prevent bigger crimes it is necessary to target the smaller ones. Crime and violence were viewed as a choice, made by the individual perpetrator. Government responses therefore had to act as a deterrent to that choice (Jensen 2010).5 The solution inevitably consisted of imposing harsher and more mandatory sentences, criminalizing membership of gangs, curbing their economic activities, and introducing specialized gang courts and gang policing units. Viewing crime as a choice effectively leads to the emergence of what criminologist André Standing (2006) denotes as a ‘parasitic’ understanding of gangs. He argues that gangs come to be seen as distinct and isolated entities whose main objective is committing crime. They therefore ‘infect’ the economy and undermine democracy, and are considered the equivalent of social ‘tumours’, which must be ‘removed’ in order to heal an otherwise healthy society.

While this particular reading of gangs lead politicians, policy people and practitioners towards endorsing and passing punitive measures, the parasitic model also explains the underlying logic of many of the ‘softer’ non-punitive measures that have also been implemented to reduce gang violence. Many of these focus on job training, income generation and other economic activities. These interventions are supposed to ‘tempt’ gangsters away from the economic opportunities offered by gang membership, rather than scare them from the ‘costs’ of gang-associated economic activities (if caught). While we are clearly more sympathetic towards such socio-economic approaches, they only work using rational choice assumptions about gangs as described by Standing (2006). As argued elsewhere (Jensen 2008), such assumptions need to be questioned. Certainly, the notion that criminality – to say nothing of life more generally – occurs on a rational basis needs to be considered with a certain scepticism.6 Furthermore, the punitive approaches are arguably still far more important in the everyday lives of people than project-based socio-economic initiatives.

While the US has paved the way in these matters, the rest of the world has followed its lead. Perhaps the most (in)famous contemporary example is the so-called ‘Mano Dura’ (‘Hard Hand’) policing campaigns that were introduced in Central America in the early 2000s (see Rodgers 2009). However, the Zero Tolerance trope can be seen everywhere. In South Africa, for example, a variety of steps have been taken over the past decade to institutionalise such an approach. In 1998, the Prevention of Organised Crime Act was passed, which targeted gangs through harsher sentencing for organized crime, through asset forfeiture (targeting the ‘life blood’ of the criminals), and through criminalizing association to ‘criminal gangs’ (sic). To assist in the identification of a gang, the South African legislation drew on a gang definition that was taken, almost word-for-word, from a Californian law – the Step Act of 1988 – which stated that the magistrate might consider a suspect a member of a gang if that person fulfilled certain criteria of residence, dress, association, and reputation. Along with criminalization of the gangs, the South African state also passed a set of mandatory minimum sentences for particular crimes. The introduction of minimum sentences is a prime example of how the power of the judiciary in South Africa has been reduced, as it can no longer deliberate over the sentencing of particular cases. This is largely due to the neutrality and independence of the courts, which has increasingly come to be seen as the problem. Finally, during the late 1990s and early 2000s, the police and the political establishment created several large-scale special policing operations – Operation Recoil, Operation Good Hope and the ‘High Flyer’ operation, to name but a few. These operations were aimed specifically at gang violence. Huge resources and considerable media attention were dedicated to these operations, which were both an attempt to focus attention on a specific problem and a way to communicate a concern about crime. As often as not, such operations met with failure.

The assumptions on which the wars on gangs are fought are inherently problematic. As over a century of gang research has highlighted, gangs are not the distinct, socially disembedded entities that the ‘parasitic’ model suggests they are (see Rodgers and Baird 2015, for a discussion of this in relation to Latin American gangs). Moreover, their purpose is by no means necessarily economic, but can involve a range of non-criminal elements including social or political concerns, such as promoting a sense of identity or vigilante order. Perhaps most critically, however, one of the striking features of gangs worldwide is that there universally exists what Scott Decker and David Pyrooz (2011: 16) have called a ‘natural desistance process’, that is to say, gang members more often than not naturally leave gangs after a variable period of time. This has obvious importance for gang interventions, especially considering the way that these can often increase rather than decrease violence (see Aguilar 2006). It might actually be better to simply ‘do nothing’ and wait for gangs to fall apart and run out of steam. Certainly, an exploration of the motivations that gang members have for leaving their gang highlights that this generally has little to do with external intervention, and is rarely spectacular, as the next section explores.

Leaving the Gang: Organic Gang Desistance

There is much to be learnt from understanding what happens to individuals after they leave gangs, especially regarding DDR projects. This is what we term ‘organic’ processes of gang desistance. We contrast such processes to the willed, external interventions of both DDR programmes and the war on gangs. As Malcolm Klein and Cheryl Maxson (2006: 154) have pointed out, ‘surprisingly little research has been conducted on gang desistance’. The few studies considering this process highlight that it can occur for a range of reasons, and in several different ways. For example, in their study of 84 former gang members in Phoenix, Arizona, Pyrooz and Decker (2011: 420) distinguish between ‘push’ and ‘pull’ motives for leaving a gang:

Pull motives were characterized by changing social controls or turning point factors that fracture the “grip of the group”… Responses that included girlfriends, jobs, or children as the motivation for leaving the gang were recorded as pulls because they are external to the gang, acting as “hooks” to restructure the lifestyle of gang members. Push motives were characterized by cognitive shifts or transformations about gang life. Responses that included “I got tired of the gang lifestyle” or “I wanted to avoid trouble and violence” were recorded as pushes because they are internal to the gang, inspiring former gang members to seek out and select into other social arenas.

The study found that two-thirds of former gang members left due to push rather than pull motives, and that leaving the gang was generally a drawn-out process rather than a sudden event.

These insights are largely confirmed by our own research. Three individual gang member case studies drawn from long-term fieldwork on gangs in Cape Town, in South Africa, illustrate the desistance process in more ethnographic detail. The first concerns Gerard, who engaged in what we might term ‘desistance through girlfriend’:

  • Gerard had joined the New Yorker gang because they dominated his part of the township. The gang had developed from a local self-defence group into an increasingly aggressive group, using more violence and then engaging in drug trafficking as low-level neighbourhood pushers. As time passed, Gerard was less and less comfortable with his role as a gang member, and the expectation of violence on his part, and he came up with frequent excuses for not attending and participating in the gang life. He slowly became less involved. This caused tensions, but he was able to point to having started ‘going steady’ with his girlfriend as a justification. Though the gang accepted this, he had to reach a compromise with them because he continued to live in New Yorker territory. As a result, his house was used as storage for guns and drugs, and occasionally he had to fight with the New Yorkers.

Shaun’s story illustrates a different exit narrative, one which we might call ‘desistance through religion’. While narratives concerning this type of desistance often focus on the transformative aspect of religion – be that Born Again evangelical churches or Islam – there is also a clear performative element to it that is important to consider:

  • Shaun was a New Yorker gang member who was persuaded to leave the gang by a reformed gangster turned pastor. After leaving the gang, Shaun joined his church, in part because he had to convince his gang, other gangs, and the local township community that his conversion was genuine. This involved a variety of interlinked strategies. Firstly, Shaun was vouched for by somebody considered ‘respectable’ – his mother. She went to the church and said, ‘My son is no longer a gangster. Please pray with me. Please pray for him to be strong’. Although Shaun became a regular church-goer, he also engaged in a round of penance within the local community: ‘After I left the gang, I went around to all the neighbours and apologized for what I had done. They said they forgave me’. Finally, Shaun needed to show gang members – former friends and enemies alike – that he had truly become religious. As he put it, ‘if you are a Christian and carry a bible, then you can get out, but not by yourself. You must be a Christian in their eyes’. This conversion had to be performed over and over again, on a daily basis.

The third desistance narrative, which we might term ‘desistance through prison and organized crime’, might initially appear slightly counter-intuitive. However, street gangs and drug dealers occupy structurally different worlds in South Africa, and moving from a gang into drug dealing is therefore an exit strategy (see Jensen 2000).

  • Kelly had belonged to a gang called the Sexy Boys since he was very young. After several stints in prison for gang-related violence, he acquired a senior position within the prison gang hierarchy. Through the network that he established in the prison, he was able to transcend the highly localized world of township gangs and establish himself as the central drug dealer. He supplied drugs to all of the different, competing street gangs whose leaders often became street level drug pushers. In the process, the violence he exercised was instrumental and his relationship to state authorities transformed from confrontational to corrupt and accommodating.

The three case studies illustrate that however desistance occurs, and whatever the reason, one universal element of leaving the gang is that the patterns of violence of ex-gang members change. In the case of Gerard, he was able to reduce but not entirely eliminate his participation in violence. Shaun, on the other hand, stopped being publically violent altogether, as this would have clashed with his new religious beliefs. Kelly’s violence became more instrumental and less visible, and was often in the service of the community as he punished violent husbands or wayward boys. This tallies very much with gang research in other contexts, which has also shown that in some cases individuals who leave their gang do not necessarily become less violent, but rather, the type of violence changes. So, even though the individual has left the gang, he continues to engage in violence on a regular basis, although not in the same way as previously (see e.g. Jones 2014; Madzou 2008; Mohammed 2014; Rodgers 2014; Utas 2014). Understanding the reasons how and why this is the case is obviously critical, and may have important ramifications for DDR programmes, especially as one of their more important features is to provide occupations for demobilized ex-combatants. To explore the relationship between practices of violence and post-gang occupational trajectories, the next section draws on ongoing research in Nicaragua,7 in particular a set of detailed life histories compiled by regularly ‘repeat interviewing’ seventeen individuals who were members of the local gang in barrio Luis Fanor Hernández, the poor neighbourhood in Managua, the capital city of Nicaragua.8

Transforming Violence

Although the nature and dynamics of gangs in Nicaragua have changed dramatically over the past two decades (see Rodgers and Rocha 2013), being a gang member has consistently remained a finite social role. Certainly, there exists a common saying among Nicaraguan gang members that ‘no hay viejos pandilleros’ (‘there are no old gang members’). Consequently, even if the age at which individuals exit from the gang has varied, there has always been an upper limit. In barrio Luis Fanor Hernández, this tended to be around 21–22 years of age during the 1990s, around 24–25 years of age in the early and mid-2000s, and around 18–19 years of age in the early and mid-2010s. Generally, individuals’ motivations for leaving the gang tend to be highly personal and idiosyncratic. Among those reported to Rodgers over the years are – in no particular order – having children, experiencing a violent trauma (e.g. being severely injured), imprisonment, boredom, having a lucky escape, being forced to join the army, emigrating, a death in their family (especially the fathers of eldest sons), moving away from the neighbourhood, parental pressure, evangelical religious conversion, betrayal by other gang members, and finding employment.9 Some of these reasons have been more common than others at different points in time over the past two decades, and they do not all consistently lead to desistance. However, one constant of gang desistance in barrio Luis Fanor Hernández – and indeed, Nicaragua more generally10 – is that it is always non-hostile (i.e. consensual).11

There have clearly only been a limited number of potential occupational trajectories open to individuals who have left the gang in barrio Luis Fanor Hernández during the past two decades. The eight major ones – in order of frequency among ex-gang members – are: (1) un(der)employment; (2) informal employment; (3) involvement in vocational criminality12; (4) migration; (5) formal employment; (6) political activism; (7) imprisonment for an extended period of time; and (8) joining the army.13 As summarised in Table 1 below, different post-gang trajectories have different consequences for individuals’ use of violence, with some leading to a clear reduction in violence, while others lead to continued use of violence, while one case lead to both in relation to different individuals:

Trajectory Violence reduction Continued violence

(1) Un(der)employment X
(2) Informal employment X
(3) Vocational criminality X
(4) Migration X
(5) Formal employment X
(6) Political activism X
(7) Going to prison X X
(8) Joining the army X

Table 1

The relationship between violence and post-gang occupational trajectories.

The violence-related consequences of some of the trajectories are by no means surprising. For example, gang members who migrated – whether to the USA or Costa Rica – all reported having to ‘behave myself’ in order to avoid attracting attention, regardless of their legal status – which is not unexpected considering the way that migrants worldwide are frequently stigmatised and targeted by the authorities of their host countries. Similarly, employment was also associated with reduced patterns of violence. Some ex-gang members explained that this was because their jobs did not provide them with any scope to be violent. This was perhaps most obvious in relation to the (few) individuals who had managed to secure formal jobs – for example, Spencer in the production line of a free trade zone assembly factory, or Pecho stacking boxes for a local paint company – but somewhat more surprisingly it also applied to individuals in informal employment too. For example, Milton, who had set up his own tortilla making business, explained that he’d sold his handgun because ‘I haven’t needed a gun since starting my business… who’s going to steal tortillas?’ Other employed ex-gang members reported curtailing their violence due to the fear of losing their job – whether they were formally or informally employed – which points to the importance of wider labour market dynamics, insofar as Nicaragua suffers high un- and under-employment rates, particularly among youth.

At the same time, however, broader labour market dynamics also clearly contributed to underemployed ex-gang members engaging in persistent patterns of violence. Jasmil, for example, worked sporadically on construction sites on a casual basis, and frequently got into scraps on construction sites in order to ‘impose myself and not have to do the shit jobs… [because] if you end up doing those, you don’t always get paid, because you’re just one more guy, the last one at the end of the queue’. The fact that he had previously been a gang member in fact gave him a comparative advantage vis-à-vis those who were not, as Jasmil explicitly recognised: ‘I’m lucky, because of my past, I know how to fight, but there’s guys like me out there who have no idea, and they’re the ones who always get stuck doing the worse jobs on the construction sites’. This continued resort to violence can be directly related to Jasmil’s precarious position, with his ‘expertise’ gained from having been a gang member one of his only assets in such circumstances. There were also other occupational trajectories that led to continued patterns of violence, partly due to the intrinsic nature of the activities that these entailed. For example, involvement in professional criminality in the form of drug dealing rather logically led to violence insofar as the illicit nature of the drug trade means that drug dealers do not have access to legally enforceable contracts or property rights, and violence rapidly emerges as a primary tool with which disputes are resolved and uncertainty minimized. Gang members effectively represented a category of individuals that Charles Tilly (2003) has labelled ‘violence specialists’, and were therefore positioned in a privileged manner to engage in local drug dealing by virtue of their brutal comparative advantage.

The same was also the case of ex-gang members who engaged in political ‘activism’ in barrio Luis Fanor Hernández, an occupation that actually corresponded more to a form of political instrumentalisation, insofar as ex-gang members were hired by the ruling Sandinista party to violently disrupt opposition rallies by armed youth. Through the intermediary of the local representative of the Juventud Sandinista, or Sandinista youth organisation, ex-gang members were supplied with mortars, guns, and machetes, as well as ammunition, and paid 200 córdobas each time they were bussed out to disrupt a demonstration – an activity which they were obviously adept at due to their experience as gang members. Weapons were to be returned within 24 hours of having been distributed, which made this an extremely attractive deal, as a gang member called Chucki explained:

I couldn’t fucking believe it, man. The bus would come and pick us up in the morning, they’d give us the weapons, ammunition, and half our money, and then we’d fight, and after that go home, and still have the guns and shit for the rest of the day and the night, which meant that we could use them for whatever we wanted, like assaulting or robbing people, which was just perfect… Then the bus would come the following morning and we’d have to give them the weapons and we got the other half of our money. It was all such fucking easy money, maje, can you believe it?

This particular form of political ‘activism’ consequently did not just lead to violence in the context of ex-gang members’ ‘politicking’, but also more extensively.

A similar situation occurred when ex-gang members joined the Nicaraguan Army. One such individual was Jhon, who served a term between 1997 and 2002. His family enlisted him because they could ‘no longer cope with him’ and hoped that it would ‘educate him’, as his mother Doña Aurora put it. This the army did, but probably not in the way that his mother expected, as Jhon explained:

[The Army is] where I learnt to use real firearms, the AK-47, the sniper rifle, the RPG—which is a rocket-launcher—all kinds of weapons! I had classes, it was like school, and they taught us to shoot, to strip and clean our weapons, and there were also exams. I can strip and re-assemble any kind of weapon—I know everything, I tell you! The basic weapon in the Army was the AK-47, but because I could shoot really well, I became a sniper, and so used a special rifle. I went and trained in Martinique and Marie-Galante, they’re French islands, and I trained with the French Army and also the Venezuelan Army…

On returning to barrio Luis Fanor Hernández after his five year term, Jhon immediately re-joined the gang and was responsible for turning it into one of the most violent and feared gangs in southeast Managua between 2002 and 2006. He drew on his specialised military training to teach the new generation of gang members, thereby replenishing the gang’s knowledge stocks about firearms which had been eroding due to generational turnover. As a result, the gang became (literally) one of the most brutal in its district (see Rodgers 2013: 26–28, for further details about this).

The final typical post-gang trajectory involved going to prison for an extended period of time. Much has been written on prisons, of course, particularly on the way that they rarely rehabilitate but more often than not actually socialize into heightened forms of violence (Steinberg 2004). This was certainly the case of an ex-gang member Mayuyu, who served a four year prison sentence for assault between 2002 and 2006, and who clearly revelled in his experience in prison, explaining: ‘fue lo máximo, maje (it was the best, mate), I learnt so much in there… There’s no shitting about in prison, it’s kill or be killed… There’s nothing hypocritical (trucha), you know? You’ve just got to impose yourself and be poderoso (powerful), and that’s it!’ Partly as a result of his socialization into prison violence, on his release Mayuyu became one of the most brutal ‘enforcers’ of the local drug ‘cartelito’ that ended up supplanting the gang between 2006 and 2010 (see Rodgers and Rocha 2013). This contrasted strongly with Mungo’s experience of being imprisoned, however. Sentenced to eight years of prison for drug dealing in 2011, he was released after serving only three, but was obviously very affected by the whole experience, as he explained: ‘it was hard, maje, real hard. You can’t trust anybody in there, you’ve always got to look out for your ass, because otherwise they’ll fuck you… The only law in there is the law of the jungle, the only way to survive is by beating the shit out of anybody who wants to fuck with you… I got so tired of that, I tell you…’ Partly as a result of this, Mungo, who had previously been an extremely violent member of the barrio Luis Fanor Hernández gang, did not rejoin the gang and very noticeably adopted a much more docile, passive way of being, which he explained as ‘keeping out of trouble’ (‘no me pongo por ningún alboroto), ‘not looking for anybody’ (‘no busco a nadie’), and ‘staying quiet’ (‘me quedo quieto’). To a large extent, however, this difference in the consequences of imprisonment came down to Mungo and Mayuyu’s different personalities, with the latter generally displaying psychopathic tendencies, while the former had always, even at his most violent, been relatively reflexive and thoughtful.

These different trajectories and their differing effects on ex-gang members’ patterns of violence are clearly the result of a combination of push and pull factors. Intrinsically, violence constitutes a comparative advantage for certain occupational trajectories, namely unemployment, professional criminality, and political activism. These trajectories can therefore be said to inherently socialize former gang members into new forms of violence (although in the latter two cases, it could be argued that there is something of an endogeneity effect insofar as recruitment into these two occupations was clearly partly dependent upon individuals’ previous status as gang member ‘violence specialists’). In other words, violence here is an effect of the occupational trajectory. Extrinsically, it is the broader context associated with formal employment, informal employment, and migration that determines the potential for violent behaviour in relation to these occupational trajectories – namely, reducing it. In other words, it is the circumstances that ‘de-socialize’ individuals from patterns of violent behaviour, or put another way, they affect violence. This was the case for Mungo after his prison term, but not for Mayuyu, for whom prison provided new skills in a way that was similar to Jhon joining the army. Overall, though, what the former gang member trajectories highlight is that the dynamics of the broader social setting within which these lives are lived matters tremendously, and needs to be taken into account a priori when thinking about violence reduction. In particular, the fact that in some cases the ability to be violent constituted a major comparative advantage in enabling former gang members to ‘succeed’ after leaving the gang needs to be acknowledged.

Conclusion

In a recent article on ‘Building Peace in Conflict and Crime-affected Contexts’, the political scientist Achim Wennmann (2014: 270) makes a case for drawing on the lessons learnt from DDR programmes in order ‘to assist negotiated exits from organized crime’. This article suggests that the opposite is also true, and that DDR programmes have much to learn from organic gang exit processes as well as from the state-sanctioned wars on gangs. Both DDR projects and wars on gangs suffer from a range of shortcomings: they work with faulty assumptions, a misplaced trust in outside intervention and through globally dispersed templates and blue prints. Robert Muggah (2010: 4) rightly argues that ‘DDR is a process of social engineering’; as such, it is different from the more organic gang trajectories that we have explored in this article. However, we contend that there are important lessons to be learnt from the organic process of gang desistance, particularly when thinking specifically in relation to violence reduction. The way different occupational trajectories have variable effects on patterns of violence, with some leading to an increase, some to a decrease, and one provoking both an increase and a decrease highlights how patterns of violence respond to a complex mix of both intrinsic and extrinsic factors. The prior experience of violence of ex-gang members clearly constituted a major advantage within the context of certain trajectories, both for endogenous reasons – for example, the illegal nature of drug dealing – as well as exogenous ones – the highly constrained nature of the broader labour market, for instance. At the same time, the fact that some trajectories can lead to both an increase or a reduction in violence highlights how we are not just talking about structural issues, but about the way that structure and individual agency interact, as Mungo and Mayuyu’s different post-prison engagement with violence highlights. These insights go against the grain of some of the central assumptions underpinning DDR programmes, including in particular the notion that their primary aim is to stop the violence in order to promote development.

As Robert Bates (2001), Gareth A. Jones and Dennis Rodgers (2016) have highlighted, it is by no means proven that peace is a necessary pre-condition for prosperity, and the occupational trajectories of former barrio Luis Fanor Hernández gang members highlight how under certain circumstances violence can constitute a major comparative advantage for individuals, sometimes even their main one. Seen in this way, if the point of a DDR programme is to provide assistance to former combatants and enable them to find jobs, including starting their own businesses, then perhaps the most efficient way to achieve this is to allow them to maximize their violence expertise. This is, of course, perhaps easily dismissed in relation to illegal economic activities, where the positive link to violence is intrinsic in nature. However, violence also emerged as a major advantage for ex-gang members for extrinsic reasons, and more specifically the constrained nature of wider labour market dynamics and the lack of opportunities that exist for young men in Nicaragua and elsewhere (in South Africa, for example, many ex-combatants sought to join private security companies or state policing structures precisely because of their comparative advantage as ‘violence specialists’).

Considering that most DDR programmes are implemented in post-conflict contexts where economic activity has (literally) been shot through, it is obvious that many of the most attractive opportunities available will often be those easily achievable through violence. The same is true of politics, which is often highlighted in DDR programmes as a potential channel through which to rehabilitate ex-combatants. In Nicaragua it was youth’s violence expertise that led to their involvement in politics. Of course, this was partly a function of the fractious and clientelistic nature of the Nicaraguan polity, but many post-conflict contexts are in fact very similar, making the political integration of ex-combatants that is often a major element of DDR programmes a potentially dangerous proposition. At the same time, the exit narratives of South African gang members highlight processes that are rarely taken into account in DDR programmes, such as founding a family or finding religion. At the very least these raise the issue that economics and politics may not be the only driving forces to demobilize and reintegrate individuals.

The general problem with template-based programmes is that they seek to impose what Roger Mac Ginty (2008: 148) refers to as ‘peace from IKEA: a flat-pack peace made of standard components’. In contexts where violence constitutes an asset, unless violence is part of the flat-pack, then the furniture is unlikely to stay up (so to speak). Ultimately, the critical issue, as Nat Colletta and Robert Muggah (2009: 432) have pointed out, is that ‘post-conflict contexts are not a terra nullius upon which discrete technical solutions are readily grafted’. Rather, we need to consider the detail of each situation and come up with tailored programmes that take into account broader structural dynamics as well as the particularities of a given situation, whether in terms of the actors involved, the spread of economic assets, and general socio-political dynamics. This includes deciding whether or not violence needs to be an acceptable part of the peace ‘package’. This would require listening to would-be ex-combatants and ex-gangsters, in order to understand their desires and aspirations, and it would also require that wider society be open to seemingly unacceptable or difficult options, such as promoting military and security-oriented careers. It would also make the central issue of any policy intervention less about reducing violence and more focused on managing it in ways that lead to the least socially destructive outcome. Inevitably, this will mean having to make unpalatable choices, privileging certain groups and activities, and accepting that there will be winners and losers.

Having said this, and thinking in terms of the gang trajectories highlighted above, one issue to explore in order to mitigate this is the way that the relationship between violence and particular trajectories waxes and wanes over time. Many DDR programmes are seen as events rather than processes – this is partly a function of their ‘checklist’ nature. Yet the passage of time is itself a critical factor to take into account, insofar as relationships and configurations rarely remain locked in place for long durations – all the more so in volatile post-conflict contexts – and situations can change very rapidly. Seen from this perspective, then, ultimately, the question – at least vis-à-vis gang members – has to be less about reducing their violence, and more about how to enable them to develop life trajectories in a way that does not make violence the bottom line, but rather situates it within a broader developmental trajectory that tends towards sustained peace. In this regard, and slightly provocatively, we could propose that one lesson that DDR interventions might take from gang research is that it is often better to do nothing, as most combatants and gangsters tend to find ways back into ‘polite society’, one way or another. Perhaps more realistically – in view of the nature of DDR as ‘big business’ (Theidon 2008: 1) – interventions should at the very least aim at supporting organic forms of desistance rather than trying to supplant them with artificial and rigid templates. Both approaches would however mean thinking of DDR interventions as processes rather than as events, with the management of violence rather than its reduction the central goal.

Competing Interests

The authors declare that they have no competing interests.