Organised Crime

Addressing organised crime and reducing the power and reach of criminal organisations requires a careful balance between supply-side enforcement – aimed at reducing the presence and operational capacity of criminal groups – and demand-side prevention – focused on lowering the incentives and opportunities for individuals to engage in illicit activities. This section reviews the main categories of interventions, evaluating their effectiveness in both the short and long term, and highlighting empirical evidence, particularly causal findings, on outcomes and unintended consequences. The analysis covers the most documented cases, yet it aims to build a broad understanding of what works in the fight against organised crime. Ultimately, the effectiveness of these policies depends on several factors: the type of criminal organisation, its level of legitimacy, the profitability of its illicit markets, and the state’s capacity to implement enforcement with due process and maintain control over the prison system (Blattman 2025, Magaloni et al. 2020, Alarcon Barrera and Sviatschi 2025, Schargrodsky and Tobón 2025).

Enforcement and State Capacity Approaches

Supply-side strategies aim to directly suppress criminal organisations through policing, military force, legal action, and institutional reforms. These interventions address the immediate ‘supply’ of organised crime by incapacitating or deterring perpetrators. Evidence suggests that traditional crackdowns can sometimes backfire by provoking violent adaptation, whereas more focused and accountable enforcement can yield better outcomes.

Policing, Crackdowns, and ‘Kingpin’ Strategies

Conventional enforcement often involves aggressive policing, raids, and the ‘kingpin’ strategy of eliminating top criminal leaders. In Latin America, governments have frequently deployed militarised crackdowns (e.g. ‘mano dura’ policies in Central America, large-scale cartel offensives in Mexico and Colombia, and heavy weaponry in police patrolling in Brazil and Colombia, among others). These approaches seek to fragment and weaken criminal groups, but research indicates that they can trigger violent retaliation and organisational fragmentation. When kingpins are arrested or killed, criminal hierarchies destabilise and factions fight for dominance, leading to surges in violence. This is particularly the case in contexts with highly lucrative markets – such as drug trafficking – where demand for illicit goods remains strong despite increased enforcement (Castillo et al. 2020).

For example, the beheading of Mexican drug cartels has been shown to increase the homicide rates in the affected areas in the short-run (Calderón et al. 2015, Lindo and Padilla-Romo 2018). Similarly, broader data from Mexico’s drug war demonstrate that confrontational crackdowns led to turf wars: intensified law enforcement activity fragmented cartels and escalated violence both locally and in adjacent communities (Atuesta and Ponce 2017, Dell 2015). These unintended consequences can explain a substantial share of the homicide increases during Mexico’s 2006–2010 drug war. Military policing approaches have proven ineffective as well. In Cali, Colombia, a large military policing intervention in areas with entrenched organised crime did not reduce crime and raised concerns about human rights violations (Blair and Weintraub 2023).

In essence, ‘divide and conquer’ tactics may temporarily disrupt cartels, but they risk violent reconfiguration of criminal markets rather than elimination of illicit activity. Importantly, these negative effects can persist: even if the immediate shock of a crackdown dissipates, newly formed splinter groups continue to drive elevated violence in the long run (Atuesta and Ponce 2017). Enforcement that weakens one group can invite new entrants or gangs to fill the void, perpetuating cycles of conflict.

Beyond violence, heavy-handed crackdowns can undermine the legitimacy of the state. In Medellín, Colombia, intensified police presence led gangs to strategically respond by providing more order and governance in their neighbourhoods – albeit to protect their drug rents rather than altruistically (Blattman et al. 2025). In Central America, mano dura policies in the 2000s – mass arrests of suspected gang members based primarily on appearance or place of residence – resulted in few convictions, yet had deeply counterproductive effects. These sweeps eroded trust in authorities, reinforced the stigmatisation of marginalised communities, and paradoxically accelerated gang recruitment by fuelling resentment and a sense of solidarity among targeted youth (InSight Crime 2010). Moreover, the absence of effective state control over prisons further enabled the consolidation and expansion of gang structures.

One of the most prominent recent examples of an intensive territorial recovery effort is the 2022 intervention in El Salvador. Under a declared state of emergency, the government launched a sweeping operation aimed at reclaiming control over gang-dominated areas. The strategy included the deployment of military forces, the creation of security perimeters, and the mass arrest of suspected gang members. The core objective was to dismantle entrenched gang networks and restore state authority in neighbourhoods where these groups had exercised near-complete control for years.

Melnikov et al. (2025) analyse how the same gang territories and nearby areas they studied during the period 1996-2019 in Melnikov et al. (forthcoming) look after the crackdown by using geocoded surveys in 2022 and 2023. The study finds that after the intervention, previously gang-controlled neighbourhoods experienced a complete removal of mobility restrictions. Moreover, recruitment also declines significantly. Before the crackdown, roughly 63% of residents reported that teenagers in their communities were being recruited by gangs; that figure dropped to just 8% following the intervention. Although income disparities between these areas and other parts of the city remain, the gap was cut in half – largely due to improved access to jobs outside former gang strongholds. Still, residents now reported to face discrimination in the labour market, as employers often still associate these neighbourhoods with gang involvement. To test this, in 2024 the researchers conducted an experiment with 400 firms and found that, even when applicants had identical qualifications, those from stigmatised areas were systematically less likely to be chosen. In terms of state legitimacy and access to public security, the study also highlights that past experiences of police abuse – reported by about 20% of respondents – significantly reduce people’s willingness to seek help from law enforcement. This lingering mistrust, along with persistent stigma and labour market exclusion, may present ongoing challenges for reintegration and the restoration of effective state presence in these communities.

While this early evidence suggests that the crackdown has weakened gang structures in El Salvador, more research is needed to assess its long-term impact on criminal activity. Moreover, it remains unclear whether El Salvador’s experience can be generalised to other settings. In this particular case, gangs operated primarily through coercion, held little legitimacy among the population, and relied on extortion – a business that was neither highly profitable nor fuelled by increasing demand. Unlike other criminal groups, they were not known for providing public goods or community services that might generate local support. In contrast, in contexts where criminal organisations enjoy higher levels of legitimacy or operate within more lucrative markets, such as drug trafficking, the effectiveness of similar strategies may be more limited. Additionally, without strong due process safeguards, comprehensive state control over prisons, and robust rehabilitation programmes, the long-term reintegration of incarcerated individuals is likely to fall short. This underscores the need for further research on effective reintegration strategies and the importance of more humane prison systems – particularly in low- and middle-income countries. Moreover, more evidence is needed to understand how the recruitment, training, and oversight of police officers, as well as the role of intelligence-led policing, can help reduce the risk of escalating violence during crackdowns or community interventions. In this regard, recent evidence from the United States suggests that targeted police training in de-escalation, procedural justice, and community engagement can significantly improve police-community relations and reduce the use of force without compromising public safety (e.g. Owens et al. 2018).

Targeted Policing and Focused Deterrence

In contrast to indiscriminate crackdowns, more targeted and intelligence-driven policing has shown promise in curbing organised violence without broad collateral damage. One such strategy is focused deterrence, which concentrates enforcement and social services on the small number of individuals or groups driving violent crime. Pioneered in US cities (e.g. the Boston Ceasefire gang intervention), focused deterrence involves directly communicating to gang members that violence will bring swift, certain punishment while offering assistance (job training, counselling) for those who choose an exit path (Abt 2019, Kennedy 2011). This ‘carrot-and-stick’ approach, often implemented via call-in meetings between gang members, police, and community leaders, seems to achieve success in reducing urban gang violence (Braga et al. 2019). Nonetheless, large-sample causal evidence is still lacking.

These programmes leverage community legitimacy and credible enforcement to fundamentally alter offenders’ cost-benefit calculations: gang members are offered a way out of violence, but also warned that any continued violent activity will bring unified law enforcement attention. Sophisticated, organised criminal structures might require a focused deterrence approach that goes beyond the traditional attention on violence and homicides. The main reason being that often criminal groups grow stronger with peace – as their businesses thrive (Blattman et al. 2025, Brown et al. 2025). The broader idea is that of conditional repression – as opposed to indiscriminate repression – to discipline the behaviour of gangs or criminal organisations, not only on violence but also on extortion or other socially costly behaviours such as recruitment of children (e.g. Lessing 2017b, Alarcón Barrera and Sviatschi 2025).

Other policing innovations include hot-spots policing (deploying patrols to crime micro-locations) and community policing (building police-community trust to improve intelligence and co-production of security). Hot-spots policing strategies have reduced street crime in various cities, though their specific impact on organised crime (gangs and extortion networks) is less clear, may require sustained presence to prevent displacement, and has not been evidenced in Latin American cities with moderate to heavy presence of organised crime (Blattman et al. 2021, Collazos et al. 2021). Community-oriented policing, which has been piloted in some Latin American contexts, aims to rebuild legitimacy by making police more accountable and responsive to residents. The idea is that when citizens trust authorities, they are more willing to report crimes and resist criminal intimidation (Abril et al. 2024). Yet, the evidence suggests that improving police-community relations is difficult to achieve in areas long dominated by gangs (Blair et al. 2021). The broader lesson is that professionalising police forces – through better training, oversight, and community engagement – is a necessary complement to any enforcement strategy. Abusive or corrupt policing not only fails to dismantle criminal groups; it can actively drive civilians into the arms of those groups for ‘protection’, undercutting the state’s authority (Magaloni et al. 2020). Thus, building a capable and trusted law enforcement apparatus is a long-term project central to any supply-side response.

Judicial Reforms and Institutional Integrity

Because organised crime often thrives on weak institutions, strengthening the justice system and governance is another critical supply-side response. This includes anti-corruption measures, legal reforms, and administrative actions to root out criminal influence from public offices. A striking example comes from Italy’s effort to combat mafia infiltration of local governments: under a 1991 law, the national government can dismiss municipal councils and replace them with temporary administrators if there is evidence of mafia control. Research on this intervention (often called commissioner dissolutions) shows encouraging results: in affected Italian municipalities, the removal of corrupt local officials was followed by higher social capital and economic growth, including significant increases in employment, new firm formation, and property values, especially in sectors previously dominated by mafia cartels (Buonanno et al. 2024, Fenizia and Saggio 2024). Notably, this intervention did not lead to an increase in violence. Rather, mafia-linked violence decreased following the dismissal of the city councils (Baraldi et al. 2024). The implication is that freeing local governance from criminal capture can unleash legitimate economic activity that organised crime had been suppressing (e.g. through extortion and cartel control of contracts), highlighting the potential long-term development gains from reclaiming captured institutions.

However, institutional reforms face challenges of sustainability. In Italy, the positive effects of city council dismissals may fade if oversight is not maintained – mafia networks sometimes re-infiltrate local governments once the temporary administrators leave. Similarly, in Latin America, attempts to purge corrupt officials or police have had mixed success; new corrupt actors can emerge if underlying incentives (low salaries, political protection for criminals) remain (Ungar 2011). This suggests that one-off purges or takeovers need to be paired with deeper institutional changes. Those include improving salaries and meritocratic recruitment for police and judges, reforming criminal codes to expedite prosecutions, and providing better protection for judges, prosecutors, and witnesses who take on organised crime cases (Dahlström et al. 2012). Legal innovations like specialised anti-mafia prosecution units, witness protection programmes, and laws that criminalise membership in illicit associations (e.g. RICO statutes or conspiracy laws) can increase the state’s fight against sophisticated criminal networks. Indeed, the mere signal of credible enforcement can deter criminals: evidence from the United States indicates that when the government successfully prosecutes high-level criminals (or confiscates their assets), it raises the perceived risk of doing illicit business (Levin 2013). Yet if prosecutions appear politicised or selective, they can backfire, eroding public trust. The integrity and capacity of institutions – police, courts, local governments – therefore form the backbone of an effective supply-side response. Without well-functioning institutions, aggressive policing campaigns or dramatic arrests become short-lived ‘wins’ that do not fundamentally diminish organised crime in the long run.

Prisons and Rehabilitation

An often-overlooked arena of the fight against organised crime is the prison system (e.g. Schargrodsky and Tobón 2025). In many countries, especially Latin America, prisons have become incubators and headquarters for organised gangs – a direct consequence of mass incarceration policies and weak prison governance (e.g. Lessing 2017a, Lessing and Denyer Willis 2019, Skarbek 2011, 2012). Thus, improving prison management and investing in rehabilitation are key policy responses on the supply side. One crucial step is to prevent prisons from serving as crime hubs: this may involve segregating the most influential crime leaders under high-security regimes to curtail their communication. Italy’s hardline prison regime, which strictly isolates mafia bosses, exemplifies an attempt to cut the command and control that kingpins can exert from behind bars (Barbarino and Mastrobuoni2014). While harsh, such measures aim to disrupt criminal networks’ ability to coordinate violence and extortion on the outside. In Latin America, by contrast, overcrowded prisons with minimal surveillance have enabled jailed kingpins to effectively run their organisations remotely. The policy lesson is that without proper prison controls, tough sentencing can unintentionally strengthen organised crime by concentrating and connecting its members.

Reforms to prison conditions and inmate rehabilitation can mitigate these risks (Tobón 2022). Reducing overcrowding and separating rival factions can lower violence within prisons (prison violence often spills over into street violence when riots or orders for revenge occur). Additionally, offering education, job training, and re-entry support to inmates – many of whom are low-level gang members – may reduce their likelihood of returning to crime after release. Rigorous evaluations of rehabilitation programmes are still limited in the context of organised crime, but evidence from related contexts is instructive. For example, cognitive-behavioural therapy (CBT) programmes for high-risk violent offenders (discussed more below) have achieved long-term reductions in reoffending by teaching emotional regulation and non-violent conflict resolution. In one study, Blattman et al. (2017) worked with Liberian ex-combatants and found that a combination of therapy and short-run economic assistance caused a sizable drop in criminal behaviour.

Prevention and Community Focused Approaches

Pure enforcement, even when refined, cannot by itself eliminate organised crime, especially in societies where social and economic conditions make illicit livelihoods attractive. Demand-side strategies address the underlying factors that generate support for, or acquiescence to, organised crime. These include poverty and inequality, youth marginalisation, lack of legitimate opportunities, and community disempowerment. By reducing the demand for participation in criminal activities (or the demand for the illicit goods and services that organised crime provides), these approaches aim for more sustainable, long-term reductions in organised crime. Below we survey preventive interventions ranging from economic development to violence mediation, emphasising evidence on their efficacy.

Recruitment Interventions

Another strategy to reduce the expansion of criminal organisations is to increase the cost of their operations by generating a shortage in labour supply. This approach targets the recruitment pipeline by limiting the availability of young individuals who might otherwise be drawn into criminal networks. Evidence from El Salvador supports the effectiveness of this strategy. In 2015, the government launched a school-based police intervention aimed specifically at reducing gang recruitment. The programme involved deploying 2–3 police officers during school entry and exit times to focus solely on preventing recruitment and ensuring the safety of students. According to Castro et al. (2025), the intervention led to a significant reduction in the likelihood that adolescents would be recruited into gangs or later appear in prison records. By disrupting the supply of new, low-cost recruits – particularly minors – the programme increased operational costs for gangs and contributed to a longer-term decline in extortion activities.

Several features explain the programme’s success. First, the police force selected for implementation had a preventive orientation and no history of violence or abuse, which helped build trust within the community. Second, the intervention was highly focused in both timing and scope: officers concentrated their efforts on the specific windows when students were most vulnerable, without engaging in broader law enforcement duties such as tackling extortion. This narrow focus likely reduced the risk of retaliation, as the gang’s immediate income sources were left untouched. This was particularly effective because newly recruited minors typically start with low-level surveillance roles – such as monitoring who enters the neighbourhood – and only later become involved in extortion. Third, the territorial structure of gangs in El Salvador – organised into semi-autonomous ‘clicas’ with exclusive control over specific areas – made it difficult to shift recruitment to non-intervened zones. Finally, since extortion was the gang’s main revenue stream and is relatively low-margin, they faced limited financial flexibility to respond – for instance, by raising incentives to attract older recruits. Combined, these features made the intervention an effective strategy for disrupting gang recruitment while avoiding a violent backlash.

Nevertheless, further research is needed to assess whether these types of strategies can be effective in other settings – particularly where criminal organisations operate more lucrative enterprises or are structured in ways that allow them to adapt and relocate their activities in response to enforcement efforts – e.g. if labour demand is highly inelastic, organised crime might just recruit others. In these other contexts, such interventions could be complemented with strategies that alter the economic incentives for minors to join gangs – for example, by expanding access to formal employment opportunities, offering targeted social programmes, or strengthening educational pathways. Combining enforcement with approaches that address the root causes of youth vulnerability may be essential to reduce recruitment and ensure more sustainable results.

Broader Socioeconomic Interventions and Incentives

A robust finding in the development literature is that improving legal economic opportunities can deter involvement in crime. Organised criminal enterprises often recruit from pools of unemployed or underemployed youth who see few alternatives to the high (if risky) income of illicit work. Accordingly, governments have tried to “dry up” the recruitment base through antipoverty programmes, education, and employment initiatives. Conditional cash transfer (CCT) programmes are a prominent example. CCTs, such as Mexico’s Oportunidades or Brazil’s Bolsa Familia, provide cash stipends to low-income households contingent on behaviours like children attending school. While primarily aimed at human capital and poverty, CCTs have shown spillover effects on crime reduction (e.g. Chioda et al. 2016, Machado et al. 2018).

In Peru, Sviatschi (2022a) shows that areas with coca cultivation experienced a significant decline in both coca production and child labour following the implementation of a CCT programme that provided monetary incentives to parents conditional on their children’s daily school attendance. Although the programme specifically targeted school-aged children and was not designed to reduce drug production, the findings suggest that CCTs could be strategically directed toward coca-suitable districts to mitigate the adverse effects of increased returns in the illegal sector – particularly in regions with high coca prices – documented in previous sections.

In Brazil, researchers found that expanding Bolsa Familia coverage in high-violence areas was associated with lower homicide rates, particularly reductions in gun-related killings and youth victimisation (Ingram and Da Costa 2014). The mechanism is likely twofold: the added income support alleviates extreme need (reducing economic motivation for joining gangs or theft), and the emphasis on schooling keeps adolescents in structured, supervised environments. This spatial analysis of violence in Brazil concluded that policies which reduce poverty and social vulnerability tend to reduce violence, and specifically highlighted Bolsa Familia’s expansion as contributing to drops in youth homicide rates.

Also focusing on Brazil, Britto et al. (2022) examine how involuntary job loss influences criminal behaviour and assesses whether unemployment insurance mitigates this effect. Leveraging administrative records that link employment histories, criminal charges, and welfare participation for the full population of Brazilian male workers, the study estimates that displacement through mass layoffs raises an individual’s propensity to commit crimes by roughly 23%. Yet, the study also finds that access to unemployment benefits entirely neutralises the crime spike following job loss, with the protective effect disappearing immediately once payments cease. The study does not focus on organised crime, but given the scope of the sample, it is not unreasonable to expect some marginal organised criminals to avoid illegal activities given the extent of social support.

Beyond cash transfers, job creation and training programmes for at-risk populations are another strategy (e.g. Chioda et al. 2024). These range from youth apprenticeship schemes to initiatives providing alternative livelihoods to farmers who might otherwise participate in the drug economy (e.g. coca crop substitution programmes in Andean countries). The success of such programmes has been mixed (Londoño Mesa et al. 2024). Some agricultural substitution efforts, like Colombia’s investments in coffee and cacao for former coca growers, have struggled unless paired with security improvements (farmers will not abandon coca without assurance that insurgents or traffickers won’t coerce them back into it). Indeed, providing subsidies to coca farmers has, unsurprisingly, increased coca cultivation in Colombia (Prem et al. 2023). Strong social support for released offenders can sustainably reduce re-entry into criminal activities and organisations (Barrios-Fernández and García-Hombrados 2025). Job training programmes for urban youth have shown positive impacts on employment, and in some cases reductions in crime or gang involvement, yet there is wide heterogeneity in who benefits and why (Davis and Heller 2020). One likely reason is that illicit earnings often dwarf entry-level wages in the formal sector, especially for low-skilled men, so the calculus to join a gang may not drastically change unless legal jobs are sufficiently attractive in relative terms. However, there are promising examples: a large-scale programme in Honduras that combined vocational training with a temporary wage subsidy for at-risk youth significantly reduced their self-reported participation in illegal activities two years later – though community violence levels did not change dramatically (Chioda et al. 2024). Another example is the READI programme in Chicago, which combines precise targeting of at-risk youth – through administrative data and community referrals – with an intense combination of CBT, an 18-month job, and social support. The programme led to substantial and sustained reductions in shootings and homicide arrests (Bhatt et al. 2024).

It is important to note, however, that most job training programmes target individuals over the age of 18 – an age group that is typically not the primary focus of criminal recruitment. Therefore, more research is needed on interventions that target younger populations, particularly during the critical windows when recruitment into organised crime is most likely to occur. In this regard, recent research on youth involvement in organised crime in Medellín, Colombia, offers some granular insights into the beliefs and constraints that channel adolescents toward gang life. Drawing on semi-structured interviews with leaders from more than eighty of the city’s roughly four hundred drug-selling gangs, Blattman et al. (2025) document detailed organisational hierarchies, promotion ladders, earnings schedules, and recruitment tactics that reward risk-tolerant teenagers yet expose them to high probabilities of violence and incarceration. These qualitative findings framed a large-scale survey of almost 10,000 seventh-grade boys enrolled in 100 low- and middle-income public schools, allowing for an unusually integrated view of the ‘demand’ and ‘supply’ sides of the criminal labour market.

The school-based instrument elicited subjective beliefs about monetary and non-monetary returns associated with four career categories – low-skill legal, medium-skill legal, high-skill legal, and gang work – together with perceived arrest and fatality risks and the social prestige attached to each option. Standardised scales captured future orientation, impulsivity, and time preferences, while respondents also stated their personal probability of joining a gang by age 25. Combining these data with neighbourhood crime statistics, the study estimated reduced-form correlates of self-reported recruitment risk and specified a structural model of occupational choice under uncertainty.

Findings indicate that nearly one in six boys placed their likelihood of gang entry at 20% or higher. Elevated risk clustered around three systematic distortions: (i) under-estimation of lifetime earnings attainable in medium- and high-skill legal professions; (ii) overestimation of the non-pecuniary rewards – status, excitement, social capital – conferred by gang membership; and (iii) the degree of future orientation and impulsivity. A randomised information treatment that contrasted actual versus perceived returns in both monetary and non-monetary dimensions produced immediate, statistically significant reductions in stated gang-entry probabilities, underscoring the malleability of recruitment intentions when beliefs are corrected. (An ongoing field experiment is testing whether a vocational orientation and behavioural intervention effectively reduces gang entry among the highest-risk adolescents identified in the survey.)

Youth Programmes and Cognitive Behavioural Therapy

Given the prominent role of youths in organised crime (both as perpetrators and victims), targeted youth development programmes are a key demand-side approach. These include after-school programmes, mentorship, behavioural therapy, and violence prevention curricula, often implemented in high-crime communities. The evidence from Medellín we discuss above suggests that labour market interventions – such as vocational orientation – should be implemented earlier, as individuals begin forming their expectations and making choices while still in school (Blattman et al. 2024). The rationale is to steer young people away from gangs during their formative years, build socio-emotional skills, and offer pro-social alternatives (education, sports, arts, job skills) that make gang life less appealing.

Recent rigorous studies in Latin America provide encouraging evidence on such interventions. For instance, an experiment by Dinarte-Diaz and Egana-delSol (2024) in El Salvador evaluated an after-school programme that combined academic support with cognitive behavioural therapy (CBT) techniques for students in some of the country’s most violent neighbourhoods. CBT sessions focused on improving self-control, coping with anger, and resisting negative peer influence. The study found measurable reductions in aggressive and disruptive behaviour in schools that received the programme, along with improvements in students’ decision-making under pressure. These outcomes suggest that even in extremely violent contexts, providing a safe space after school and training youths in emotional regulation can curb the minor violent incidents and conflicts that often serve as entry points to more organised violence. This echoes findings from the United States, where CBT-oriented programmes for at-risk adolescents (such as the Becoming a Man programme in Chicago) produced sharp drops in youth arrests (Heller et al. 2017). It also parallels the Liberia study noted earlier, in which high-risk young men saw their antisocial behaviour decrease after short-term therapy and employment (Blattman et al. 2017, 2023). However, more research is needed to determine whether CBT interventions can be effective in the long run in contexts of organised crime, where much of the violence is premeditated and individuals may join not out of impulsivity, but due to other factors such as perceived safety, as well as pecuniary and non-pecuniary returns. Thus, in the context of criminal organisations, the mechanisms through which CBT operates may need to be adapted or complemented with other interventions targeting structural incentives and social dynamics.

The policy implication is that youth interventions work best as part of an ongoing support system rather than one-off inoculations. Moreover, when after-school programmes, mentoring, or counselling are continuously available through critical years, they can interrupt the school-to-gang pipeline. In particular, interventions targeting the critical ages for gang recruitment can be effective in producing long-term impacts on organised crime (Sviatschi 2022a, Castro et al. 2025). Several Latin American cities have expanded after-school sports leagues, music programmes, and tutoring in violence-plagued neighbourhoods (e.g. Medellín’s network of library parks and youth centres established in the 2000s alongside its crime decline). These efforts, in tandem with improved policing, are credited with boosting social cohesion and offering youths status and belonging outside of gangs. Evaluations of some of these initiatives find reduced self-reported gang involvement and higher school retention among participants, though robust causal evidence on violent crime outcomes is still emerging. Nevertheless, the accumulation of micro-level successes points toward a preventive strategy centred on youth empowerment and psychological resilience as a valuable complement to policing. By changing attitudes and skills – instilling non-violent norms, future orientation, and conflict resolution abilities – such programmes address the ‘demand’ for criminal life at its source.

Changing Peers’ Norms at Critical Ages

Other types of school-based interventions aimed at reducing early violent behaviour at critical ages have focused on shifting social norms and attitudes toward reporting violence. One example is a study conducted in Peru by Gutierrez et al. (2024), which evaluated a preventive programme targeting adolescents aged 12 to 16. Through guided tutoring sessions, the programme emphasised the importance of reporting violent incidents to enable early intervention and prevent escalation. It also incorporated a digital platform that allowed students to submit anonymous reports and promoted a less punitive, more restorative approach to conflict resolution. The intervention aimed to increase the social cost of violent and antisocial behaviour by changing peer norms around the acceptability and value of reporting such incidents. This age group was specifically targeted because it represents a critical period when youth are particularly vulnerable to recruitment into gangs or engagement in criminal activities. The underlying rationale was that fostering a school environment where reporting violence is seen as socially acceptable and supported by peers would raise the perceived risks and reputational costs associated with antisocial behaviour, both in school and in the broader community.

The results showed a significant increase in the reporting of school violence, which was accompanied by a decline in students’ exposure to such incidents. Notably, students previously identified with aggressive behaviours were significantly less likely to commit crimes four years after participating in the programme. These findings suggest that interventions implemented during critical stages of youth development – by strengthening cooperative norms and deterrence from within the peer group, rather than relying solely on punitive measures or external incentives – can produce lasting reductions in criminal involvement.

Mediation, Negotiation, and Violence Interruption

A controversial set of demand-side responses involves directly engaging with criminal groups to negotiate a reduction in violence. These mediation or truce strategies treat violence as, in part, a problem of broken communication or strategic tit-for-tat that might be de-escalated through third-party facilitation (Blattman 2023). Examples include gang truces in El Salvador, where evidence indicates that when gangs successfully collude, divide territories, and reduce internal violence, they often repurpose the youth previously recruited for violent confrontations into other criminal activities (Brown et al. 2025). This shift allows them to expand their territorial control and scale up operations such as extortion, leading to increased financial extractions from local residents. As a result, gangs not only strengthen their economic base but also their capacity for coercion. These findings highlight a potential trade-off: while short-term reductions in violence may be achieved, they can come at the cost of long-term consolidation and territorial expansion of criminal organisations (Blattman et al. 2022). Furthermore, the truce’s gains are often proved short-lived, as such equilibriums are unstable. Apart from the El Salvador example – where violence spiked after the 2012 truce ended – a similar pattern was observed in a 1990’s gang truce in Los Angeles and a ceasefire in Trinidad and Tobago: temporary lulls in killings were followed by surges when the informal peace agreement broke down (Katz et al. 2016).

That said, mediation approaches are not monolithic. Violence interruption programmes like Cure Violence do not negotiate formal truces or rewards, but instead use credible community insiders to intervene in interpersonal disputes (such as retaliatory shootings) before they escalate. In some US cities and a few Latin American pilot projects (e.g. in Honduras, Mexico, and Colombia), violence interrupters have been credited with cooling down neighbourhood conflicts and preventing cycles of killings. Yet, rigorous causal evidence is still absent (e.g. Butts et al. 2015). These programmes treat violence as a contagious disease, aiming to interrupt transmission by identifying brewing ‘hot’ conflicts and mediating personal disputes (for instance, by separating feuding gang members or negotiating agreements to stand down). These programmes have the advantage of not requiring governments to concede any official authority to criminals (unlike formal truces), but they rely heavily on the skills and integrity of the mediators.

Finally, at the highest level, there are instances of demobilisation agreements akin to those in civil wars: encouraging entire armed groups to lay down arms in exchange for leniency or integration. Colombia’s peace deals – the AUC paramilitary demobilisation in 2003–2006 and the FARC guerrilla peace accord in 2016 – can be seen in this light, as both insurgent groups were deeply involved in organised crime (e.g. drug trafficking, extortion, and illegal gold mining). These processes indeed led to large drops in violence initially, with increases in economic activity (e.g. Bernal et al. 2024, Perilla et al. 2024). However, criminal successors emerged in many zones, which reintroduced violence and criminal activities (Prem et al. 2021, 2023). The takeaway is that negotiated reductions in violence can save lives in the short term, but they must be bolstered by long-term strategies – including economic reintegration of ex-combatants, community reconciliation, changes in the economic incentives to re-engage in organised criminal activities, and security efforts to prevent splinter factions from taking over. Without those, the vacuum left by a truce or demobilisation can invite new violent actors, as we have seen in both Central America’s gang truces and Colombia’s post-conflict challenges.

For full reference list see the end of the conclusion chapter.