Latin America · Security · Public Policy
Is police repression the most effective way to manage violence in Latin America?
Verdict: Neither alone works. Repression without prevention usually fails long-term. Prevention without enforcement fails where gangs control territory. The evidence points to a combination — with El Salvador as the most dramatic exception to study, and Medellín as the most sustainable model.
Last updated Mar 24, 2026 · Atemporal — review quarterly
Background
Latin America has eight of the world's ten most violent countries by homicide rate. Governments across the region have repeatedly turned to 'iron fist' (mano dura) policies — mass incarceration, military deployments, emergency powers — as the primary response. El Salvador's dramatic crime reduction under President Bukele has renewed debate about whether repression works. This page examines the historical record.
The context: why this question is urgent
8
of the world's 10 most violent countries are in Latin America and the Caribbean (UNODC)
1.9
homicides per 100,000 in El Salvador in 2024 — down from 51 in 2018. Same rate as Canada. The most dramatic crime reduction on record.
CSMonitor, CSIS, 2025
80K+
people detained under El Salvador's state of emergency — 1.3% of the entire population, the highest per-capita incarceration rate in the world
↑
Most LAC nations that employed mano dura have seen an increase, not decrease, in homicide — per British Journal of Criminology (2025)
10–15%
annual violence reduction achieved by community-based prevention programmes in Latin America (IFLC programmes, cited by IADB)
Geopolitical Monitor, 2025
5+
Latin American countries now attempting to replicate the "Bukele model" — Ecuador, Honduras, Panama, Paraguay, Jamaica
CSIS, 2025
The case studies — what actually happened
✅ El Salvador — Bukele's "Plan Control Territorial" (2022–present)
Homicide rate: 51/100K (2018) → 1.9/100K (2024)
The most dramatic crime reduction in modern history. Bukele declared a state of emergency in March 2022, suspended due process rights, and ordered mass arrests targeting gang members. 80,000+ detained. A 40,000-capacity megaprison (CECOT) built. Homicides fell from 2,398 in 2019 to 114 in 2024. Bukele won re-election with 85% of the vote. The model is now being copied across the region.
Verdict: Dramatic short-term results — but at documented human rights costs and unresolved long-term questions
⚠️ Ecuador — Noboa's emergency crackdown (2024)
Homicide rate fell 17% in 2024 (Interior Ministry)
Ecuador declared an "internal armed conflict" in January 2024 after dramatic gang attacks including a live TV studio takeover. Military deployed nationally, mass arrests, state of emergency. 17% reduction in homicides in 2024 per official data. But: International Crisis Group notes uncertainty about sustainability — gangs displaced, not dismantled. Noboa re-elected on security platform.
Verdict: Early short-term results; long-term outcome uncertain
❌ Honduras, Guatemala, Brazil — mano dura without results
2000s–2020s: repeated failed iron-fist experiments
Both Honduras and Guatemala implemented aggressive police and military crackdowns repeatedly from the 2000s. Result: homicide rates increased, gang structures were strengthened (prison-based coordination), corruption deepened, and state capacity was weakened. Brazil's approach under Bolsonaro — shoot first, ask questions later — produced record police killings but no sustained reduction in violent crime. British Journal of Criminology (2025): "Most LAC nations that employed mano dura have seen an increase in homicide."
Verdict: Repression alone consistently failed when not combined with institutional reform
✅ Medellín, Colombia — transformation through institutions
Homicide rate: 380/100K (1991) → ~16/100K (2023)
Colombia's most dramatic transformation came not from repression but from a combination: targeted police intelligence against leadership structures, urban infrastructure investment in poorest neighbourhoods (escalators in comunas, libraries), education investment, and gang reintegration programmes. Medellín went from the world's most dangerous city to a celebrated urban transformation model. The security improvement was real and sustained.
Verdict: Most cited sustainable success case — combines repression with prevention and investment
⚠️ The human rights dimension — El Salvador
Human Rights Watch (2026 World Report): "Authorities have committed widespread abuses, including mass arbitrary detention, enforced disappearance, torture and ill-treatment of detainees, and due process violations."
The specific documented problems: innocent people arrested with no evidence; at least 34 cases reported of people detained while searching for missing family members; deaths in custody. Amnesty International has declared prisoners of conscience. The right to challenge detention before a court was suspended.
The central tension: the same policy that reduced homicides from 51 to 1.9 per 100,000 also put 1.3% of the population in prison without due process. For most Salvadorans, the tradeoff is acceptable — Bukele's approval rating remains above 80%. For human rights organisations, it is not.
The specific documented problems: innocent people arrested with no evidence; at least 34 cases reported of people detained while searching for missing family members; deaths in custody. Amnesty International has declared prisoners of conscience. The right to challenge detention before a court was suspended.
The central tension: the same policy that reduced homicides from 51 to 1.9 per 100,000 also put 1.3% of the population in prison without due process. For most Salvadorans, the tradeoff is acceptable — Bukele's approval rating remains above 80%. For human rights organisations, it is not.
The real debate
✅ Repression works — the case for
❌ Not the answer alone — the case against
✅ What the evidence says actually works — the synthesis
- •Targeted enforcement, not mass incarceration. Intelligence-led policing targeting leadership structures is more effective than mass arrests. Medellín and Bogotá's successful transformations used this — break networks from the top, not sweep up everyone at the bottom.
- •Prevention AND enforcement together. Social programmes alone don't work in gang-controlled territory. Enforcement alone creates a cycle of suppression and resurgence. The combination — like Medellín — produces sustainable results.
- •Institutional quality matters more than intensity. Repression through corrupt, impunity-ridden institutions makes things worse. The same level of enforcement through professional, accountable institutions produces different results.
- •Community-based programmes show 10–15% annual reductions — smaller and slower than El Salvador, but sustainable and without due process violations (IADB-cited programmes).
- •El Salvador may be a special case. Small country, gang problem highly concentrated, government had unusual political capital and support. The same model applied to Brazil, Mexico, or Colombia would not work the same way — the context is too different.
Key voices
"Most LAC nations that employed mano dura have seen an increase in homicide. Although this form of policy is widespread and popular across the region, it is rarely empirically evaluated with rigorous methods."
"Authorities have committed widespread abuses, including mass arbitrary detention, enforced disappearance, torture and ill-treatment of detainees, and due process violations."
"The hard-line approach to law enforcement risks perpetuating the lawlessness it is trying to stamp out."
"El Salvador's homicide rate dropped from 51 murders per 100,000 people in 2018 to 1.9 per 100,000 in 2024 — the same as Canada's."
Christian Science Monitor — "Iron fist security winning voters in Latin America", Feb 2026
How different regions and publications frame it
Guardian / NYT / Economist
Anglo-American centre-left
Acknowledge El Salvador's dramatic crime reduction while emphasising human rights costs and replicability concerns. Frame the debate as a tradeoff between security and civil liberties. Broadly sceptical of exporting the model to larger, more complex countries.
El País / Folha
Spain/Brazil · centre-left
Both cover the debate through the lens of their own countries' specific contexts — Spain's experience as a former colonial power, Brazil's deep inequality and police violence problem. El País covers the Bukele model with serious critical analysis. Folha documents Brazil's failed iron-fist approaches under Bolsonaro.
Latin American right/populist
Conservative populist
Milei (Argentina), Noboa (Ecuador), Bolsonaro (Brazil), Hernández (Colombia) — all ran on or implemented iron-fist security platforms. CSMonitor (Feb 2026): "iron fist promises are winning voters." The political appeal is genuine because the public is exhausted by violence.
HRW / Amnesty / academic
Human rights / research
The clearest opposition. Both document specific human rights abuses in El Salvador and Ecuador. CSIS: "tough on the weak, lenient on the powerful." British Journal of Criminology: most mano dura experiments increased homicide. These institutions argue systemic reform, not repression, is the evidence-based path.
CSIS / ICG / IADB
Think tanks / development
The most nuanced institutional view: both enforcement and prevention are necessary; the question is design. El Salvador worked partly because of specific conditions. Community prevention programmes show 10–15% annual reductions sustainably. Institutional quality matters more than enforcement intensity.
The honest bottom line
Repression alone is not the most effective approach — but it can be a necessary component of a more effective one.
The evidence from most of Latin America is clear: iron-fist approaches without addressing root causes, institutional quality, and gang network structure tend to fail or make things worse. Brazil, Honduras, and Guatemala tried this for decades.
El Salvador's case is genuinely remarkable and cannot be dismissed. The most dramatic crime reduction in modern history happened through mass incarceration and suspension of due process. For the Salvadoran public, the tradeoff has been worth it. For human rights organisations, it hasn't been.
The most important question is not "repression or prevention?" but:
— Is enforcement targeted (leadership networks) or indiscriminate (mass arrests)?
— Are institutions accountable or corrupt?
— Is there a social investment strategy alongside enforcement?
— Is the gang structure small and concentrated (El Salvador) or diffuse and embedded (Brazil, Mexico)?
Medellín's 30-year transformation — from the world's most violent city to an urban regeneration model — is the most credible evidence-based answer. It required both targeted enforcement AND massive social investment. Neither alone was sufficient.
The evidence from most of Latin America is clear: iron-fist approaches without addressing root causes, institutional quality, and gang network structure tend to fail or make things worse. Brazil, Honduras, and Guatemala tried this for decades.
El Salvador's case is genuinely remarkable and cannot be dismissed. The most dramatic crime reduction in modern history happened through mass incarceration and suspension of due process. For the Salvadoran public, the tradeoff has been worth it. For human rights organisations, it hasn't been.
The most important question is not "repression or prevention?" but:
— Is enforcement targeted (leadership networks) or indiscriminate (mass arrests)?
— Are institutions accountable or corrupt?
— Is there a social investment strategy alongside enforcement?
— Is the gang structure small and concentrated (El Salvador) or diffuse and embedded (Brazil, Mexico)?
Medellín's 30-year transformation — from the world's most violent city to an urban regeneration model — is the most credible evidence-based answer. It required both targeted enforcement AND massive social investment. Neither alone was sufficient.