The word exception implies a return. A temporary departure from the rule, bounded by time and circumstance, after which normal life resumes. In El Salvador, the exception has been renewed fifty times. It has outlasted the crisis that justified it. It has survived a presidential reelection, a constitutional reform, and four years of uninterrupted enforcement. At some point — and no one can say exactly when — the exception stopped being an interruption and became the operating system.
On March 27, 2022, following a spike of 87 homicides in a single weekend, El Salvador's Legislative Assembly declared a state of exception. Constitutional rights were suspended: freedom of association, access to legal counsel, the right not to be detained without a warrant. The initial order allowed arrests without probable cause, extended pre-charge detention from three to fifteen days, and authorized warrantless surveillance of private communications. It was supposed to last thirty days.
It is now May 2026. The state of exception has been extended fifty consecutive times. Over 91,990 people have been arrested. The vast majority have not faced trial. In September 2025, the Legislative Assembly passed a law allowing preventive detention of up to seven years — meaning a person can be held in a cell, without seeing a judge, for the better part of a decade. At least 504 people have died in state custody.
And here is the number that makes this signal structurally distinct from a standard human rights crisis: 94% of Salvadorans approve of the president who ordered it. Not grudgingly. Not in the absence of alternatives. Enthusiastically. In the most recent CID Gallup poll, conducted in March 2026, Nayib Bukele recorded the highest approval rating of his presidency. Security and education both scored 95% satisfaction. The state of exception is not something being done to the population. By every available measure, it is something the population is choosing.
This is not a story about a dictator. It is a story about a mechanism.
To understand why a population would voluntarily surrender its constitutional rights, you have to understand what those rights were worth before the surrender.
El Salvador was, for a generation, the most violent country in the Western Hemisphere that was not at war. In 2015, the homicide rate reached 103 per 100,000 inhabitants — roughly one murder for every thousand people, every year. Entire neighborhoods were controlled by MS-13 and Barrio 18, gangs that operated as parallel states: taxing residents, controlling transit, deciding who could live where. The police were widely perceived as either complicit or powerless. Constitutional rights existed on paper while extortion operated on the street.
The state of exception collapsed the homicide rate. By 2025, El Salvador reported 1.3 homicides per 100,000 — making it, statistically, one of the safest countries in the Western Hemisphere. People who had not walked in their own neighborhoods at night for twenty years could suddenly do so. Businesses that had paid monthly extortion fees were freed from them. For millions of Salvadorans, the rights they lost under the exception were rights they had never functionally possessed.
The legal architecture that supports this has hardened over time. The initial emergency decree was a blunt instrument — warrantless arrests, extended detention, suspended counsel. But the subsequent reforms have been surgical. In March 2026, the Legislative Assembly approved a constitutional amendment permitting life imprisonment for murder, rape, and terrorism. In April 2026, Bukele signed a law allowing life sentences for individuals as young as twelve. The exception is no longer an emergency posture. It is a constitutional renovation.
The institutional response has been proportionally escalating. The Inter-American Commission on Human Rights has documented systematic patterns of arbitrary detention, torture, and deaths in custody. In its 2026 monitoring report, the IACHR stated that the conditions that originally justified the emergency "have since been overcome" and that the continued extension is incompatible with the American Convention on Human Rights. Human Rights Watch's 2026 World Report dedicated a full country chapter to El Salvador, documenting arbitrary detentions and due process violations at scale. Amnesty International described the model as "security at the expense of human rights."
None of these reports have moved the approval rating a single point.
The signal here is not the exception itself. States of exception are common — from 2020 to 2023, more than thirty were declared across Latin America alone, in countries ranging from Ecuador to Peru to Belize. Emergency powers are a constitutional feature, not a bug. The signal is what happens when the exception outlasts its justification and the population asks for it to continue anyway.
This creates a category problem for democratic theory. The standard framework for analyzing authoritarian erosion assumes a population that is either coerced or deceived. El Salvador fits neither model cleanly. The elections are contested. The media, while under pressure, operates. The approval polling is conducted by independent firms. The population is not being fooled about the trade-off — 504 deaths in custody are not hidden; they are weighed against the memory of 6,000 homicides per year and found, by a decisive majority, to be an acceptable cost.
The 504 deaths in custody crystallize the structural cost. According to Socorro Juridico Humanitario, 90% of those who died did not fit profiles associated with gang membership. They were people swept up in the dragnet — detained without evidence, held without trial, dead without accountability. Each of those deaths is both a human catastrophe and a data point that the majority has already factored into its calculation. The population is not unaware. It has decided the arithmetic works.
This is the core of the signal: not that rights are being violated, but that the violation has been democratically ratified. When a population votes — repeatedly, overwhelmingly — to suspend its own constitutional protections, the tools democracy uses to self-correct have been absorbed into the mechanism they were designed to check.
The second-order effect is already in motion: the model is being exported.
Honduras declared its own state of exception in December 2022, suspending constitutional rights in 162 neighborhoods across San Pedro Sula and Tegucigalpa. It built an island prison facility modeled explicitly on El Salvador's CECOT mega-prison. The results have been different — Al Jazeera reported in August 2025 that the Honduran emergency "deepened the crime crisis" rather than resolving it — but the political template was identical: suspend rights, deploy military, build visible detention infrastructure, campaign on the optics of toughness.
In January 2026, Guatemalan President Bernardo Arevalo declared a state of exception after gang members killed ten police officers. The language was familiar — "violent criminals who commit acts of terrorism" — and the mechanics were the same: freedoms of movement, assembly, and protest restricted. Costa Rica, Chile, and Ecuador have all had political candidates or sitting leaders invoke "Plan Bukele" as a security model.
The International IDEA's 2025 Global State of Democracy report found that 94 countries — 54% of all countries assessed — suffered a decline in at least one factor of democratic performance compared with five years earlier. El Salvador is not an outlier. It is a leading indicator. The specific mechanism — popular ratification of rights suspension — is a template that fits any country where the state has failed to deliver basic security and the population has exhausted its patience with institutional process.
Three scenarios emerge from this pattern:
Scenario one: entrenchment. The exception becomes permanent constitutional architecture. Preventive detention norms, life sentencing for minors, warrantless surveillance — all become normal law. El Salvador ceases to be a democracy in exception and becomes a security state by design. This is the trajectory the legal reforms already describe.
Scenario two: replication failure. Other countries adopt the aesthetic without the execution. Honduras's experience suggests this is likely. The model's effectiveness may depend on El Salvador's specific conditions — small geography, identifiable gang structures, concentrated political authority — that do not transfer. The copies fail, but not before suspending rights in the attempt.
Scenario three: democratic mutation. The most structurally significant outcome. A new category of governance emerges — electorally legitimate, popularly supported, constitutionally hollow. Not authoritarianism in the traditional sense. Not democracy in any functional sense. Something that uses the vocabulary of both while operating as neither. A system where elections happen, approval is genuine, and rights are voluntarily suspended — permanently.
El Salvador has not just suspended its constitution. It has demonstrated that a population can be brought to request the suspension, ratify it at the ballot box, and celebrate the result. That demonstration — replicable, televised, and carrying a 94% approval rating — is the signal that regional governance will be processing for the next decade.
The question is not whether the exception will end. The question is whether the word "exception" still means anything at all.