The Presumption Deferred: Pre-trial Detention and the Fracture in Albania’s Rule-of-Law Architecture

Albania has built the institutional scaffolding the EU requires. What it has not built is the doctrine. Pre-trial detention is the clearest proof of that gap, and it sits at the centre of the country’s accession credibility.

Renada Bici  | Tirana Examiner Legal Desk

Numbers have a particular authority in legal analysis. They resist the comfortable ambiguity of political argument and force a reckoning with what is actually happening in a system, as opposed to what that system claims to be doing. Albania’s pre-trial detention figures demand that kind of reckoning. As of November 2025, between 56 and 60 percent of the country’s prison population had not been convicted of anything, a range confirmed by Albania’s own government in its submission to the United Nations Committee Against Torture. The rate of approximately 94 pre-trial detainees per 100,000 inhabitants is more than four times the European Union average. These are not the numbers of a country that has absorbed the presumption of innocence as a governing principle. They are the numbers of a country that has adopted it as formal language while continuing, in practice, to invert it.

56–60%
Share of Albania’s prison population in pre-trial detention, November 2025. Source: Albanian government submission to the UN Committee Against Torture. The EU average is approximately 20 percent.

253 days
Average length of pre-trial detention in Albania, per the Albanian government’s own data. In Germany, pre-trial detainees make up 21 percent of the prison population.

This is not an abstract critique. Albania is, by general consensus and by its own declared ambition, on the path to European Union membership, with 2030 cited as a plausible target. The accession process has produced genuine results. Negotiating clusters have been opened at a pace that has drawn genuine praise from Brussels. The institutional architecture demanded by the process, special anti-corruption structures, vetting mechanisms, a restructured judiciary, has been constructed with notable speed. EU Council President António Costa has spoken of Albania’s accession as “our common goal” and pointed to the country’s accelerated pace of reform. None of this is fabricated. The problem is structural, and it concerns what the institutions do once they exist.

There is a distinction, underappreciated in the standard enlargement discourse, between importing institutions and importing doctrine. Albania has done the former with considerable efficiency. A special prosecution structure against corruption, a special court to adjudicate the resulting cases, a constitutional vetting process to clean the bench: these are real and, in institutional design terms, serious. But institutions do not carry their doctrine with them when they are transplanted. The doctrine, the body of interpretive principle that governs how institutions exercise power, must be built separately. In Albania’s case, the doctrine of proportionality in pre-trial detention has not been built. The structure exists. The animating principle does not.

Proportionality is not a refinement of due process. It is its foundation. The European Court of Human Rights is unambiguous on this point. Pre-trial detention is permissible only where it is lawful, necessary and proportionate to a legitimate aim. It is a measure of last resort, not a default instrument of case management. The presumption of innocence is not merely a philosophical commitment about how society should regard the accused; it is an operational rule that constrains the state’s power to punish before it has proved anything. When a person spends an average of 253 days in detention before trial on charges they deny, the state has already administered a significant punishment. The trial that follows is, in functional terms, an appeal against a sentence already being served.

The proportional share of pre-trial detainees in Albania’s prison population is roughly three times the EU member state average of approximately 20 percent, and nearly double the broader European continental average of one in three. On the per-population rate, Albania’s figure of approximately 94 detainees per 100,000 inhabitants is more than four times the EU average. To place that in concrete terms: Germany, a jurisdiction with a well-developed constitutional culture of liberty and proportionality review, holds pre-trial detainees at a rate of 21 percent of its prison population. Albania holds them at nearly three times that share. That gap is not explained by a different crime environment or an unusual density of flight risks. It is explained by a different doctrine, or more precisely, by the absence of one.

At that scale, detention is no longer precautionary. It is structurally punitive. It is the state imposing deprivation of liberty as a routine consequence of serious accusation, without the judicial finding that is the only legitimate basis for doing so. The burden has been reversed. The accused must argue for liberty; the state need not argue, beyond the fact of the charge, for its removal. That reversal is the abolition of the presumption of innocence in practice, whatever the constitutional text may say.

~94 per 100,000
Albania’s pre-trial detention rate per 100,000 inhabitants, more than four times the EU average. At this scale, detention is not a precautionary measure. It is punishment without verdict.

In January 2026, the Council of Europe convened a conference in Tirana to address precisely this. Representatives from the judiciary, prosecution services and the Albanian parliament were brought together with international experts to examine what the Council’s own Director General of Human Rights and Rule of Law, Gianluca Esposito, described as the need for “strict scrutiny” of every deprivation of liberty. Former justice minister Besfort Lamallari acknowledged that the excessive use of pre-trial detention is “not just a statistical problem but a fundamental problem of human rights, the presumption of innocence and the proportionality of state intervention.” The Council of Europe stated its readiness to support reform. The diagnosis was precise, the defect was named, and the offer of remedial assistance was formally made.

Six weeks later, the European Parliament issued a draft report on Albania in which pre-trial detention did not appear. These two positions do not merely diverge. They represent a collision between two European logics operating in the same institutional space, producing incompatible conclusions about the same legal system at the same moment. The Council of Europe has formally assessed Albania’s pre-trial detention regime as a human rights failure requiring urgent remediation. The EU accession framework has elected not to register that finding. What follows from that choice is not a technical discrepancy between bodies with overlapping mandates. What follows is the EU accession process placing its endorsement on a system that a parallel European institution has formally found to be incompatible with European human rights standards. Brussels is validating outputs, the accumulation of charges, the detention of prominent accused, the closure of cases, while declining to assess the procedural legitimacy of the process generating them. That is not rigorous conditionality. It is conditionality hollowed out at its core.

The case of Tirana’s mayor, Erion Veliaj, held in pre-trial detention since early 2025 on corruption charges he denies, is not an anomaly in this system. It is the system behaving consistently with its own logic under conditions of high visibility. This is not the place for an assessment of the underlying merits. What is appropriate here is a structural observation: a three-term mayor of the national capital, charged but not convicted, was held for over a year before being granted access to thousands of case documents essential to his defence. That sequencing is precisely what a system looks like when the doctrine of proportionality is absent and when the right to mount a defence is subordinated to the operational tempo of the prosecution. Access to one’s case file is not a procedural courtesy. It is a prerequisite for the right of defence, and its delay over that timescale is a due process failure that the seniority or political identity of the accused does nothing to alter.

The case has drawn convergent criticism from voices across Albania’s political spectrum, including critics who would normally find nothing to agree on. That breadth of concern is itself a diagnostic signal. When a procedural failure produces convergent criticism from political opponents, the failure is almost certainly real and almost certainly systemic, rather than a dispute about the guilt or innocence of a particular individual. The critics are not defending Veliaj. They are identifying a pattern that the verified data already confirms and that his case has made impossible to dismiss as exceptional.

The pattern, stated plainly, is this: Albania has constructed a special anti-corruption architecture that is capable of arresting and detaining senior public figures. It has not constructed the proportionality review mechanisms, the judicial culture of detention scepticism, or the practical access-to-defence protections that would make that architecture compatible with Chapter 23 commitments on fundamental rights and fair trial standards. The result is a system that produces the output Brussels has asked for, high-profile cases and detention of prominent accused, while generating that output through a process that fails the procedural standards Brussels has also required. The outputs are certified. The process is not examined. That gap is not incidental to Albania’s accession credibility. It is its central unresolved problem.

Chapter 23 is not satisfied by the existence of institutions. It requires that those institutions operate in accordance with European standards of procedure and rights protection. A prosecution architecture that detains at more than four times the EU per-population rate, holds detainees for an average of 253 days before any conviction, and delays defence access to case materials by more than a year in high-profile proceedings, is not operating in accordance with those standards. This is not a political verdict. It is a technical one. And it is one that the EU accession framework has, thus far, chosen to avoid making.

A system that punishes before it proves cannot be certified as rule-of-law compliant, regardless of how many institutions it builds or how many clusters it closes. Albania’s accession process has reached the point where that sentence can no longer be deferred. The institutional Albania is real. The doctrinal Albania, in which detention is a last resort, proportionality is rigorously reviewed, and the right to mount a defence is not contingent on the prosecution’s convenience, does not yet exist. Certifying the first as though it were the second is not a technical oversight. It is a choice, and Brussels must be held accountable for making it.

Renada Bici is a Tirana-based lawyer practicing in civil, criminal, and administrative law. She holds a law degree from the University of Tirana and has experience in both private legal practice and public administration.