The Criminal Justice Legal Clinic draws attention to the publication, on 29 September 2026, of the report Espulsioni. Un'analisi della politica di rimpatrio italiana, 2015-2025 (Expulsions. An Analysis of Italian Return Policy, 2015-2025). The report was produced within the framework of the Trattenuti project by ActionAid Italia and the Department of Political Science of the University of Bari "Aldo Moro". It was coordinated by Fabrizio Coresi, with scientific coordination by Giuseppe Campesi, who are also its authors. The report is a valuable resource for the study of the administrative detention of foreign nationals and, more broadly, of the intersection between criminal law and immigration law that forms the subject of our observatory on crimmigration. Drawing on eleven years of data, it empirically assesses the effectiveness of Italian return policy. It also examines that policy's compliance with the principles of necessity and proportionality which, under both EU and domestic law, govern recourse to the coercive enforcement of removal and to administrative detention.

Content

The report begins with a methodological premise. It challenges the reliability of the indicator occasionally used in Ministry of the Interior documents, which compares returns to sea arrivals, on the grounds that it relates two populations that do not overlap. In its place, the authors propose an enforcement rate comparing returns actually carried out with the removal orders issued, namely expulsions and deferred refusals of entry ordered by the Questore under Article 10(2) of Legislative Decree No. 286/1998 (the Consolidated Immigration Act, TUI). The findings show that between 2015 and 2025, against some 302,000 removal orders, approximately 66,000 returns were carried out: just over one in five, with the ratio never exceeding one third in any given year.

Enforcement capacity also proves to be highly selective depending on the nationality of the person concerned. The authors identify three ideal-type regimes.

The first is that of multilateral corridors, concerning countries that have concluded readmission agreements with the European Union and benefit from visa liberalisation, such as Albania, Moldova, Georgia and Kosovo. In these corridors, voluntary compliance with the removal order plays a significant role. This is partly because the entry ban may be withdrawn under Article 11(3) of Directive 2008/115/EC and Article 13(14) TUI.

The second is that of bilateral corridors with a high concentration of coercion, such as those to Tunisia, Egypt and Nigeria. Enforcement in these corridors relies heavily on charter flights and on intergovernmental cooperation on identification and readmission that must be continuously renegotiated. That cooperation is therefore structurally fragile, as shown by the suspension of charter flights to Tunisia from June 2025.

The third is the regime of non-enforceability, concerning Syria, Afghanistan, Iran, Iraq and Algeria. Here, the principle of non-refoulement or inadequate consular cooperation precludes enforcement. The repeated issuance of removal orders thus ends up entrenching the irregular status of those concerned.

In every year considered, coercive enforcement accounts for between 70.6% and 92.8% of all returns. This stands in marked contrast to the priority that the Return Directive accords to voluntary departure, as interpreted by the Court of Justice in El Dridi (C-61/11 PPU). In the domestic legal order, voluntary departure is governed by Article 13(5) TUI.

Of particular interest from a legal perspective is the analysis devoted to Pre-Removal Detention Centres (Centri di permanenza per il rimpatrio, CPRs). The data refute the assumption that an expansion of detention capacity translates into an increase in returns. In 2025, admissions to CPRs reached the highest level in the series, while returns carried out from the centres fell by 45.1% compared with 2022. Over the same period, returns carried out without prior detention quadrupled.

The authors infer from this a question of legality, rather than merely one of administrative efficiency. Under Article 15 of Directive 2008/115/EC and Article 14 TUI, detention is permissible only as a measure of last resort. It must serve the enforcement of removal, may be ordered only where less coercive measures cannot be applied effectively, and must last no longer than strictly necessary. Its expansion should therefore be justified by evidence of an additional enforcement benefit, which the data fail to show.

The issue becomes all the more pressing with regard to detention ordered against nationals of countries to which removal is practically precluded. In such cases, the "reasonable prospect of removal" required by Article 15(4) of the Directive is lacking. This requirement was articulated by the Court of Justice in Kadzoev (C-357/09 PPU) and, in substantially similar terms, by the European Court of Human Rights in Mikolenko v. Estonia. In these circumstances, detention risks turning into a form of containment of irregularity lacking any autonomous legal basis.

In economic terms, documented expenditure on the detention system exceeded EUR 217 million between 2018 and 2025. Between 2023 and 2025, detention expenditure per return carried out from the centres rose by 156.5%, while the number of such returns fell by 42%.

Methodology

In the absence of a single public database, the dataset was reconstructed by integrating and cross-checking heterogeneous official sources. These include data from the Court of Auditors, the Ministry of the Interior (Department of Public Security) and the National Guarantor for the Rights of Persons Deprived of Liberty. They also include the Ministry of the Interior's Performance Reports and its Reports to Parliament on the activities of the police forces. Data on the functioning of CPRs and on costs were largely obtained through generalised civic access requests (Italy's freedom of information regime) submitted to ministries, Prefectures and Questure.

The report adopts a consistent functional classification distinguishing between forced returns, voluntary compliance with removal orders and assisted voluntary returns under Article 14-ter TUI. It also develops original indicators of detention intensity, such as the total number of detention days per return, and of detention expenditure per return. The authors stress that these are aggregate indicators rather than cohort rates. As such, they do not allow conclusions to be drawn as to the lawfulness of any individual detention measure.

The process of data collection itself becomes an object of analysis. The report documents the full or partial refusals issued by the administration in response to certain access requests. Some of the grounds invoked, such as the need to prevent the data from "being subject to interpretation" or the "political and media sensitivity" of the management of the centres, raise serious concerns as to the democratic accountability of a public policy that affects personal liberty.

The open data platform

The report forms part of a broader project that includes the platform trattenuti.actionaid.it. The platform makes data on the administrative detention of foreign nationals from 2014 to 2025 publicly accessible.

A summary section presents the key indicators. Between 2018 and 2025, 16.4% of admissions to detention centres came directly from prison. Over the same period, the share of applicants for international protection among detainees rose from 15.4% to 40.1%, peaking at 45.2% in 2024. This growth was accompanied by an increase in releases following non-validation or non-extension of detention, which in 2025 amounted to 27.9% of admissions. The authors note that the aggregate nature of the data does not allow a causal link to be established. They nonetheless consider this trend consistent with the transfer of judicial review over the detention of asylum seekers from the Justice of the Peace to the specialised sections of the ordinary courts.

The data can also be explored interactively through charts broken down by individual facility, including the centres at Gjadër in Albania, along three axes. The first covers admissions, with separate figures for those coming from prison and for asylum seekers. The second covers releases, broken down by non-validation or non-extension of detention, expiry of the maximum detention period and enforcement of return. The third covers the costs of the system, divided into operating expenses, ordinary maintenance and extraordinary maintenance. A dedicated section allows the entire dataset to be downloaded in open format, making it available to researchers, legal practitioners and students for further analysis.

The full report and press release are available on the ActionAid Italia website, while the data can be freely consulted and downloaded from the Trattenuti platform.