Restorative Justice in the Italian Criminal System: A Doctrinal Reading of the Cartabia Reform Three Years On

1. Introduction

There is a close relationship between crime and conflict, and the criminal trial has historically been the privileged place where the State has dealt with this relationship. As Christie observed long ago1, conflicts are not only events to be neutralised through punishment, but also goods that belong to the parties involved and that the State, by appropriating them, often subtracts from the persons who would have the greatest interest in dealing with them. The Italian penal system has built itself around the offender, the proceeding and the sanction, and only in recent decades has it begun, slowly, to look at the victim as a subject endowed with a procedural role and not only as the bearer of an interest in compensation. To date, the Italian legal system has recognised two main areas of intervention for the victim: the probative solicitation, on the one hand, and the control and impulse of the criminal action, on the other, not giving due importance to possible requests for reparation2.

In this perspective, restorative justice (hereinafter, RJ) has progressively gained ground both at the European and at the national level, as a paradigm capable of integrating the criminal proceeding rather than replacing it. The Italian legislator, with Law No. 134 of 27 September 2021 and the implementing Legislative Decree No. 150 of 10 October 2022, has finally introduced into the criminal system a comprehensive framework regulation of restorative justice, theoretically applicable to any type of crime. The reform is the first explicit attempt to regulate restorative justice in detail in Italy, overcoming the previous attempts of reduced, sectoral and partial interventions that have taken place in the past3. The matter has been further refined by Legislative Decree No. 216 of 27 December 2024, the so-called Correttivo, which has clarified some operational aspects and has rebalanced the institutional architecture of the new system4.

The present article aims to analyse the doctrinal physiognomy of this reform three years after its entry into force, in dialogue with the international and European framework and with the early Italian scholarship. After a brief methodological clarification, the analysis is structured along four lines. The first concerns the European and international background, with particular reference to Directive 2012/29/EU and to Recommendation CM/Rec(2018)8. The second focuses on the architecture of Title IV of Legislative Decree 150/2022 and on the model of complementarity between RJ and the criminal proceeding. The third addresses the most controversial points of the reform, namely the application of restorative programs at all stages of the proceeding, the case of gender-based violence, and the institutional configuration of the Centres for restorative justice. The fourth, finally, considers the doctrinal questions raised by the early implementation of the reform and the scenarios that the next years will allow to test on an empirical basis. The thesis that this paper supports, in fact, is that the reform has opened the doors to a relational paradigm but has not yet built the cultural and institutional conditions to allow it to operate without distortions.

2. Method and Scope of the Analysis

Before entering the substance of the reform, it seems useful to clarify the method and the scope of this contribution. The analysis is doctrinal in nature, that is, it works on the legal text and on the dogmatic categories that the legislator has placed at the basis of the new framework, rather than on empirical data on the application of the institutes. The object of the inquiry is in particular Title IV of Legislative Decree No. 150 of 10 October 2022 and the partial reformulation introduced by Legislative Decree No. 216 of 27 December 2024, read in the light of the European framework. The choice of a black-letter approach seems justified by the still very early stage of practical application of the Centres for restorative justice, which does not yet allow a proper empirical assessment.

The selection of sources has followed three criteria. The first concerns primary sources, namely the Italian normative texts, the Lattanzi Commission report on the reform of the criminal procedure and the speech delivered by Minister Cartabia in Sassari in June 2022, which is here treated as a document of legal-political value rather than as a mere institutional address. The second criterion concerns the secondary literature, in Italian and in English, published between 2017 and 2026 on restorative justice in the Italian system, with particular attention to the contributions of Mannozzi and Lodigiani and to the more recent commentaries on the Cartabia reform. The third criterion concerns institutional reports, among which the report of the Garante of Emilia-Romagna of 2026 has been used as a recent and territorially situated source on the early functioning of the new system.

Two European instruments are taken as benchmarks against which the Italian solution is measured. Directive 2012/29/EU is relevant here because it sets out the minimum standards on the rights, support and protection of victims of crime and contains, in Article 12, the safeguards that Member States must ensure when restorative justice services are made available. Recommendation CM/Rec(2018)8 of the Committee of Ministers of the Council of Europe is relevant because it represents, to date, the most advanced soft-law document on the matter and offers an articulated catalogue of principles, ranging from voluntariness to confidentiality, on which the quality of restorative programs may be assessed. The reading of the Italian reform in the light of these two instruments allows to verify whether the domestic regulation is in line with the European standards or whether it presents margins of friction.

Some limits of the present analysis have to be declared in a single block, so that the reader can weigh the strength of the claims that follow. The article is doctrinal in scope. It does not perform any statistical analysis, it does not test hypotheses on samples, and it does not produce empirical findings of its own. No court case-law of the merit courts on the new Centres is examined, since the available decisions are still too sparse to allow a doctrinal systematisation. The criminological and empirical literature on the effectiveness of restorative programs is not surveyed, since it belongs to a different research strand. The focus is on adult criminal justice. The juvenile system is recalled only by way of systematic contrast, as a model of earlier experimentation that has historically anticipated the choices later made by the legislator for adults. A direct consequence follows. The article cannot make claims about how the reform actually works in practice, about territorial variation in the implementation, or about the operational performance of the Centres for restorative justice. Where institutional reports or administrative documents are referenced, they are treated as documentary sources that map the early state of implementation, not as evidence of effectiveness. The conclusions reached are therefore doctrinal hypotheses, falsifiable by future empirical work.

Two further clarifications are useful before the analysis proper. On the inclusion logic of the cited literature, the choice has been made along three lines. The doctrinal commentary on the Cartabia reform was selected when it discussed Title IV provisions or the architecture of the Centres for restorative justice. The foundational works of Mannozzi and Lodigiani were retained as a doctrinal anchor on the notion of restorative justice in Italy. The international literature was included only in the form of counterpoints to the Italian model, and not as a comparative dataset. Criminological evaluation studies and empirical effectiveness research were excluded for the reasons given above. On the relation between method and argument: the four sections that follow are organised around a single thesis, namely that the reform has formally opened the doors to a relational paradigm but has left three areas of friction unresolved. Section 3 sets the European background, which is the benchmark of the analysis. Section 4 reconstructs the architecture of Title IV, which is the descriptive base of the argument. Section 5 develops the three areas of friction, each presented by stating the relevant statutory provision first and then separating the doctrinal implication from the normative reading. Section 6 places the Italian solution in a comparative frame, used as counterpoint and not as empirical proof. The scope limits stated above apply to all four sections.

3. The International and European Background

This section reconstructs, in a descriptive vein, the European supranational framework that binds the Italian legislator. The normative reading of those instruments is reserved, instead, to the discussion of the domestic regime in the sections that follow.

Restorative justice finds a binding definition in Directive 2012/29/EU establishing minimum standards on the rights, support and protection of victims of crime. Article 2(1)(d) of the Directive defines restorative justice services as "any process whereby the victim and the offender are enabled, if they freely consent, to participate actively in the resolution of matters arising from the criminal offence through the help of an impartial third party"5. The Directive does not require Member States to introduce restorative justice tout court, but it imposes on them a series of duties of information, safeguards and access whenever such services are made available. In particular, Article 12 sets out a list of safeguards, ranging from the free and informed consent of the victim to the confidentiality of the discussions, that the Member States must ensure when they provide restorative justice. The transposition of the Directive in Italy, through Legislative Decree No. 212 of 15 December 2015, was however limited and did not provide for an organic regulation of restorative programs6.

At the European level, Recommendation CM/Rec(2018)8 of the Committee of Ministers, on restorative justice in criminal matters, represents the most advanced soft-law instrument in the field. The Recommendation broadens the notion of facilitator, no longer linking it to the figure of the offender alone but to "those responsible for that harm", and it states that restorative programs "should be available at all stages of the criminal justice process". The Italian legislator, in fact, has chosen to align the new domestic regulation with this principle of full accessibility, with consequences that, as the article will discuss further on, are not without ambiguities.

A further reference is offered by the Venice Declaration of 2021, sponsored by the European Forum for Restorative Justice, which envisages restorative justice "as a culture that should permeate the criminal justice system based on the participation of the victim and the offender on a voluntary basis, as well as other affected parties"7. This trajectory of progressive institutionalisation of RJ in Europe has been the necessary condition for the Italian reform, and explains, at least in part, why Law 134/2021 expressly required the Government to draft the new regulation in compliance with international and European standards.

4. The Architecture of Title IV of Legislative Decree 150/2022

The analysis that follows operates predominantly on the descriptive plane, with the aim of reconstructing the institutes of the Title IV in their literal content. The doctrinal evaluation of those institutes is offered, in turn, in Section 5, where the three areas of friction are discussed.

Through Legislative Decree 150/2022, restorative justice has properly entered the Italian penal system8. The reform is contained in Title IV of the decree, articles 42 to 67, and it provides what the legislator itself defines as a "comprehensive framework regulation". The notion of restorative justice adopted by Italian law, at article 42(1)(a), recalls the definition of the European Recommendation: a process that "enables the victim of the crime, the person named as the perpetrator of the offence and other subjects belonging to the community to participate freely, in a consensual, active and voluntary way, in the resolution of the issues deriving from the crime, with the help of an impartial third party, adequately trained, called mediator". The reference for the first time to the community responds to an enlightened vision of restorative justice, which should stop being a private matter between the offender and the victim, to become of social interest.

The decree provides three categories of restorative programs: victim-offender mediation, the so-called reparative dialogue, and any other dialogic program between perpetrator, victim and community. Access to these programs is free and is permitted at every stage and level of the criminal proceeding, as well as in the execution phase of the sentence and even after its execution. The mediators, who must be at least two, and the staff of the restorative justice centres are bound to confidentiality of activities, deeds, statements and information, which are unusable in criminal proceedings. The meetings between the subjects involved must be held in suitable spaces and places to ensure confidentiality and independence.

At the end of the chosen program, a report is sent to the proceeding judicial authority, drawn up by the mediators and containing the description of the activities carried out and the results achieved, or the failure to carry out the program, the interruption of the same, or the failure to achieve a remedial result. However, the latter hypothesis cannot produce unfavourable effects on the person indicated as the perpetrator of the offence. The proceeding authority will evaluate the implementation of the program and any remedial outcome for the determinations of its competence, also for the purpose of assessing the seriousness of the crime.

Lastly, the new legislation made inevitable changes to the penal code, introducing the provision of the tacit remission of the lawsuit at article 152 c.p. and a common mitigating circumstance, at article 62 No. 6 c.p., for the case in which the alleged offender has participated in a restorative justice program with a restorative result9. A similar circumstance is also relevant for the purposes of applying the conditional suspension of the sentence in its short and special variants. Legislative Decree 150/2022 has today also amended some rules relating to juvenile criminal proceedings and the execution phase, in order to include explicit reference to restorative justice programs. In particular, the modification concerns the institute of the suspension of the trial and probation under article 28 of Presidential Decree No. 448/1988, which had already been used in the past to allow the inclusion of mediation, at the discretion of the judge10. Today the judge, with the order suspending the trial, can not only give the accused prescriptions aimed at repairing the consequences of the crime and promoting conciliation with the person offended by the crime, but also invite him or her to participate in a restorative justice program.

5. Critical Issues in the Early Phase of Implementation

Before turning to the three areas of friction, a methodological precision is in order. In what follows, the analysis works on two distinct planes, that are kept separated through explicit signposting. On the descriptive plane, the statutory text of Title IV of Legislative Decree 150/2022 (articles 42 to 67) and the integrative provisions of Decree 216/2024 are reconstructed in their literal and systematic content. On the normative plane, the doctrinal evaluation of those provisions is offered, with due attention to the counterarguments that have been put forward in the Italian and international scholarship. The two planes communicate, but they do not coincide, and the conclusions drawn at the second level cannot be ascribed to the first as if they were a direct reading of the legislative text.

A preliminary caveat is in order, before turning to the three areas of friction. Italian scholarship on the Cartabia reform has, on the whole, embraced the relational paradigm of justice with a degree of enthusiasm that, on this point, deserves a corrective reading. As Daly has argued in a contribution that remains, on this matter, a reference point in the international literature, the discourse on restorative justice often operates through narratives that overstate the opposition between restoration and retribution, that romanticise its supposed indigenous origins, and that promise transformative effects on persons that the empirical record does not always support11. A doctrinal analysis of the Italian reform has to take seriously, then, the cautionary force of these arguments, since the regulatory enthusiasm shown by the legislator does not, in itself, neutralise the more sober findings of the comparative literature.

5.1. The Applicability at Every Stage of the Proceeding

The relevant provision is article 129-bis of the criminal procedure code, introduced by article 7 of Decree 150/2022. It allows the proceeding judicial authority, in any state and degree of the proceeding and also ex officio, to send the parties to a Centre for restorative justice when the program may be useful to the resolution of issues arising from the fact for which the proceeding is conducted. On a strictly doctrinal reading, the provision is silent on the timing or the suitability of the program, since the judgment of usefulness is left to the proceeding judicial authority and no statutory criterion governs the exercise of that judgment. The further claim that the breadth of the provision distorts the essence of restorative justice is, by contrast, a normative reading that goes beyond the letter of the statute and rests on a substantive view of how restorative justice should operate. The two planes are kept separate in what follows.

The first critical issue concerns precisely what the legislator has presented as the most ambitious aspect of the reform, namely the applicability of restorative justice programs in each phase of the process and also in the execution phase of the sentence. As Palermo has noted12, the prediction of the applicability of restorative justice programs at every stage of the proceeding poses a problem of opportunity. In fact, it does not seem to be in the interest of the offender and, perhaps not even of the victim, to start a communicative-relational process at a time when the criminal act is temporally far away and criminally already defined. There is, on this front, a risk that the decree ends up distorting the essence of restorative justice programs, as it jeopardises the informal and voluntary nature of restorative justice and its alternative language, which could irreparably be included and homogenised with that of the criminal justice system.

This approach can be shared as a premise to support the usefulness of restorative justice interventions, but only if it is accompanied by an awareness that the time of restorative justice does not coincide with the time of criminal procedure. The preliminary investigation phase appears, from this point of view, the most fertile ground for restorative programs, since it is free from the prejudices and distrust that occur in the subsequent phase, and because the objective of these programs should also be to facilitate the release of the offender from the criminal circuit as soon as possible. The trial and post-conviction phases retain a possible space for restorative justice, but the risk that the program is reduced to a strategic move for sentencing or for the access to alternative measures is real and has to be taken seriously by judges and mediators.

A normative reading of the same provision could nonetheless support a different conclusion. The extension of restorative programs to every phase of the proceeding may be read as answering a legitimate deflationary objective, in the sense that Cavaliere has indicated as part of the broader logic of the reform13 and that Daraio has linked to the targets assumed by Italy in the framework of the National Recovery and Resilience Plan. The position is, however, hard to sustain without further distinctions. Recommendation CM/Rec(2018)8, at point 17, states that the use of restorative justice should not be motivated by efficiency considerations alone, and the Lattanzi Commission had originally proposed an instrument of cosiddetta archiviazione meritata, that is a deserved dismissal upon successful completion of the restorative path, which the legislator has not transposed14. The omission has weight: in the absence of that instrument, the deflationary effect on the proceeding remains weak, while the risk that the program is used in an instrumental manner, of the kind Mannozzi has warned against, is not neutralised by the safeguards of article 43 alone.

5.2. Restorative Justice and Gender-Based Violence

Article 44 of Decree 150/2022 makes restorative programs available without an exclusion based on the type of offence and entrusts the safeguard function to a case-by-case evaluation by the proceeding judicial authority. On a strictly doctrinal reading, two things follow from the text: the statute is open in scope, and the work of safeguarding is delegated downstream to the judicial authority. Whether that delegation is sufficient, and whether the absence of a categorical exclusion is in tension with the obligations Italy has assumed under the Istanbul Convention and with the standards of the international monitoring bodies, is a normative question. It can be answered only by reading article 44 against external benchmarks, and not from the text of article 44 alone. The two planes are kept separate in what follows.

The second area of friction, perhaps the most delicate of the entire reform, regards the application of restorative programs to crimes of gender-based and intimate partner violence. As Ollino and Pertile have recently argued15, the implementation of RJ in cases of violence against women poses significant risks, including secondary victimisation, exposure to physical and psychological harm, and potential manipulation by offenders seeking to exploit RJ processes. In this respect, the Italian reform has chosen a position of formal openness, since article 44 of the decree does not exclude any type of crime from the scope of the new regulation, and has entrusted the safeguard function to the proceeding judicial authority, which has to evaluate, case by case, the appropriateness of the program.

This solution, however, has been the subject of strong criticism by the international monitoring bodies. The CEDAW Committee and the GREVIO group have repeatedly underlined that "mandatory mediation" or any form of obligatory restorative path in cases of gender-based violence is incompatible with Italy's obligations under the Istanbul Convention16. In fact, the Italian system knows the prohibition of mediation only in the context of family proceedings, under article 6 of the Istanbul Convention as transposed in Italian law, but does not provide a similar express prohibition for the criminal proceeding. The reform, therefore, leaves on the proceeding authority a heavy burden, which requires specific training and an awareness of the dynamics of intimate violence that, to date, is not yet uniformly distributed within the judiciary. On a doctrinal reading, the failure to expressly exclude these crimes from the application of restorative programs may be regarded as one of the open questions of the reform, on which an interpretative or legislative clarification appears desirable. Whether this open question will translate into systematic problems in practice is, in turn, a matter for empirical observation in the years to come.

In normative terms, the open model of article 44 can be read as preferable to a per se exclusion, since it preserves the residual space of victim agency and avoids a paternalistic stance towards women who freely choose to engage in a restorative path. The reading is not implausible. It has been articulated, by way of comparative reference, in the literature that reads article 12 of Directive 2012/29/EU as compatible with a calibrated use of restorative justice in cases of gender-based violence, provided that the safeguards are effective. The same reading is, however, hard to sustain in light of the objections raised by the international monitoring bodies and by Ollino and Pertile, since the formal openness of articles 43 and 44 transfers to the single judge a burden of evaluation that, in the absence of mandatory training, of dedicated risk assessment tools and of an explicit prohibition of mediation in pending or recently closed situations of intimate partner violence, is unlikely to be discharged in a uniform and protective manner. Article 12 of the Directive, for this reason, demands a level of operational detail that the Italian legislator has not provided.

5.3. The Institutional Design of the Centres for Restorative Justice

The third issue concerns the institutional architecture of the new system. Articles 63 and 64 of Decree 150/2022 place the Centres for restorative justice within local authorities and entrust them with the task of ensuring essential and uniform levels of provision of services for restorative justice. On a strictly doctrinal reading, the statutory choice has two recognisable consequences: the Centres are public bodies of administrative nature, and their institutional position belongs to regional and municipal administration rather than to the judicial system. The further claim that this institutional choice carries a structural deficit is a normative reading. It rests on a comparison with what the institution of the Centre would have required, in light of the European framework and of the foreign models of dedicated public agencies, and not on the words of articles 63 and 64 alone. Read systematically, the legislative text assigns to these Centres an exclusively organisational purpose: to optimise the implementation of restorative justice programs from a management point of view, in a function that the legislator has framed as oriented to service efficiency for all the subjects involved17. The doctrinal reading suggests, on this point, an asymmetry within the design of the reform. The legislator has not coupled the institution of the Centres with a parallel infrastructure of victim-support bodies inserted within the spaces of the criminal proceeding, and an institutional choice of that kind would have been consistent with the spirit of Directive 2012/29/EU and with the central role that the European framework assigns to the victim. Whether this asymmetry will translate, in practice, into a marginalisation of victim-support functions is, however, an empirical question that this article does not, and cannot, answer.

Two documentary sources are referenced in what follows strictly as background. They are not introduced as empirical findings of the present article, nor as the result of any statistical analysis carried out by the author. On the administrative plane, the first interministerial decree on the financing of the Centres was adopted only at the end of December 2025, with an allocation of approximately 14.4 million euros and the identification of 36 Centres distributed across 34 entities, mostly municipalities, one province and four regions18. On the institutional-monitoring plane, the report of the Garante of Emilia-Romagna of 2026 is read here as a documentary source, not as a study with general external validity. It has a regional scope, which limits the reach of its observations beyond Emilia-Romagna. The report examined the restorative programs activated in adult prisons of the region between September and November 2024. It noted structural, cultural and organisational obstacles, in particular in the coordination between penitentiary institutions and social cooperatives, aggravated by normative fragmentations, administrative delays and a frequently discretionary management19. These documents, considered in isolation, signal only that the early implementation of the reform raises questions of uniformity. They do not support, on their own, any general claim on the actual functioning of the Centres across the country, since no national survey, no consolidated court case-law and no comparative empirical study is in place. The article does not advance, therefore, propositions on territorial variation, on the functioning of the Centres, or on the implementation deficit as a fact. What can be advanced, in strictly doctrinal terms, is a hypothesis: the statutory objective of essential and uniform levels of service formulated by articles 63 and 64 may stand in tension with the implementation choices that have been made so far at the administrative level. Whether this hypothesis holds is a question for future empirical research, with adequate sample design, national scope and comparative measures, and falls outside the scope of the present contribution.

On a strictly doctrinal plane, articles 63 and 64 are to be read systematically together with articles 42 and 43, which set out the general principles and the safeguards of the restorative programs, with article 67, which regulates the entry of mediators in the lists held by the Ministry of Justice, and with article 129-bis of the criminal procedure code, which governs the referral by the proceeding judicial authority. The combined reading of these provisions outlines a two-layered architecture. The first layer, that of articles 42 and 43, articulates the substantive standards of the restorative programs, including voluntariness, confidentiality and safety. The second layer, that of articles 63, 64 and 67, articulates the institutional and professional infrastructure entrusted with their delivery. Article 129-bis links the two layers by providing the procedural channel through which the proceeding judicial authority addresses the parties to the Centre. The legislative text, read in this systematic key, leaves to the executive normative sources the task of giving operational density to that infrastructure, but it does not articulate, at the level of the primary source, a mechanism that ties the substantive standards of articles 42 and 43 to the organisational standards of articles 63 and 64. The doctrinal observation is therefore a textual observation: the safeguards of the restorative program and the organisational standards of its delivery rest, in the structure of the decree, on parallel and not interlocked normative tracks.

According to the legislative text, articles 63 and 64 of Decree 150/2022 entrust the Centres for restorative justice to local authorities, with organisational and managerial functions, and article 67 establishes the conditions for the entry of mediators in the lists held by the Ministry of Justice. A reading of opposite sign could be advanced, namely that the model of organisational Centres, with no protective function towards the victim, may be sufficient if flanked by the existing victim-support services governed by Legislative Decree 212/2015 and by the network of anti-violence centres established at regional level. The reading is plausible on the wording of the statute, since the legislator has indeed designed the Centres alongside, and not in substitution of, those parallel infrastructures. It rests, however, on a presupposition that the coordination between Centres, courts, penitentiary institutions and victim-support services is, in practice, in place. Whether this presupposition is realised is a question for future empirical investigation, since the materials available to the present article do not allow any general claim on the matter. The Garante of Emilia-Romagna has noted, on a regional scale, a fragmentation of competences and the absence of stable coordination protocols, but the geographic scope of that observation does not justify, by itself, a national doctrinal generalisation.

A further critical issue, that intersects all three points discussed above, concerns the training of mediators. The decree of the Minister of Justice of 9 June 2023, no. 80, has fixed the minimum training requirements for the entry of mediators in the lists held by the Ministry, with a threshold of 240 hours including theoretical, practical and tirocinium components. The implementing guidelines drafted at the Ministry of Justice, building on Recommendation CM/Rec(2018)8, articles 22 to 24, have set out a minimum standard of 200 hours and have insisted on the necessity of a uniform preparation across the territory20. In its regulatory architecture, however, the system rests on a plurality of training pathways: university masters and ad hoc training courses, hosted by different universities and shaped by different scholarly traditions. Whether this plurality translates into substantive heterogeneity of methodology, and whether such heterogeneity affects the quality of the restorative programs activated, is an empirical question that the present article does not address. On the doctrinal plane, what can be observed is more limited: the regulatory architecture does not articulate, within the implementing decrees, a mechanism designed to ensure substantive uniformity beyond the formal threshold of training hours.

6. A Comparative Note on the Italian Solution

The comparative reading proposed here is doctrinal in nature, and the references to foreign jurisdictions are used as counterpoints, not as empirical demonstrations. The aim is to test the choices of the Italian legislator against the institutional alternatives that other systems, of comparable legal tradition, have selected over the past two decades.

In comparative terms, the Italian model is distinctive for the breadth of its scope and for the strong link it builds between the restorative program and the criminal proceeding. Other European jurisdictions have followed different paths. The Polish system, for instance, has recently looked at the Italian Cartabia reform as a model for the strengthening of mediation, with particular attention to its fiscal incentives and to the procedural integration of dispute resolution mechanisms21. The English system, on the other hand, has historically maintained a more sectoral approach, with restorative justice mainly developed in the youth justice setting and in the post-sentence phase, and with a lesser ambition of integration with the ordinary criminal proceeding22. The German experience, finally, has long been characterised by the Täter-Opfer-Ausgleich, but with a more limited normative architecture than the Italian one.

The Italian solution looks bold but also exposed. Its boldness lies in the choice of a generalised applicability and of a structured institutional system. Its exposure lies in the gap between the normative ambition and the cultural and operational conditions in which the reform has to live. As the article has argued in the previous sections, this gap is doctrinally visible in the case of gender-based violence, in the placement of the Centres within local administration, and in the regulatory architecture of mediator training. Whether each of these doctrinal tensions also produces empirical effects on the ground is, in turn, a question for future research.

A first counterpoint is offered by Belgium. The Loi du 22 juin 2005 introducing provisions on mediation in the Preliminary Title of the Code of Criminal Procedure has built a model in which restorative mediation is offered, on a voluntary basis, in every phase of the proceeding and also in the execution phase. The Belgian system, from this angle, presents a structural similarity with the Italian model. The point of divergence is the development, in operational practice and in the circulars of the Colleges of Public Prosecutors, of an explicit caution towards the use of mediation in cases of violence within close relationships, with a stratification of safeguards that the Italian reform has not articulated23.

A second counterpoint is the Norwegian Konfliktrådet, the National Mediation Service. Established with the Act on Mediation of 1991 and operative since 1992, this service is configured as a public agency that handles both criminal and civil cases, with a network of local Mediation Services organised on a national scale. The institutional position of the Norwegian Service is autonomous from the criminal justice system, although most of the cases are referred by the public prosecutor. The Italian Centres for restorative justice, by contrast, are placed at the local authorities and are entrusted with an organisational function only, without an autonomous institutional weight comparable to that of the Konfliktrådet24.

A third counterpoint is offered by New Zealand, where the Family Group Conferences (FGCs) introduced by the Children, Young Persons and Their Families Act 1989, today the Oranga Tamariki Act 1989, have been in operation for over thirty-five years in the youth justice system. The empirical literature on FGCs is rich and, at the same time, contested. Some studies have suggested a measurable diversionary effect, while more recent contributions, on this point, have warned against the consolidation of mythologies surrounding the model and have invited a more cautious reading of the data25. The Italian reform, in importing some elements of the conferencing tradition into the adult criminal system, has not yet developed an evaluative apparatus comparable to the one that has accompanied the New Zealand experience, and the lessons learned from that jurisdiction remain, to a large extent, still to be acquired. The three counterpoints, taken together, indicate a recurrent pattern. Belgium has built, in operational practice, the safeguards that the Italian decree leaves to the discretion of the single judge, especially in cases of intimate partner violence. Norway has invested in the institutional autonomy of a public mediation agency, while the Italian Centres are placed within local administration and entrusted with an organisational function only. New Zealand has accompanied a long-standing restorative model with an evaluative apparatus that the Italian system has not yet developed. Read against these three reference points, the Italian solution emerges as ambitious in scope and parsimonious in infrastructural detail. The doctrinal tensions identified in section 5 are, on a comparative plane, the visible counterpart of three institutional choices that other jurisdictions have addressed with explicit safeguards, with autonomous public agencies or with sustained empirical assessment.

7. Conclusions

The Italian regulatory intervention certainly appears relevant, albeit with some critical issues, and to be kept under observation for future application developments. This reform represents the first explicit attempt to regulate restorative justice in detail in Italy, overcoming the previous attempts of reduced, sectoral and partial interventions that have taken place in the past. Some critical issues can nonetheless already be identified, three of which stand out as particularly important for the coming years.

The first concerns the temporal articulation of restorative programs. In fact, it does not seem to be in the interest of the offender, and perhaps not even of the victim, to start a communicative-relational process at a stage of the proceeding in which the criminal event is already defined and the relational dimension of the conflict has been crystallised. The second concerns gender-based violence, where the absence of an express prohibition or limitation seems to expose the system to the risk of secondary victimisation, in tension with the obligations assumed by Italy under the Istanbul Convention. The third concerns the institutional architecture of the new Centres. On a doctrinal reading, articles 63 and 64 entrust the Centres with an organisational function, while the protective function towards the victim remains delegated to the parallel infrastructures of victim-support services. Whether this division of tasks proves sufficient in practice is a question that future empirical research will have to address; on a normative plane, it can be observed that a stronger articulation between the Centres and victim-support services within the courts would be consistent with the spirit of Directive 2012/29/EU.

In any case, this regulatory intervention could promote a cultural change, facilitating the transition from the rehabilitative, criminal-centric model to a "relational justice" model, in which the focus shifts not only to the offender, but also to the victim and to the society. This new model represents much more than the simple application of a conflict management technique, because it takes the form of a process capable of producing a new sociality. Therefore, if the Italian legislator has formally opened the doors to restorative justice, it will be necessary to see how effectively the law will be willing to lose the monopoly in the management of crime and the power of differential control of illegalities. The next decade is, properly speaking, the decade of implementation, and the doctrinal task, going forward, is to keep the conceptual coordinates of restorative justice stable against the structural inertia of the criminal-centric paradigm.

A final clarification on the limits of the present analysis seems necessary. The doctrinal approach adopted here works on the legal text and on its systematic coordinates, and does not allow inferences on the actual application of the institutes by the courts and by the Centres. The conclusions reached must therefore be read at two levels. De iure condito, they identify tensions between the statutory framework and the European benchmarks, and they propose a reconstruction of the institutes through which those tensions may be governed in interpretation. De iure condendo, they suggest specific points on which the legislator could intervene, namely an express limitation for crimes of intimate partner and gender-based violence, an articulation of the deflationary effect through an instrument of the kind envisaged by the Lattanzi Commission, and a strengthening of the institutional position of the victim within the courts. The de facto question, that is whether and how these tensions will materialise in the practice of the next years, remains open. It is falsifiable, in turn, by future empirical evidence on the activity of the Centres, on the flows of cases sent under article 129-bis of the criminal procedure code, and on the territorial distribution of the programs.