Since 1991, the Ukrainian penitentiary system has undergone many reforms – too many for such a period of time. It is entirely natural for a new team of leaders and managers of the penitentiary system to attempt to change the state of affairs, the dynamics and the use of resources in the system for the sake of a declared goal related to the resocialisation of offenders – however this category may be presented in speeches and policy documents. Yet the number of reforms carried out over 35 years shows that these declared reforms have not achieved their goal.
This paper advances three propositions. First, the previously developed periodisation of penitentiary policy [1] is supplemented by a seventh, bellicarceral modulation, which reflects the change in the nature of the system under conditions of war. Second, the repeated failure of reforms is explained by means of the theory of gradual institutional change [2, p. 15–18; 3, p. 18–31] through the category of institutional drift behind a reform façade. Third, a five-phase model of the failed penitentiary reform cycle is proposed, and the conditions under which this cycle is not triggered are identified.
The most important reform was that of 1998–1999, when the penitentiary system severed its ties with the police. There was, in fact, no other way, since the European Prison Rules require the separation of the penitentiary system from law enforcement bodies, and this applies not only to penal institutions as such but also to pre-trial detention centres. This was required by the commitments Ukraine undertook upon accession to the Council of Europe, but 1999 is now long behind us, and the number of reforms carried out since then speaks of quantity rather than quality.
Nor can one fail to mention the least successful penitentiary reform. This was the liquidation of the State Penitentiary Service and the transfer of management functions directly to the Ministry of Justice. The failure of this step was subsequently, in effect, officially acknowledged.
A name in itself does not guarantee a policy, and it is precisely this thesis that we shall defend below. Yet a name records the declared goal, and therefore a change of name is an indicator of goal conversion, though not of a change in practice. The name “State Penitentiary Service of Ukraine” reflected a social conception of the prison – in contrast to today’s “execution of sentences”, which does not merely dominate: today it is in effect the sole function of the penitentiary system.
The 2016 reform was unsuccessful for many reasons, but today it is probably more appropriate to speak of its consequences: the liquidation of an established structure, the liquidation of a philosophy, a complete lack of forward planning, and the creation of the wholly unlawful institution of “penitentiary investigators”, which in April 2018 “failed” before the Constitutional Court of Ukraine [4].
In our 2024 study, we compared the dynamics of penitentiary system funding with the dynamics of its key indicators [1]. The system’s budget grew from UAH 227 million in 1998 to UAH 8,056 million in 2021 and UAH 9,523 million in 2024. At the same time, over 2009–2019, the number of crimes in institutions per number of prisoners increased almost fourfold – against the background of a 3.5-fold increase in funding, a threefold reduction in the number of prisoners and a 1.2-fold increase in staff numbers. Mortality and suicide rates also rose: mortality – from 3.91 per 1,000 prisoners in 2009 to 9.25 in 2021; suicides – from 3 per 10,000 prisoners in 2009 to 10.1 in 2021.
A legitimately provocative question arises: can one conclude that the more we fund the prison, the more this prison kills, drives people to death or fosters the spread of crime in penal institutions? Of course, such a conclusion would be extremely simplistic. It is therefore worth naming and rejecting three alternative explanations.
The first is inflation: the nominal growth of the budget is partly absorbed by rising prices; however, per prisoner, funding over 2009–2019 increased roughly tenfold, which substantially exceeds inflation in that period. The second is the denominator effect and changes in the composition of the prison population: the reduction in the number of prisoners does indeed raise relative indicators, and those who remain in institutions serve longer sentences, are older and in poorer health, which in itself may raise mortality. However, the number of crimes in institutions also increased in absolute terms – from 422 in 2009 to 764 in 2019 and 861 in 2021 – so the growth cannot be reduced to the arithmetic of the denominator, and changes in the composition of the prison population do not explain the rise in suicides, which is above all an indicator of detention conditions and the quality of management. The third is the change in registration practice following the entry into force of the Criminal Procedure Code of Ukraine of 2012: it may have affected overall statistics, but it explains neither the sharp acceleration of growth precisely after the 2016–2017 reforms (the average crime rate in institutions rose from 3.85 per 1,000 prisoners in 2009–2014 to 7.6 in 2015–2019) nor the dynamics of mortality, which do not depend on registration practice. Hence, the answer lies in the improper management of the penitentiary system and the ineffective use of resources.
Accordingly, the theme of cyclical failure that we have announced is not a slogan. It is our statement of a state of affairs that requires attention. In speaking of cyclicality and of the classification of modulations we have developed, we consider that this classification reflects certain chronic problems of the penitentiary system, the first of which is its absolute lack of transparency.
The central analytical category is the modulation of penitentiary and probation policy – a period of policy identified not by its formal organisational structure but by its nature and essence. The term “modulation” was chosen instead of “period” or “stage” precisely because it denotes a change in the nature of policy rather than a calendar or organisational boundary. This approach makes it possible to see that a change of signboards and structures by no means always signifies a change of policy, whereas the preservation of a signboard may, on the contrary, conceal a profound change in its essence. At the same time, in peacetime the boundaries of modulations mostly coincided with organisational decisions. This does not contradict the methodology: in the Ukrainian case, reorganisations served as the main means of launching changes in the essence of policy, whereas the substantive features of each modulation – policy priorities and management style – have been defined separately from the organisational form. War became the first case in which the essence of policy changed without a change of form.
Contemporary theory of gradual institutional change distinguishes four main mechanisms: displacement, layering, drift and conversion [2, p. 15–18; 3, p. 18–31].
Institutional drift denotes a situation in which formal rules and structures remain unchanged while their actual impact and outcomes gradually shift owing to a failure (or unwillingness) to adapt the institution to changed conditions [5]. The essence of drift is that change occurs not through the adoption of decisions but through their absence: the institution “drifts with the current”, and its actual functions move away from the declared ones.
The Ukrainian penitentiary case is paradoxical and therefore theoretically interesting, since here formal structures change not rarely but far too often. The classical understanding of drift therefore requires refinement by distinguishing between levels of analysis. At the level of forms, displacement takes place: liquidations, renamings, transfers of functions. At the level of outcomes, drift takes place: the system’s actual indicators – crime, mortality, suicides, transparency – change of their own accord, outside any measurable goal. At the level of goals, conversion takes place: the declared rehabilitative goal is gradually washed away. Since 2022, layering has been added to this. The link between the level of forms and the level of outcomes is most precisely described by the category of decoupling: formal structure serves as a ceremonial façade detached from the organisation’s actual activity [6, p. 356–358]. The excessive frequency of formal reorganisations creates an appearance of constant movement, yet this movement is not directed towards measurable goals and is not subject to evaluation. Hence, the cyclicality of forms is the mechanism of the drift of substance: drift occurs not in spite of reforms but through them. It is proposed to designate this configuration as institutional drift behind a reform façade.
This drift is combined with the conversion of the declared goal. From the social conception of the prison reflected in the name “penitentiary service”, the system moved to the “execution of sentences” as its sole function, then to perceiving institutions as a costly property complex subject to a “great prison sell-off” without any economic calculations, and, under conditions of war, to perceiving the prisoner primarily as a mobilisation resource. Tellingly, the release of convicted persons for military service is officially presented as a mechanism of resocialisation through the exercise of civic responsibility. Yet this change of goal has never once been openly discussed or substantiated by statistics or scholarly research. It is precisely the absence of outcome evaluation and open data that makes the drift imperceptible and therefore self-reproducing.
Our previous study identified six modulations of penitentiary and probation policy [1, p. 19–43].
The first is the vassal-bureaucratic militarised modulation (1991–1998). The penitentiary system, as part of the Ministry of Internal Affairs (MIA), occupies, in line with Soviet tradition, a vassal and marginal position within the police system; the management style is police-militarised bureaucracy.
The second is the transitional bureaucratic militarised modulation (1998–1999). Its substance was the creation of the State Department of Ukraine for the Execution of Sentences, temporarily subordinated to the MIA, and the gradual withdrawal of the system from the political influence of the police in fulfilment of commitments to the Council of Europe. It is with this modulation that the reform we consider the most important of the entire period of independence is associated – the severing of the penitentiary system’s ties with the police.
The third is the autonomous traditional-bureaucratic modulation (1999–2015). The Department, and from 2010 the State Penitentiary Service of Ukraine (SPS), functions as a central executive body; the management style is classical militarised bureaucracy.
The fourth is the centralised theatrical modulation (2015–2017). It consisted in the liquidation of the SPS and the direct implementation of policy by the Ministry of Justice (the reform of P. Petrenko and N. Bernatska). Penitentiary policy was in effect replaced by a theatrical performance, vivid for the ordinary citizen but simplistic from an expert’s point of view, with no political accountability of its leaders for the results.
The fifth is the decentralised pseudo-managerialist modulation (2017–2020). The Administration of the State Criminal Executive Service within the Ministry of Justice marked a limited return to the SPS model; it was liquidated by Resolution of the Cabinet of Ministers of Ukraine No. 20 of 24.01.2020. This modulation is singled out largely conventionally, since its achievements and failures are a continuation of the theatrical modulation.
The sixth is the mixed pseudo-managerialist modulation (since 2020). Its features are the Department for the Execution of Criminal Sentences of the Ministry of Justice, the “great prison sell-off” without economic calculations, and paid cells in pre-trial detention centres (SIZO).
For the last three modulations, the management style is defined as unbalanced civil-militarised management. At the same time, the term “bureaucratic” in this classification carries no negative connotation: in 1991–2015 the service transformed within its own discourse and economic constraints under the influence of the Council of Europe and pressure from the ECtHR. By contrast, the “theatrical” and “pseudo-managerialist” modulations carry a negative connotation.
Thus, the theatrical modulation is characterised by eight principles: short-term goals (tactics instead of strategy); disregard for the sociological and criminological components; populist economic dominance; lack of transparency; absence of criminological forecasts; disregard for field experience, international practice and the subcultural factor; imposition of reforms without support on the ground; non-transparent renewal of personnel. The Accounting Chamber (2017) found that the goal of the reform was defined only after the decision to liquidate the SPS had been taken, and that the new structural units operated for six months without defined functions. One of the greatest failures was the introduction of “investigators of the Criminal Executive Service”: from 05.01.2017, crimes in institutions were left without authorised investigators until Decision of the Constitutional Court of Ukraine of 24 April 2018 No. 3-r/2018, which declared part six of Article 216 of the Criminal Procedure Code of Ukraine unconstitutional [4; 1, p. 41–42]. According to the Accounting Chamber (2021), the reform has not been completed, and in 2018 – the first half of 2020 about UAH 414 million was spent without proper calculations and more than UAH 60 million in bonuses was paid in breach of the law [7; 1, p. 44].
The last of the six modulations, the mixed pseudo-managerialist one, began in 2020 and formally continues to this day (the Department for the Execution of Criminal Sentences of the Ministry of Justice of Ukraine). War, however, has changed not the signboard but the very nature of the penitentiary system. Therefore, under the same methodology that places essence above organisational form, there are grounds to speak of a seventh modulation that began in February 2022 and has been layered upon the pseudo-managerialist institutional shell. In terms of the theory of institutional change, this is a classic case of layering [2, p. 15–18]: the new logic does not replace the old structure but is superimposed on it, changing its essence without changing its form. Hence, the updated periodisation is two-tiered: the six successive modulations form the institutional shell, while the seventh is a layer that determines the logic of policy within that shell.
It is proposed to call the new modulation the Bellicarceral Mobilisation Modulation. The term “bellicarceral” (from Latin bellum – war, and carcer – prison) extends the author’s conceptual triad proposed for the analysis of wartime crime: the Digibellum State describes the transformation of the state, the Omnifront Condition – the transformation of society, and the Bellicriminal Continuum – the transformation of crime and criminal justice. The bellicarceral modulation is a projection of this triad onto places of deprivation of liberty: the prison ceases to be a “rear of punishment” separated from the war and becomes simultaneously a resource of war, a target of military attack and a mirror of wartime public sentiment.
The bellicarceral mobilisation modulation is a period of penitentiary and probation policy in which the goals of the execution of sentences are subordinated to the needs of waging war, prisoners are regarded as a mobilisation resource, penal institutions become part of the omnifrontal space of risk, and the rehabilitative paradigm finally gives way to a utilitarian-mobilisational and punitive logic. The management style within this modulation may appropriately be designated as military-utilitarian resource management: the system’s success is measured not by reduced recidivism or compliance with European standards but by the number of people “released for defence” and the ability of institutions to function under shelling.
The first characteristic is mobilisation utilitarianism. In May 2024, conditional early release from serving a sentence for the purpose of military service under contract was introduced. As of February 2026, more than 12,000 convicted persons had been conditionally released early for military service under contract, while the mobilisation potential was estimated at 20–30% of the total number of convicted persons. For the first time since independence, the state regards the prisoner not as a subject of correction and resocialisation but as a defence resource. Release to the front performs the function that in peacetime ought to be performed by resocialisation programmes, thus becoming a kind of “rehabilitation through war”, devoid of any penological methodology and measurable criteria of success. Tellingly, this modulation reproduces the principal flaw of the previous ones: no public evaluation of the results has been conducted, and it is unknown how many of those released have been killed, how many have left their military units without authorisation, and how many have committed criminal offences again.
The second characteristic is decarceration without a decarceration policy. The number of prisoners continues to decline: 48.3 thousand in 2022, about 44 thousand at the time of the CPT visit in October 2023, 43.4 thousand as of 1 February 2024, about 42 thousand in 2024, and about 37 thousand convicted and remand prisoners in May 2025. However, unlike in 1999–2019, this decline is ensured not by a balanced criminal law policy but by the mobilisation of convicted persons, the overall reduction of the population as a result of loss of life, migration and the occupation of part of the territory, and the declining effectiveness of criminal justice. It is here that crime statistics become not a background but a mechanism of penitentiary processes. After a prolonged decline (321,443 criminal offences in 2021), the number of registered offences began to rise sharply: 362,636 in 2022 (rising every month to a peak in August), 475,595 in 2023, 492,479 in 2024, 608,101 in 2025 (the highest since 2013) and 397,180 in the first eight months of 2026. At the same time, the ratio of registered proceedings to proceedings with a notice of suspicion increased from 1.86 in 2021 to 2.74 in 2022, 2.51 in 2023, 2.53 in 2024, 3.66 in 2025 and 5.27 in 2026. The declining share of proceedings with a notice of suspicion means a fall in the “throughput capacity” of criminal justice and, consequently, a reduction in the inflow of convicted persons to penal institutions. At the same time, the population decline also makes relative indicators questionable: in the absence of reliable population data, the imprisonment rate per 100,000 inhabitants loses comparability with pre-war series.
The third characteristic is the prison as a target of military attack. The Omnifront Condition extends to places of deprivation of liberty as well. A prisoner deprived of freedom of movement becomes doubly vulnerable: he cannot evacuate or take shelter on his own, and therefore the state’s positive obligation to protect his life in wartime is not weakened but strengthened. The most acute dimension is the fate of institutions in the temporarily occupied territories: some convicted persons have remained under the control of the aggressor state. A separate dimension is the assignment to the Ukrainian prison service of the detention of prisoners of war, which directly embeds the penitentiary system in the architecture of war.
The fourth characteristic is the lowering of the threshold of tolerance for pain and suffering. War normalises violence and devalues the suffering of others. Already at the start of the full-scale invasion, part of society, including some well-known lawyers and scholars, approved of the public corporal punishment of looters as a manifestation of “popular vengeance”. For the penitentiary system, this means that ill-treatment is perceived less and less as a scandal and more and more as “just retribution”. CPT reports record the persistence of informal prison hierarchies, of the status of the “degraded” and of the system of “orderlies”; torture practices are becoming more concealed, and bringing perpetrators to justice more difficult.
The fifth characteristic is punitive populism and the bellicriminal expansion of the prison population. Public sentiment is shifting towards severity, and the legislature responds by criminalising new forms of conduct: collaborationist activity, aiding the aggressor state, evasion of military service, unauthorised abandonment of a military unit and desertion. Bellicriminal categories of convicted persons are being added to the traditional prison population, for whom public demand for resocialisation is virtually absent, and public discourse is built in terms of treason rather than correction.
The sixth characteristic is the maximum decline of the rehabilitative paradigm. It began to fade as early as the theatrical and pseudo-managerialist modulations, when the penitentiary system was regarded primarily as a costly property complex subject to the “great prison sell-off”. War completes this process: educational, work and psychosocial programmes give way to issues of security and survival.
The seventh characteristic is financial illusion and wartime secrecy. The financial illusion lies in the fact that the nominal growth of the budget – from UAH 8,056 million in 2021 to UAH 9,523 million in 2024 – creates the appearance of state support for the system. In reality, given cumulative inflation of almost 50% in 2022–2024, this 18% growth means a real cut in funding of about one fifth. At the same time, the absence of open data makes it impossible to answer the question of what exactly society receives for this money. War provides a convenient justification for the further closure of data, and the Soviet tradition of secrecy acquires new legitimacy through the martial law regime. The 2016 amendments to the Constitution of Ukraine deprived the prosecutor’s office of “penitentiary” oversight, yet an independent system of penitentiary inspections that would fill this gap has still not been created. This leaves the system without a “critical friend” precisely when one is needed most.
The bellicarceral mobilisation modulation is the first modulation in which penitentiary policy is determined not by the internal logic of the system itself (bureaucratic, theatrical or pseudo-managerialist) but by the external logic of war. The prison becomes at once a mobilisation reservoir, a frontline zone and an arena of public revanchism. The reduction in the number of prisoners, which under other circumstances would be considered an achievement, is in this modulation a by-product of mobilisation, demographic catastrophe and the weakening of criminal justice. The main risk is that wartime utilitarianism and punitive sentiment legitimise the abandonment of standards for the treatment of prisoners. Yet the prohibition of ill-treatment is absolute and admits no derogation, even in time of war.
Thus, the updated periodisation of 1991–2026 covers seven modulations: six successive ones – vassal-bureaucratic militarised (1991–1998), transitional bureaucratic militarised (1998–1999), autonomous traditional-bureaucratic (1999–2015), centralised theatrical (2015–2017), decentralised pseudo-managerialist (2017–2020), mixed pseudo-managerialist (since 2020) – and the bellicarceral mobilisation modulation (since 24 February 2022), layered upon the sixth.
A comparison of the modulations shows that failures are not accidental: they are reproduced according to a recurring pattern. We therefore propose a five-phase model of the failed penitentiary reform cycle.
The first phase is declaration. A new management team proclaims a “new penitentiary policy”, a “fresh start”, a “European choice” or a “reset”, appealing to historical phantasms and ideological slogans rather than to statistics and scholarly research. The goal of the reform is either not formulated at all or formulated only after the key decisions have been adopted.
The second phase is destruction. The previous management structure is liquidated together with its accumulated institutional memory, personnel and philosophy. Liquidation is presented as a result valuable in itself, and the expediency of the new construction is explained neither to society nor to experts.
The third phase is vacuum. Functions are transferred to an entity that does not yet exist or lacks the resources to perform them. The most striking example is the institution of “investigators of the Criminal Executive Service”: from 5 January 2017, the police lost the power to investigate crimes in institutions, while the new body never became operational. Another example is the structural units of the Ministry of Justice that operated for six months without defined functions. It is precisely in the vacuum phase that the reform causes the greatest harm to specific people.
The fourth phase is external correction. Since there is no internal evaluation system, mistakes are corrected from outside: by the Constitutional Court of Ukraine [4], the Accounting Chamber (2017, 2021) [7; 1, p. 39–40, 44], the ECtHR (Sklyarenko v. Ukraine), and the CPT. External correction always comes years late and remedies only the legal form, not the consequences.
The fifth phase is quiet restoration. The system returns to the previous model under a new name: the SPS was liquidated in 2015–2016, but as early as 2017 the Administration of the State Criminal Executive Service was created as a limited return to the SPS model; in 2020 it too was liquidated and replaced by a Department of the Ministry of Justice of Ukraine. Restoration is never accompanied by an acknowledgement of error, an assessment of the damage or personal political accountability. It therefore imperceptibly passes into the declaration of the next cycle.
An important feature of the cycle is the shortening duration of modulations: seven years (1991–1998), one year (1998–1999), sixteen years (1999–2015), two years (2015–2017), three years (2017–2020). The autonomous traditional-bureaucratic modulation, the longest and most stable, was at the same time the most transparent, pro-social and “politically honest”: it did not conceal its bureaucratisation and carried out profound reforms with almost no funding, with conservative staff, an enormous number of prisoners and overcrowding. The theatrical modulations, having considerably greater resources, moved from inaction to unplanned chaotic movements proclaimed a “new penitentiary policy”. The acceleration of the cycle is a symptom of drift: the more often structures change, the less time there is for accumulating data, evaluation and learning from mistakes. The sixth modulation, which has now lasted more than six years, at first sight contradicts the thesis of acceleration. However, its duration is due not to the stabilisation of policy but to the fact that war has frozen the cycle: the declaration of the next cycle has been postponed, and the internal logic of reform has been supplanted by the external logic of war.
The contrasting case is the 1998–1999 reform. It too involved the liquidation of the previous structure and the transfer of functions, yet it did not trigger the failure cycle. The comparison makes it possible to identify the conditions under which the cycle is not reproduced: the presence of an external normative anchor (commitments to the Council of Europe); a goal defined before the decision was taken (the withdrawal of the penitentiary system from the MIA had already been envisaged by the 1991 Main Directions for Reforming the Criminal Executive System); the preservation of personnel and institutional memory; and a long implementation horizon.
The central mechanism that closes the cycle is the absence of accountability. Throughout the entire period of independence, not a single official evaluation of penitentiary and probation policy has been conducted; the only exceptions are two evaluations by Council of Europe experts (1996, 2003). Without evaluation there is no memory, without memory – no learning, and without learning each new team starts “from a blank slate”, repeating the mistakes of its predecessors.
The proposed model also has predictive value. After the end of the war, the most likely development is the declaration phase of a new cycle – in the form of a “European reset” as part of fulfilling European integration commitments and a new Strategy for Reforming the Penitentiary System. Repetition of the cycle can be avoided only under the conditions identified above: the goal and measurable indicators must be formulated before decisions are adopted, the baseline state of the system must be independently assessed, institutional memory must be preserved, and an independent system of penitentiary inspections must ensure continuous external evaluation of results.
References:
1. Yagunov D. The Ukrainian Prison and Probation Policy (1991–2024): Modulations and Key Indicators. European Journal of Economics and Management. 2024. Vol. 10, No. 4. P. 17–86. DOI: https://doi.org/10.46340/eujem.2024.10.4.2.
2. Mahoney J., Thelen K. A Theory of Gradual Institutional Change. Explaining Institutional Change: Ambiguity, Agency, and Power / ed. by J. Mahoney, K. Thelen. Cambridge : Cambridge University Press, 2010. P. 1–37.
3. Streeck W., Thelen K. Introduction: Institutional Change in Advanced Political Economies. Beyond Continuity: Institutional Change in Advanced Political Economies / ed. by W. Streeck, K. Thelen. Oxford : Oxford University Press, 2005. P. 1–39.
4. Decision of the Constitutional Court of Ukraine in the case upon the constitutional petition of the Ukrainian Parliament Commissioner for Human Rights on the conformity of the provisions of part six of Article 216 of the Criminal Procedure Code of Ukraine with the Constitution of Ukraine (constitutionality), 24 April 2018, No. 3-r/2018.
5. Hacker J. S. Privatizing Risk without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States. American Political Science Review. 2004. Vol. 98, No. 2. P. 243–260.
6. Meyer J. W., Rowan B. Institutionalized Organizations: Formal Structure as Myth and Ceremony. American Journal of Sociology. 1977. Vol. 83, No. 2. P. 340–363.
7. Report on the results of the performance audit of the use of state budget funds allocated to the Ministry of Justice of Ukraine for the execution of sentences by institutions and bodies of the State Criminal Executive Service of Ukraine, and of the use and disposal of state-owned property, tangible and other state assets having financial consequences for the state budget / Accounting Chamber of Ukraine. Kyiv, 2021.