Аннотация
Background: Parole research across Council of Europe member states has long debated which model to adopt: discretionary, mandatory, or mixed. Yet, this framing misses a more consequential variable: how deeply any model is institutionally embedded. In terms of justice, insufficiently institutionalized parole risks reducing early release to a formal entitlement detached from fair decision-making, whereas robust institutional design enables it to function as a corrective mechanism against disproportionate punishment and as a practical expression of fairness in the execution of sentences. Method: The study employs a multi-method comparative legal approach combining documentary analysis, index construction, and non-parametric statistical testing. Standardized data for 2010–2024 were drawn from the World Prison Brief and SPACE I (Annual Prison Statistics of the Council of Europe — Statistiques Pénales Annuelles du Conseil de l’Europe) reports. We selected eight jurisdictions to maximise institutional variance: Kazakhstan, France, England and Wales, Sweden, Germany, the Netherlands, Poland, and Georgia. The IIP was constructed as a five-criterion composite instrument: standardized risk assessment tools, formalized decision-making body, court–probation coordination, mandatory post-release supervision, and institutionalized reintegration programmes, each scored 0–2 (composite range 0–10). Cohen's kappa (a statistical measure of agreement among independent appraisers) was used to assess inter-rater reliability. The Spearman rank correlation was used to test the association between IIP scores and average prison population rates. Formal model type (discretionary, mandatory, or mixed) served as an independent variable for comparative benchmarking. Results and Conclusions: A comparative analysis of eight Council of Europe jurisdictions for the period 2010–2024 shows that IIP indicators exhibit an inverse association (rₛ = −0.76, n = 8), exceeding the critical threshold at α = 0.05; given the small sample size, this result is considered preliminary. A robustness check on a core subgroup of four jurisdictions (France, England and Wales, Sweden, and Kazakhstan) yielded a weaker inverse association (rₛ = −0.63). These preliminary findings are consistent with an institutional model of parole effectiveness in which outcomes may be more closely associated with the organization of procedures than with the legal framework for release. For Council of Europe member states and the ongoing discussion within the CEPEJ (European Commission for the Efficiency of Justice – Commission européenne pour l’efficacité de la justice), these findings suggest that the question of reform may concern not only which model to adopt, but also the depth of its procedural implementation.