—Engin Firat, PhD Candidate, The City Law School, City St George’s, University of London
The European Court of Human Rights (ECtHR/Strasbourg Court) is increasingly reviewing not only governmental decisions but also the institutional machinery through which those decisions are produced. This post argues that these developments reflect a distinct mode of review, which I will call audit review: review directed not at the applicant, and not at the impugned measure, but at the apparatus that generates decisions. If that assessment is correct, the ECtHR is increasingly asking not whether governments reached the correct answer, but whether they have built institutions capable of reaching Convention-compliant answers in the first place.
The Court’s judgment in Ferrieri and Bonassisa v Italy (8 January 2026) shows the shift in isolation. This judgment concluded that the Italian Revenue Agency’s access to taxpayers’ banking data during a fiscal audit violated Article 8 of the Convention. It did not hold that the intrusion was disproportionate. It held that, on the Italian Court of Cassation’s own case law, ‘authorisation does not have to contain reasoning’ (§ 81), and that the measures ‘were not subjected to an effective ex post judicial or independent review’ (§ 103).
Nothing in that reasoning turns on what happened to the applicants. The ECtHR did not assess their tax affairs, or whether the Revenue Agency should have had access. It assesses the design of the institution as such: whether the power was bounded, and whether an independent entity could examine its exercise afterwards.
That is an ordinary way of reasoning in administrative law, and an unusual one in an ECtHR judgment about a tax audit. It is also, increasingly, a common approach used by the Strasbourg Court to decide cases. The obvious objection is that this is the familiar ‘procedural turn’ renamed. It is not.
What the Procedural Turn Was About
None of this is new. The requirement that the decision-making process leading to an interference be fair and afford due respect to the interests protected by the Convention dates to W. v the United Kingdom (1987), on parental access to a child in public care. In Hatton (GC, 2003), concerning night flights at Heathrow, the ECtHR said openly that it may both assess ‘the substantive merits’ of a decision and ‘scrutinise the decision-making process’ (§ 99). A substantial literature has mapped what followed, from Brems and Gerards to Arnardóttir and Huijbers.
But note its organising question. That literature asks how far the ECtHR should defer, and answers in terms of the quality of the domestic process. Process, in Popelier and Van de Heyning’s phrase, gives proportionality teeth. Even the accounts most attentive to institutional safeguards treat them as inputs to deference.
That is a question about intensity, and intensity has always varied. Craig’s survey of judicial review identifies eight distinct ways in which it does, with that survey intended to be exhaustive. Every one varies in terms of how searchingly a court looks. None varies in terms of what a court looks at.
This post therefore brackets deference altogether. Whether review is intensive or restrained is a second-order question. The prior one is what the ECtHR has chosen as the object of its scrutiny.
From Good Governance to Carbon Budgets
The best evidence for this shift is textual rather than interpretive. It turns not on how one reads the case law but on a formula the ECtHR has carried from one Convention Article to another.
In Moskal v Poland (2009), the ECtHR identified the principle explicitly: ‘good governance’ requires public authorities ‘to act in good time, in an appropriate manner and with utmost consistency’ (§ 51). Poland lost because, on discovering its own error in granting a pension, the authorities ‘failed in their duty to act in good time and in an appropriate and consistent manner’ (§ 72). The formula is older than its name, appearing unlabelled in Beyeler (GC, 2000, § 120) and consolidated in Rysovskyy (2011, § 70). Its habitat is Article 1 of Protocol No. 1: property, expropriation, welfare administration. It governs how public authorities must conduct themselves when exercising power over entitlements.
Now read Verein KlimaSeniorinnen Schweiz v Switzerland (GC, 9 April 2024). Assessing whether Switzerland had remained within its margin of appreciation under Article 8, the Grand Chamber asked whether the competent authorities had ‘had due regard to the need to’ do five things: quantify a remaining carbon budget; set intermediate sectoral targets; provide evidence of compliance; keep those targets ‘updated with due diligence, and based on the best available evidence’; and, fifth, ‘act in good time and in an appropriate and consistent manner when devising and implementing the relevant legislation and measures’ (§ 550(e)).
This phrasing is taken word for word from Moskal § 72. The significance is not linguistic coincidence but doctrinal migration.
Read as a whole, § 550 is not a rights standard. It is a specification for an administrative apparatus: a quantified baseline, disaggregated pathways, an evidentiary record, a scientific updating duty, timeliness. The violation followed from an absence: ‘in the absence of any domestic measure attempting to quantify the respondent State’s remaining carbon budget,’ the Strasbourg Court had ‘difficulty accepting that the State could be regarded as complying effectively with its regulatory obligation under Article 8’ (§ 572).
The ECtHR’s immediate inquiry was not whether Switzerland had selected the optimal emissions target, but whether it had constructed an institutional framework capable of producing Convention-compliant climate policy at all.
Three Modes Distinguished by Object
The traditional distinction between rights review and proportionality review does not fully capture the ECtHR’s case law. A third mode has emerged, and it deserves to be distinguished from the other two not with reference to the question the ECtHR asks but with regard to the object it examines:
Rights review examines the applicant’s situation. Has a Convention right been infringed?
Proportionality review examines the measure. Was this interference justified by a sufficient reason?
Audit review examines the apparatus that produced the measure. Did the State possess and maintain a decision-making system capable of generating Convention-compliant outcomes?
By ‘apparatus’ I mean the legal and institutional arrangements through which decisions are produced: the allocation of authority, the bounding of discretion, requirements of reasoning and evidence, and mechanisms of independent oversight and review.
The distinction is not merely terminological. Each allocates judicial authority differently: under the first two the ECtHR adjudicates outcomes, under the third it supervises institutions. Nor is this a new name for the vocabulary already in circulation: structural review, systemic review, process review, due diligence. Those concepts describe what the ECtHR requires of a State. Audit review describes what the ECtHR is looking at. That question is not peculiar to Strasbourg: any constitutional or administrative court reviewing the exercise of public power must decide what, exactly, it is examining.
This is an analytical framework, not a claim to have found the definitive architecture of Convention review; the modes are not watertight, and one judgment may contain more than one. Three features nonetheless suggest the distinction is real rather than verbal.
The object need not be a measure at all. In Mugemangango v Belgium (GC, 2020) what the ECtHR examined was a parliamentary chamber’s validation of its own members’ credentials. It required the deciding body to offer ‘sufficient guarantees of … impartiality’ (§ 94), discretion that ‘cannot be excessive’ (§ 109), and a procedure guaranteeing ‘a fair, objective and sufficiently reasoned decision’ (§ 115). Those are specifications of what a decision-making body must be, not statements about how intensively some measure was scrutinised.
Big Brother Watch (GC, 2021) is clearer still. The ECtHR did not hold that the United Kingdom’s regime for the bulk interception of communications was disproportionate. It held that such a regime must carry ‘end-to-end safeguards’: independent authorisation at the outset, supervision while it runs, and independent review afterwards (§ 350). Note where the proportionality assessment goes: not to Strasbourg, but to the domestic level, at each stage. The ECtHR satisfies itself that an institution exists to perform that exercise, that the institution is independent, and that it gives reasons. That is not the same enquiry conducted more intensively. It is a different enquiry.
The two enquiries can produce opposite results, and in either direction. In Kurt v Austria (GC, 2021) a child was shot dead at school by his father three days after a barring order, and Austria won. The authorities’ risk assessment had been ‘autonomous, proactive and comprehensive’ (§§ 168, 190). In Ferrieri Italy lost although nobody suggested that the intrusion into the applicants’ records was excessive; the same intrusion would have survived had the authorisation carried reasons. If process quality merely adjusted how closely the ECtHR examined outcomes, neither result would follow.
Both modes can appear in one judgment. In Savran v Denmark (GC, 2021) the ECtHR made its own assessment of the medical evidence under Article 3 and found no violation. It then found a violation of Article 8, because the domestic authorities ‘had not taken the interests duly into account, nor had they properly balanced them’ (§ 189). One applicant, one proposed removal, two objects, two modes, opposite results.
Conclusion
The claim here is taxonomic rather than a claim of discovery. That international human rights courts review the design of governmental machinery is not new. In Velásquez Rodríguez (1988), the Inter-American Court said that States must ‘organize the governmental apparatus … so that they are capable of juridically ensuring the free and full enjoyment of human rights’ (§ 166). What has not been done is to treat that activity as a mode of review in its own right, individuated by its object rather than by its intensity. Nor does audit review explain everything: it has no purchase where a right is absolute, and it still requires an applicant.
The classification matters because each mode implies a different relationship with the State. Where the object is the applicant’s situation, subsidiarity means the ECtHR should not retry the case; where it is the measure, that it should respect a domestic judgment. Where it is the apparatus, subsidiarity becomes more demanding: States must be equipped to decide well, and must be able to show it. Compliance changes too. Executing a judgment becomes less a matter of paying an award than of rebuilding an institution: in 2025 the Committee of Ministers closed its supervision of the Strand Lobben group once Norway had adjusted its decision-making processes. What Italy must now do about Ferrieri is not compensate the applicants but require its Revenue Agency to give reasons.
The familiar reassurance that Strasbourg does ‘not act as a court of fourth instance’, repeated as recently as the Chișinău Declaration of May 2026 (§ 11), therefore describes what the Strasbourg Court is leaving behind rather than what it has begun. Increasingly its attention is directed beyond individual decisions to the institutional arrangements that generate them: less an appellate tribunal reviewing isolated governmental acts, more a supervisor of the conditions under which such acts are produced. As litigation moves further towards climate frameworks, algorithmic decision-making and pandemic governance, that is the form most of the hard questions will take. The vocabulary in which the Convention system discusses its own future has not caught up.
Suggested citation: Engin Firat, Not the Decision but the Machinery: Is the ECtHR Developing a New Mode of Review? Int’l J. Const. L. Blog, Aug. 25, 2026, at: http://www.iconnectblog.com/not-the-decision-but-the-machinary-is-the-ecthr-developing-a-new-mode-of-review/