Absoluteness affirmed and absoluteness denied: how the Chișinău process is reviving the ‘relatively absolute’ approach to Article 3

By Professor Natasa Mavronicola

***

Strasbourg Observers is currently hosting a blog symposium on Article 3 ECHR and the forthcoming Chișinău Declaration on the ECtHR’s approach in migration cases. The symposium was convened by Natasa Mavronicola. It also features contributions from Mary Rogan, Rishika Sahgal, Eva Sevrin, and Elaine Webster.

***

Introduction

In 2013, Francesco Messineo and I wrote a case comment criticising key facets of the European Court of Human Rights’ judgment in Babar Ahmad and Others v UK. Babar Ahmad and a preceding judgment in Harkins and Edwards v UK, both judgments issued by the Fourth Section of the ECtHR on 10 April 2012 and 17 January 2012 respectively, contained reasoning which did not cohere with the earlier Grand Chamber judgments in Chahal v UK (1997) and Saadi v Italy (2008). This reasoning appeared to ‘relativise’ the application of Article 3 ECHR contrary to its absolute character. The particular reasoning we took issue with in our case comment all but disappeared from Article 3 case-law since. However, the Chișinău process has seen efforts to ‘resurrect’ it. In particular, a preliminary draft text produced on 5 March 2026 by the Steering Committee for human rights (CDDH) reproduced two strands of reasoning from these judgments:

1. that ‘treatment which might violate Article 3 because of an act or omission of a Contracting State might not attain the minimum level of severity which is required for there to be a violation of Article 3 in an expulsion or extradition case’ (Harkins and Edwards, para 129; Babar Ahmad, para 177); and
2. that, in the context of a cautious approach (meaning a reluctance) towards finding a real risk of ill-treatment, the Court, ‘save for cases involving the death penalty, […] has even more rarely found that there would be a violation of Article 3 if an applicant were to be removed to a State which had a long history of respect for democracy, human rights and the rule of law’ (Harkins and Edwards, para 131; Babar Ahmad, para 179).

The CDDH outcome document has removed the former and retained the latter of the two strands of reasoning. Yet, although abandoning the relativist reasoning of paragraphs 129 and 177 of those cases, the CDDH outcome document retains some of their relativist spirit, not least by restating a fundamentally relativist – and empirically dubious – presumption of Article 3-compliance in respect of some States, and by openly adopting balancing rhetoric. Although the outcome document underlines the absolute character of Article 3 ECHR, notions of balancing permeate the analysis and proposed wording, not least in the idea that the application of Article 3 must involve ‘avoiding unnecessary constraints on decisions to extradite, or to expel foreign nationals’ (CDDH outcome document, p.10). Crucially, the document also revisits a dictum from the 1989 judgment in Soering v UK, to the effect that ‘inherent in the whole of the Convention is a search for a fair balance between the demands of the general interest of the community and the requirements of the protection of the individual’s fundamental rights’ (Soering, para 89; CDDH outcome document, p. 3, para 7). In addition, the document refers extensively to the 2022 Grand Chamber judgment in Sanchez-Sanchez v UK to support a relativist stance.

These textual references occur against the backdrop of a statement from 27 states on 10 December 2025 demanding that Article 3 case-law be ‘constrained to the most serious issues in a manner which does not prevent State Parties from taking proportionate decisions on the expulsion of foreign criminals, or in removal or extradition cases’.

In the ensuing discussion, I unpack Harkins and Edwards and Babar Ahmad to contextualise the relativism adopted therein. I then consider what Sanchez-Sanchez v UK has contributed in relation to relativismand balancing in the application of Article 3 to the expulsion/extradition context. Finally, I contemplate what these references are being deployed towards in the context of Chișinău and warn against the distortion and dilution of Article 3 and its absolute character.

For the purposes of this blog post, I use ‘expulsion/extradition’ to refer to deportations, extraditions, and other mechanisms of removal in which non-refoulement becomes relevant.

What happened in Harkins and Edwards and Babar Ahmad?

Both Harkins and Edwards and Babar Ahmad concerned the prospect of punishment. Harkins and Edwards faced the prospect of being sentenced to whole life imprisonment without parole, while the applicants in Babar Ahmad faced the likelihood of being detained in a supermax security prison as well as being subjected to whole life imprisonment without parole.

In determining whether the applicants’ extradition would violate Article 3 ECHR, the ECtHR was invited to address the UK House of Lords judgment in Wellington v the Secretary of State for the Home Department. In Wellington, Lord Hoffmann, Baroness Hale and Lord Carswell had sought to draw a distinction between extradition cases and other non-refoulement cases, arguing that the interests of justice should be taken into account in applying Article 3 in the context of extradition. Lord Hoffmann had put it as follows: ‘the desirability of extradition is a factor to be taken into account in deciding whether the punishment likely to be imposed in the receiving state attains the “minimum level of severity” which would make it inhuman and degrading’. Accordingly, for him: ‘Punishment which counts as inhuman and degrading in the domestic context will not necessarily be so regarded when the extradition factor has been taken into account’ (Wellington, [24]).

Lord Brown and Lord Scott had disagreed, citing the absolute character of Article 3 and the unequivocal rejection of such a relativist stance by the Grand Chamber of the ECtHR some months earlier in Saadi v Italy. In Saadi, the Grand Chamber had specifically addressed a Third Party Intervention by the UK government arguing for a balancing exercise in applying Article 3 to expulsion/extradition decisions and rejected it. As Lord Scott put it: ‘[T]he standard of treatment or punishment apt to attract the adjectives “inhuman or degrading” for article 3 purposes ought to be a constant. I do not see how otherwise the article 3 prohibition regarding such treatment or punishment can be regarded as an absolute one’ (Wellington, [41]).

In both Harkins and Edwards and Babar Ahmad, the ECtHR reaffirmed Article 3’s absolute character before unravelling it. The ECtHR was emphatic that ‘the question whether there is a real risk of treatment contrary to Article 3 in another State cannot depend on the legal basis for removal to that State’ (Babar Ahmad, para 168; Harkins and Edwards, para 120). While acknowledging that it had ‘always distinguished between torture on the one hand and inhuman or degrading punishment on the other’, the ECtHR went on to clarify that, as repeatedly affirmed in relevant case law, the non-refoulement obligation did not vary based on whether the prospective treatment would amount to torture or to inhuman or degrading treatment or punishment (Babar Ahmad, paras 170-171; Harkins and Edwards, paras 122-123). The ECtHR accordingly appeared to dismiss Lord Hoffmann’s relativist approach of adjusting, according to the desirability of extradition, the substantive reach of Article 3 or the degree of risk triggering the non-refoulement obligation.

The ECtHR also rejected any notion of balancing, referring back to Chahal and underlining ‘that it was not to be inferred from paragraph 89 of Soering that there was any room for balancing the risk of ill-treatment against the reasons for expulsion in determining whether a State’s responsibility under Article 3 was engaged’ (Babar Ahmad, para 172; Harkins and Edwards, para 124; citing Chahal, para 81). Highlighting that ‘in the twenty-two years since the Soering judgment, in an Article 3 case the Court has never undertaken an examination of the proportionality of a proposed extradition or other form of removal from a Contracting State’, the ECtHR concluded that ‘the Court must be taken to have departed from the approach contemplated by paragraphs 89 and 110 of the Soering judgment’ (Babar Ahmad, para 173; Harkins and Edwards, para 125).

Yet relativism (re)surfaced a few paragraphs later:

‘[T]he Convention does not purport to be a means of requiring the Contracting States to impose Convention standards on other States … This being so, treatment which might violate Article 3 because of an act or omission of a Contracting State might not attain the minimum level of severity which is required for there to be a violation of Article 3 in an expulsion or extradition case.’ (Babar Ahmad, para 177; Harkins and Edwards, para 129)

This relativist reasoning distorts and dents the obligation under Article 3 ECHR. The application of the non-refoulement obligation borne by a Contracting State evidently does not involve a transfer of treaty obligations to a non-Contracting State. Insofar as the Court’s approach involves making the factor of geographic location determinant of the character of the prospective ill-treatment, this constitutes distortion rather than interpretation. As Francesco Messineo and I argued shortly after the judgments, the Court’s reasoning in Harkins and Edwards and Babar Ahmad could be read as seeking to (re-)incorporate balancing into Article 3’s application to extradition, effectively resurrecting paragraph 89 of Soering and covertly weighing the desirability of extradition against the imperative of Article 3 ECHR. 

The negative outcome for the applicants in these cases is not my present focus, though it is evident that the Court’s conclusions on the facts in both these cases were impacted by its relativist approach. But an important factor leading towards the Court’s conclusions was a matter-of-fact statement that revealed a normative stance: that the Court ‘has been very cautious in finding that removal from the territory of a Contracting State would be contrary to Article 3 of the Convention’ – caution here meaning reluctance – and that

‘save for cases involving the death penalty, [the Court] has even more rarely found that there would be a violation of Article 3 if an applicant were to be removed to a State which had a long history of respect for democracy, human rights and the rule of law’ (Harkins and Edwards, para 131; Babar Ahmad, para 179).

The wording adopted here reveals something akin to a presumption of Article 3-compliance where someone is facing expulsion/extradition to a state that bears the characteristics outlined, and a dubious application of this characterisation to the USA in the context of the already (at the time) well-documented US torture and rendition practices. Besides another Fourth Section judgment reference to the USA in July 2014, the ECtHR has not employed this wording in respect of any destination state since.

Absoluteness affirmed in Sanchez-Sanchez v UK?

The ECtHR’s jurisprudence appeared to have moved on after the judgments in Harkins and Edwards and Babar Ahmad (whose reasoning had been criticised, eg here and here). Prominent among subsequent jurisprudential developments was the Grand Chamber’s judgment in Vinter and others v UK, which established that, in order to be compatible with Article 3 ECHR, sentences of life imprisonment must be reducible both de jure and de facto, and that this necessitated a review mechanism providing a concrete prospect of release. In Trabelsi v Belgium in September 2014, in which the Court held that Belgium had violated Article 3 by extraditing someone to the USA to face a real risk of an irreducible life sentence, Judge Yudkivska noted with approval that the Court had left the relativist reasoning and assumption of Article 3 compliance in paragraphs 177 and 179 of Babar Ahmad behind.

In 2022, the Grand Chamber of the ECtHR pronounced on Sanchez-Sanchez v UK, yet another proposed extradition to the USA involving the prospect of a sentence of life imprisonment without parole. The Grand Chamber reaffirmed that ‘it does not consider that any distinction can be drawn between the minimum level of severity required to meet the Article 3 threshold in the domestic context and the minimum level required in the extra-territorial context’, specifically indicating the rejection of the Harkins and Edwards (and Babar Ahmad) relativist reasoning in brackets (Sanchez-Sanchez, para 99). I assume that said reasoning, which had been included in an earlier draft, has been removed from the CDDH outcome document on this basis (see CDDH outcome document, p. 8, para 25).

However, other elements of Sanchez-Sanchez are being employed in the CDDH outcome document to similar – relativist – ends. The Grand Chamber leaned into the speculative character and concomitant uncertainty of the assessment involved (Sanchez-Sanchez, para 92) to push for a cautious (i.e. reluctant) approach to findings of a prospective Article 3 violation. It buttressed this ‘cautious’ stance by stressing that a bar on extradition in such circumstances could allow someone to ‘escape with impunity’ (Sanchez-Sanchez, para 94), the latter being suggestive of a balancing posture. On this basis, the Grand Chamber in Sanchez-Sanchez employed an ‘adapted’ approach to the assessment of the prospect of life imprisonment without parole, diluting the procedural requirement for review of whole life sentences in the extradition context and raising the evidentiary burden on the applicant through a two-stage assessment to determine the real risk of an irreducible life sentence. (In doing so, it overruled the particular application of Vinter in the case of Trabelsi.) All these elements of Sanchez-Sanchez feature heavily in the CDDH outcome document.

According to Paul Arnell, the Grand Chamber at best ‘fudged’ the question of the relevance of Wellington-style relativism in its judgment. Ultimately, while the judgment in Sanchez-Sanchez may chiefly be understood as the Grand Chamber adopting a particular (chiefly evidentiary) stance, not on the principle of non-refoulement writ large, but on the specific prospect of life imprisonment without parole, it struck some significant contradictory notes that are now being employed to revive relativism in the run-up to Chișinău.

Conclusion

We have, in some ways, been here before. The CDDH document drafting process clearly encompassed a reincarnation/synthesis of the UK’s Third Party Intervention in Saadi, the Wellington majority reasoning, and the since-discredited relativist reasoning in Harkins and Edwards and Babar Ahmad. The final product has been amended to remove the most openly relativist reasoning of Harkins and Edwards and Babar Ahmad but retains a relativist slant, not least with its reference to a presumption of Article 3 compliance for certain states. Even though the Grand Chamber in Sanchez-Sanchez confirmed the Court’s rejection of the overt relativism of Harkins and Edwards (and Babar Ahmad), the reasoning employed by the Grand Chamber in adapting the assessment of the specific prospect of life imprisonment without parole is now being deployed in the CDDH outcome document towards a reassertion of balancing and relativism in the application of Article 3.

Article 3’s absoluteness is thereby simultaneously affirmed and denied in the CDDH outcome document. There is no scope for performing the acrobatics outlined in the document – affirming its absoluteness in theory while insisting on reining back its full substantive application in respect of some states and some people in practice – without undermining the integrity of Article 3.

—

By way of postscript: I have written and talked about Article 3 for over 15 years now, and have found that those who invoke relativism in the application of Article 3 sometimes seek to depict the Court’s ‘relative’ assessment of whether something violates Article 3 as being necessarily relativist, and therefore as standing in tension with notions of absoluteness. It is no surprise, in the current context, that the CDDH outcome document also emphasises the relative assessment involved in delineating the Article 3 ‘threshold’. Yet it warrants underlining that the Article 3 assessment is relative, not relativist. The assessment determining whether a particular treatment experienced by a particular individual constitutes inhuman or degrading treatment (or torture) is relative simply in the sense that it relates to a number of potentially intersecting factors shaping the character of the treatment at issue.

More extensive discussion of absoluteness, relativism, the Article 3 ‘threshold’, and the Article 3 non-refoulement obligation can be found in my monograph on Article 3, which is available open-access.