This article belongs to the debate » COVID 19 and States of Emergency
05 May 2020

Dissecting Covid-19 Derogations

Does the pandemic require derogation from human rights treaties? This question has sparked significant debate, notably spurred by Alan Greene’s provocative argument that failing to derogate would denature ordinary human rights law and leave the start and end points of the crisis unclear. Others disagree: Scheinin argues the principle of normalcy, contained in General Comment 29, should continue to apply. Only where ordinary human rights provide inadequate flexibility should derogation be considered, and even then the principle should continue to limit the derogations. Several analyses have complemented this debate, analysing the ECtHR’s practice (Molloy), the detail of the European derogations save San Marino’s (Zghibarta), the prospect of enhanced political supervision of derogation particularly following PACE Resolution 2209 (2018) (Epure), the mechanics of notification under the ECHR (Holcroft-Emmess) and the overarching Treaty frameworks (Emmons). The UN Human Rights Committee itself has weighed in, issuing a statement criticising aspects of Covid-19 derogation practice on 24 April.

This post seeks to complement this debate in two ways. First, it will summarise the state practice during this crisis, mapping the derogations to date from European, American and international human rights systems (I). Second, it will draw some tentative conclusions from this practice (II).

This partly intends to serve as a starting point for future empirical research which will no doubt exhaustively analyse how far derogation is correlated with the severity of the virus, the restrictiveness of measures and other factors (compare Hafner-Burton et al.). Such research is already underway as to national states of emergency. It will also bring into the above debate the American Convention on Human Rights (ACHR), whose derogation practice has been neglected until now. For a good overview of the Inter-American system’s approach to the crisis more broadly, see Piovesan and Morales Antoniazzi.

I. The derogations

As of 5 May 2020, and based on the information available on the relevant websites:

  • 10 of 47 states (21%) have notified derogations from the European Convention on Human Rights (ECHR);
  • 12 of 24 states (50%) have notified suspensions of guarantees under the ACHR; and
  • 14 of 173 states (8%) have notified derogations from the International Covenant on Civil and Political Rights (ICCPR). 12 of these are states which are also party to, and have derogated from, the ECHR or ACHR. One (Colombia) appears to have ended its ICCPR derogation as of 17 April; but contrast its ACHR notifications of 17.04 and 27.04.

A brief remark on the ACHR’s derogation regime is necessary. Whilst art.27 ACHR uses the word ‘suspension’ rather than derogation, the rights in question are not ‘suspended’. The regime operates indistinguishably from derogation, with particular measures needing to be strictly necessary in the particular context: C-275 J v. Peruparas 124-147; C-289 Espinoza Gonzáles para 132. This post will use the terms interchangeably. See further Emmons on the limits of derogations.

Figure 1 sets out the derogations to date in tabular form. Where the derogated provisions were not specified in the letter notifying derogation but are apparent from the attached laws, this is indicated by ‘(in Decree)’; where this does not appear, the letter itself includes the detail. For simplicity and consistency, where a notification letter specifies the end dates of relevant laws or a state of emergency, this is taken as the projected end of the derogation (even where a state says it will send a further notification when the derogation ends).

(You can download the graph here.)

(The figure was updated on 6 May 2020.)

Figure 1

* Argentina’s first letter singles out its border controls, and recitals 5-7 of Decree 313/2020 can be read as implying that it is these measures that require suspension whereas other measures (such as internal movement restrictions) can be justified as ordinary restrictions. But this is unclear, and it also includes details on its state of emergency and the relevant Decrees, leaving it unclear whether those measures require derogation.
** Bolivia’s third letter specifically refers to the ordinary restriction provisions of these articles (arts. 7(2), 16(2) and 22(3)), leaving it unclear whether the intention is to suspend or not. Similarly, Chile, the Dominican Republic and Guatemala’s letters refer to restrictions (rather than suspensions) of the relevant articles.
*** Guatemala’s second letter states that ‘the measures restricting articles 13, 15 and 22 [ACHR] continue in force’ and the attached decree refers to no new measures. However, this may be a mistake: the first and third notifications in fact only refer to articles 15 and 22, and the second letter uses identical language to the others (‘which concern the freedom of assembly and movement’). **** Honduras’ first notification’s first attached Decree refers to earlier measures in force from 16-21 March which suspended additional constitutional rights (art. 13 and further aspects of art.7 ACHR) to those suspended from 21 March onwards. The second notification includes its only decree to refer to international human rights; interestingly, it cites arts 11 and 13 ACHR (and arts 4, 17 and 19 ICCPR) in respect of restrictions on the dissemination of photographs, interviews or personal data of Coronavirus victims without their express consent.

Figure 2 sets out the most commonly-invoked derogations or suspensions. Where rights are cited but it is unclear whether the intention is to derogate, the table errs on the side of assuming derogation.

(You can download the graph here.)

(The figure was updated on 6 May 2020.)

II. Analysis

Three patterns are immediately apparent.

First, the identification of derogated provisions varied. Only five of the twelve states complied with the obligation to state which ACHR rights were suspended (art.27(3) ACHR; C-166 Zambrano Vélez et al. v Ecuadorparas 69-70); of these five, one (Argentina) was particularly ambiguous as to its scope and the remainder (Bolivia, Chile, Dominican Republic and Guatemala) stated that they were restricting, not suspending, the rights. Conversely, the majority of ECHR states did clearly identify the derogated provisions, despite this not being strictly required (art.15(3) ECHR; Şahin Alpay v. Turkey para 73). And all states identified the derogated ICCPR provisions, including those that had failed to do so in their regional notifications (Armenia, Romania, San Marino, Colombia, Ecuador, El Salvador and Peru) and despite the less binding nature of Human Rights Committee Views (van Alebeek and Nollkaemper, p.383). The importance of identifying provisions will be returned to below, but an immediate point is that this paradoxical variation requires further attention. It perhaps suggests that fear of subsequent judicial sanction plays a limited role in eliciting precision and that other factors (perhaps use of templates and/or contemporary control by the Secretary General) are more important.

Second, regional derogations frequently appeared to be conceived of as ancillary to constitutional derogation. In other words, several states seemed primarily to analyse their state of emergency and consequent restrictions under constitutional law, and assumed that such restrictions entailed derogation from international human rights law. For instance, no fewer than eight states (Armenia, Romania, Colombia, Ecuador, El Salvador, Honduras, Panama and Peru) referred to restricted constitutional rights without any express reference to ECHR or ACHR ones. Where states referred to international and constitutional right derogations, the lists were typically identical (Albania, Georgia, Argentina; compare Suriname’s attached Decree, art.6). This is also evident in states’ language: almost all notifications refer to the national state of emergency, the (typically national) rights restriction this entails and the prescribed period of these national measures (see, for instance, Estonia and El Salvador), leaving it implicit that the human rights derogation is coterminous with these. Such references to international human rights law’s requirements as are made consist of stock phrases about non-derogable rights, non-discrimination and respect for proportionality (see, for instance, El Salvador (though cf. art.2 of the attached Decree) and North Macedonia’s first notification). Chile is the exception that proves these rules: its letter alone distinguished between constitutional and international human rights, and expressly noted that derogation will end earlier than the state of emergency if possible.

On one view, this practice suggests certain states agree with Greene: they view derogation as a natural result of a state of emergency. On another, this practice is seriously problematic: it suggests a failure to conduct a separate human rights analysis to determine whether the restriction of constitutional rights in fact requires derogation from international human rights law. This violates the principle of normalcy and risks states taking extremely restrictive measures without a clear view on what human rights allows or prohibits. This brings us to the final point.

Third, there is a striking lack of consistency in practice. It is true that the majority of derogating states derogate from freedom of movement and freedom of assembly/association. Yet those states still remain a minority, with the great majority of states globally having not derogated despite imposing equally or still more drastic lockdown measures. Half of ACHR states have derogated whereas only a fifth of ECHR ones have, and only two states derogated from the ICCPR without a regional derogation. This pattern requires further exploration. The ACHR notifications are particularly vague and focussed on national constitutional rights, as outlined above, which may explain in part the ACHR’s higher derogation rate.

Ten states derogate from regional rights without derogating from the equivalent ICCPR provisions (Albania, N. Macedonia, Moldova, Serbia, Argentina, Bolivia, the Dominican Republic, Honduras, Panama and the Republic of Suriname). Some states provide details on every law that has been passed, with Colombia providing no fewer than eight notifications; others send a high-level law that empowers state authorities to take measures (e.g. Moldova) or send nothing at all (San Marino, Serbia); and most fall in between, providing some but not all detail. A number of states’ notifications are patchy (notably Colombia, El Salvador, Honduras and Serbia) despite the requirement to notify all relevant measures.

The most striking shortcoming, however, is that states provide little to no explanation as to why derogation, rather than restriction, is necessary. Only Argentina’s highly unclear notification appeared to draw a line between measures requiring suspension and measures requiring restriction. Indeed, some states’ notifications appear to conflate derogation and restriction (Bolivia, Chile, Dominican Republic, Guatemala). This is perhaps understandable: the majority of the rights in play are ordinary qualified one, and it is quite unclear what derogation adds to the ordinary proportionality test (see Simpson p.875). Difficult and uncertain legal analysis is required to determine the limits of the ordinary rights law (Coghlan, McBride). One can sympathise with Estonia’s vague statement that ‘some of these measures may involve a derogation from certain obligations’. Derogation might then be seen as a ‘backup’: states take the measures they consider necessary and derogate, granting them further leeway in case it turns out the measures were not justifiable under the ordinary tests.

Yet this undermines a key purpose of notification, which is to ‘prevent the abuse of the exceptional powers…and allo[w] other State Parties to evaluate if the scope of this suspension is consistent with the provisions of the Convention.’ (Zambrano Vélez, para 70; compare General Comment 29, para 17). Scholars have placed great emphasis on this role in the current debate, with Greene stressing the ‘transparency, additional oversight and supervision’ derogation allows, Holcroft-Emmess noting it ‘promotes transparency and adherence to the rule of law’ and Scheinin accepting that notification ‘may have the positive effect of taming emergency powers by constraining the State to articulate their emergency measures under the terms of necessity, proportionality, exigency in the situation, temporality and a commitment to human rights as a framework for legitimate emergency measures’.

If notifications fail to articulate which aspects of their measures exceed ordinary human rights law so as to require derogation, and why that derogation is necessary in the circumstances, then notification fails to achieve this. Rather than requiring states to apply a human rights lens and facilitating public debate between states as to the limits of rights and derogations here, notification collapses into the roles described above: an afterthought to constitutional analysis, a backup, or even a simple public relations device to signal that the situation is being taken seriously (for those derogating) or that draconian measures do not violate rights (for those not derogating: compare Harris, O’Boyle and Warbrick, p.810).

III. Conclusion

This leads to two thoughts by way of closing. First, should the regional courts and Treaty bodies be stricter as to derogation’s procedural requirements? Perhaps. The case-law to date has at least encouraged relatively timely notification and, in most cases, at least some level of detail as to the measures taken. See particularly Espinoza Gonzáles para 120 emphasising the importance of these steps in allowing contemporaneous national court control. But international courts will not review these notifications for many years, and at that point will need to strike a balance between procedural rigour and justice (as well as respect for the state sovereignty underpinning derogation) in the individual case. It is in particular difficult to see how courts and Treaty bodies could directly control the state’s previous legal analysis of the restriction/derogation boundary.

Second, then, this analysis perhaps reinforces arguments that stronger contemporaneous control by the regional organs, Secretaries General, other states and NGOs is required (notably Epure; and Harris, O’Boyle and Warbrick, pp.831-2). This requires scrutiny that measures are consistently reported; that the line between constitutional rights, human rights restriction and human rights derogation is clearly articulated and justified; and that the derogations in play and their relationship with the measures is clear from the notification letters themselves. Where a number of states have considered measures to require derogation, it may conversely require scrutiny of those states that have failed to derogate despite taking similar measures. The Human Rights Committee’s statement, referred to above and reminding States parties not to derogate where restriction suffices, is a first step in this process. In this way, notification of derogation might come closer to achieving the crucial aim of ensuring international, contemporary transparency.

The author wishes to thank Professor Martin Scheinin for his comments on a draft of this post.

Note from the editors: The graphs as well as the text were updated on 06 May 2020.