In Secretary of State for Business and Trade v Mercer [2024] UKSC 12 (Mercer), the Supreme Court was confronted with whether and how to exercise its powers under sections 3 and 4 of the HRA 1998. This blog post will focus on assessing how the Supreme Court approached the judicial discretion to grant a declaration of incompatibility under section 4 HRA (s.4). Section 4 states that if the court is satisfied that the provision is incompatible with a Convention right, it may make a declaration of that incompatibility” – ‘may’ demonstrates the judicial discretion at the core of the mechanism.

Elsewhere, I have explored how courts have grappled with the exercise of their s.4 discretion and argued that s.4 could be understood as establishing a “double filter mechanism” (E. Adams, ‘Judicial Discretion and the Declaration of Incompatibility: Constitutional Considerations in Controversial Cases’ [2021] PL 311). The first filter is the judicial ‘decisional space’ which relates to whether the courts should grant a declaration, and the second filter is the political ‘decisional space’ which relates to whether the political institutions should ‘accept a declaration’. Crucially, the filters are interconnected: the courts’ anticipation of the operation of the second filter might influence the judicial approach to the first filter.

In using the double filter mechanism as an explanatory device, we can assess how courts have approached their discretion in especially controversial constitutional cases. For example, controversial human rights cases might concern contested social or moral questions or become the subject of heightened political dissonance, resistance or hostility (examples include R (Chester) v Secretary of State for Justice; McGeoch v Lord President of the Council [2013] UKSC 63 on prisoners’ voting rights (Chester), R (Nicklinson) v Ministry of Justice [2014] UKSC 38 on assisted suicide (Nicklinson), and Re Northern Ireland Human Rights Commission’s Application for Judicial Review [2018] UKSC 27 on abortion (Re NIHRC)). In these cases, broader constitutional considerations, such as qualms about the appropriate role of the courts or the wider constitutional context – for instance, tensions between courts, government and/or Parliament – can be prioritised. Essentially, judicial concerns as to possible constitutional issues at the second ‘political’ filter stage can, in some cases, constrain the exercise of the courts’ first filter discretion – leading the court to refrain from granting a declaration. This can have a deleterious effect on rights protection.

Whilst Mercer is undoubtedly a significant judgment as it involves ‘politically and socially sensitive issues’ in relation to rights concerning industrial action, when compared to the controversial cases listed above, it arguably does not meet the definition of an especially controversial constitutional case, [102]. Analysis of Mercer is, however, valuable as it facilitates further understanding as to how courts navigate their s.4 discretion. As this blog post will show, where constitutional considerations arise, depending on the case they can be framed differently and can shape how the double filter mechanism operates. It will also be shown how unclear dicta from previous judgments can perpetuate confusion regarding the judicial discretion, leading to an inconsistent approach to s.4. Overall, I will argue that Mercer further demonstrates that a clear and consistent approach to the s.4 discretion is necessary.

Brief Background

Mercer centred on section 146 of the Trade Union and Labour Relations (Consolidation) Act 1992 (TULRCA) which provides that: ‘A worker has the right not to be subjected to any detriment as an individual by any act, or any deliberate failure to act, by his employer if the act or failure takes place for the sole or main purpose of […] preventing or deterring him from taking part in the activities of an independent trade union at an appropriate time, or penalising him for doing so’. Whilst employees who have been dismissed for taking part in industrial action might claim unfair dismissal, there is no statutory protection ‘against action short of dismissal for employees, or […] workers’ who take part in industrial action, [1]. Therefore, the issue in Mercer was whether s.146 TULRCA could ‘be interpreted as extending to provide such protection’, [1]. The appellant, Fiona Mercer, had been suspended, for both planning and participating in lawful strike action. During her suspension, whilst Mercer still received pay, she was not paid for the overtime that she would usually have worked. Mercer submitted that under s.146 TULRCA she had suffered detriment as the purpose of the suspension was to prevent or deter ‘her from taking part in the activities of an independent trade union “at an appropriate time”’, [2].

In the Supreme Court, Lady Simler (with whom Lord Lloyd-Jones, Lord Hamblen, Lord Burrows and Lord Richards agreed), held that s.146 TULRCA ‘does not provide protection against detriment short of dismissal for workers taking part in industrial action’, [47]. The Court explained that ‘the words “at an appropriate time” […] exclude working time (save where the employer has consented to the activities in question) so that they limit the protection available to activities which are outside working time and/or not inconsistent with the worker’s performance of their primary duties to their employer’, [3]. Analysis of Strasbourg case law led the Court to conclude that whilst Article 11 ECHR protects the right to strike, crucially, ‘it is not a core right’ and it is not an ‘absolute’ right, [81]. However, the Court held that the total lack of ‘any legislative protection at all against any sanction short of dismissal for lawful industrial action […] does put the United Kingdom in breach of its positive obligation to secure effective enjoyment of the right to participate in a lawful strike that is protected by article 11’, [90]-[91].

Ultimately, the Supreme Court granted a declaration of incompatibility under s.4 ‘that section 146 of TULRCA is incompatible with article 11, insofar as it fails to provide any protection against sanctions, short of dismissal, intended to deter or penalise trade union members from taking part in lawful strike action organised by their trade union’, [121]. The following sections unravel the Court’s reasoning in deciding to grant a declaration.

Can a Declaration be Made?

In assessing whether a s.3 HRA interpretation was possible, and in agreement with the Court of Appeal, Lady Simler held ‘that a Convention compatible interpretation of s.146 of TULRCA is not possible’ and would constitute ‘impermissible judicial legislation rather than interpretation’, [102].

Therefore, the Supreme Court considered its powers under s.4. Notably, in the Court of Appeal, Lord Burnett had declined a declaration as it ‘would be directed not so much at section 146 as at TULRCA as a whole’ and the incompatibility concerned a more general ‘lacuna in the law’ as opposed to ‘a specific statutory provision’ ([2022] EWCA Civ 379, [82], [88]). Lord Burnett added it would be inappropriate to grant a declaration where ‘the extent of the incompatibility is unclear’, [88]. As Mallory and Tyrrell explain, in constructing the judicial discretionary space, the court must be satisfied that a declaration ‘can’ be granted (‘Discretionary Space and Declarations of Incompatibility’ (2021) 32(3) KLJ 466). This is a factor that arises prior to the actual ‘exercise’ of the s.4 judicial decisional space – the decision as to whether a declaration should be granted. Where a lacuna in the law exists, there is essentially an absence of legislation to declare incompatible and, therefore, a declaration cannot be made.

By contrast, whilst expressing ‘some sympathy for the Court of Appeal’s conclusion’, the Supreme Court held that ‘section 146 is the only provision which limits the common law in this context and has the implicit effect of legitimising sanctions short of dismissal imposed for participating in a lawful strike’, [116]. Lady Simler persuasively held that the appellant’s only ‘means of vindicating her article 11 right in the domestic courts or tribunals is blocked by the conventional interpretation given to section 146 of TULRCA. On this basis, section 146 is incompatible with article 11 of the Convention’, [117].

Judicial Discretion

Having clarified that s.146 was incompatible with article 11, the Court then considered the exercise of its discretion under s.4. It appears that the reasoning of the Court fits with the idea that a double filter mechanism operates in relation to s.4: the decision at the first filter, the judicial ‘decisional space’ whether to grant a declaration, was made taking into consideration the operation of the second filter, the political ‘decisional space’ regarding whether to ‘accept a declaration’.

For instance, the Court aptly held that s.4 is ‘a discretionary power’ which alerts Parliament to ‘incompatibility’, [112]. The Court added that a s.4 declaration ‘does not affect the validity, continuing operation, or enforcement of the legislation it concerns’ and crucially, ‘it requires no action from the executive or Parliament’ – this demonstrates the Court has regard to the second filter and the ways in which the political branches might react to the judgment, [113]. At this point in the judgment, the Supreme Court adopted what I would argue to be the “preferable” approach to understanding s.4. It is preferable as it recognises that when legislation is found incompatible and a section 3 interpretation is not possible, courts have a vital role in notifying Parliament that legislation is incompatible. Section 4 allows for the political branches to exercise their discretion at the second filter stage – it maximises the opportunity for democratic determination of the issue.

Despite this, the Court then proceeded to note that whilst s.146 of TULRCA is incompatible with article 11, s.4(2) allows for discretion and there are cases where a declaration might be inappropriate. Lady Simler cited Nicklinson in support of this understanding, that ‘the power to grant declaratory relief is anyway inherently discretionary. The possibility of not granting a declaration to enable the legislature to consider the position is by no means a novel notion’, [118]. Lady Simler also referred to Lord Kerr’s assertion in Steinfeld that ‘the circumstances in which self-restraint […] should be exercised have not been comprehensively catalogued’, [119]. Lady Simler noted that ‘the wide variety of circumstances in which the court may be called upon to make a declaration make it difficult to identify particular considerations favouring one course rather than the other’, [119].

Yet, analysis of parliamentary debates during the passage of the Human Rights Bill reveals that whilst s.4 is discretionary, the scope of the judicial discretion was in fact intended to be relatively narrow as ‘in the great majority of cases courts would … make declarations’ (HL Deb 18 November 1997, vol 583, col 546). As the scope of the judicial discretion was not set out in fuller detail, in practice, courts have adopted conflicting approaches to navigating their s.4 discretion. In some cases, courts have readily granted declarations (e.g. R (Thompson & JF) v Secretary of State for the Home Department [2009] EWCA Civ 792; [2010] 1 WLR 76 [33]), whereas in other cases courts have narrowed the discretion (e.g. R (on the application of Rusbridger) v Attorney General [2003] UKHL 38; [2004] 1 AC 357 [35]-[36] (Lord Hutton)). In particular, as discussed above, the discretion has become a source of confusion in some controversial cases (such as Chester, Nicklinson and Re NIHRC) where constitutional considerations were prioritised and used as an impediment to granting declarations, resulting in confused approaches to s.4.

Therefore, perpetuating the view in Mercer that ‘self-restraint’ might be required is potentially problematic as it leaves open the possibility for constitutional considerations, such as showing deference to Parliament, to be prioritised in a way which blocks a declaration. For instance, whilst s.4 is declaratory, there has been a high rate of political compliance with declarations of incompatibility which can inform some judges’ views that s.4 is a “strong” tool. The perception that the declaration is “strong” means that it might be considered more deferential to exercise ‘self-restraint’ and refrain from granting a declaration to, for instance, avoid creating or exacerbating tensions between institutions (see for example, R (on the application of Animal Defenders International) v Secretary of State for Culture, Media and Sport [2008] UKHL 15 [33] (Lord Bingham)). This demonstrates how, in some cases, judicial concern as to how a declaration might be perceived at the second filter stage might constrain the operation of the first filter, leading to a more reticent and restrictive approach to s.4. Courts may utilise ‘self-restraint’ as a judicial “get out” to avoid granting a declaration, despite the legislation being incompatible and despite a declaration being arguably warranted in the case. Yet, although there has been high political compliance with declarations, it is important to remember that a declaration of incompatibility remains declaratory. Therefore, I argue that the preferable approach is to frame s.4 as a tool that inherently demonstrates deference to Parliament – it respects parliamentary sovereignty. It allows for the exercise of political discretion at the second filter stage.

Despite retaining scope for reticence, ultimately, Lady Simler held that Mercer is not a case ‘where it is inappropriate to make a declaration of incompatibility’, [120]. This raises further questions as to when it would be appropriate to decline a declaration, injecting further uncertainty into the case law. Nevertheless, Lady Simler did delineate the reasons why the Court considered that a declaration was appropriate. The Court stated that the fact questions of policy are raised ‘is a reason in favour of making a declaration of incompatibility, not refusing one’ (emphasis added), [120]. Yet, conversely, in other cases, policy considerations might be utilised as a reason for declining a declaration e.g. it could be argued that it is not the court’s place to even adjudicate on contentious policy issues and the court should adopt a hands-off approach and decline a declaration (see e.g. Re NIHRC, Lord Reed [344]). Here, institutional defensiveness might be a constitutional consideration – the court’s approach might be tied up with concerns regarding how the declaration would be received politically (and might seek to avoid political reproach), especially in cases where there is a fraught constitutional context.

These divergent approaches can create uncertainty as to how the judiciary should exercise their first filter discretion. For instance, whilst the Court of Appeal primarily held that a declaration could not be made due to there being a lacuna in the law, the Court also added that it considered that it would be inappropriate to grant a declaration as ‘the legislative choices are far from being binary questions’, [36]. The Court of Appeal was evidently still mindful of potential constitutional issues that might arise at the second filter stage. Yet I argue that the fact that legislative choices are not binary should be a reason for granting a declaration. Lady Simler’s approach is preferable as it recognises that s.4 allows for democratic consideration of policy issues. Lady Simler also stated that ‘it is for Parliament to decide whether to legislate’ and Parliament might conclude that ‘section 146 is not after all the correct vehicle to remedy the problem. That too is not a reason for refusing a declaration in this case’, [120]. The Supreme Court’s judgment therefore supports my view that allowing the exercise of political discretion at the second filter stage is a reason for granting a declaration.

In explaining the reasons for granting a declaration, the Supreme Court also stated that ‘no legislation is pending or envisaged in this area, that might make it premature to make a declaration’, [120]. But what is meant by ‘pending’? What is meant by ‘envisaged’? When and why would it be ‘premature’ to grant a declaration? This constitutional consideration creates uncertainty. Arguably, Lady Simler’s approach reflects Lord Neuberger’s judgment in Nicklinson that as the legislature was ‘actively considering the issue’ this was a factor for declining a declaration in that case (Nicklinson [116]). Therefore, in some cases, in assessing the second filter, the political decisional space, the potential or actual political consideration of legislation, might be a reason for declining a declaration (depending at what point in the political and/or legislative process legislation is at).

However, just because legislation is potentially ‘pending’ or ‘envisaged’ does not guarantee that amendments to legislation and/or new legislation will be enacted. For example, in Chester the Supreme Court held, following the declaration in Smith v Scott, that s.3(1) of the Representation of the People Act 1983 was incompatible with Article 3 of Protocol 1, that the issue of prisoners’ voting rights was under ‘active consideration’ by Parliament and there was ‘no point in making any further declaration’ (Lord Mance, [39]). However, ultimately, legislation was not introduced, rather the prisoner voting clash was seemingly “resolved” by mere administrative amendments. These amendments have left the impugned legislation intact and consequently, the majority of prisoners remain disenfranchised, undermining rights protection. A further declaration in Chester could have had an important role in emphasising that the legislation was flawed and legislative change was necessary if the incompatibility was to be remedied. Courts have a crucial and legitimate role in demonstrating to the political branches that the judiciary also considers the legislation wanting.

Conclusion

Generally, Lady Simler’s approach to the s.4 judicial discretion is valuable as it highlights that s.4 is an important mechanism that facilitates democratic determination of rights issues. This might help bolster judicial confidence in subsequent cases, as it demonstrates that where constitutional considerations arise at the first filter stage (such as deference and institutional defensiveness), they should be framed as factors that support granting a declaration of incompatibility, because making a declaration of incompatibility is an example of deference to Parliament. Section 4 is an important tool as it notifies Parliament of incompatible legislation to be considered at the second filter stage.

However, the clarity of the judgment is partly clouded by maintaining space for reticence, which is defined in uncertain terms. Further, the constitutional consideration as to whether legislation is ‘pending’ or ‘envisaged’ creates more uncertainty and could be used in future to support a reticent approach. These unhelpful points in the judgment muddy the waters and represent additional layers of complexity for the judiciary to grapple with in subsequent cases. Whilst a declaration was made in Mercer, in other cases, dicta which promote reticence can be utilised as a “get out” to “justify” a decision to decline a declaration of incompatibility where warranted, potentially undermining rights protection. Instead, to ensure that the mechanism of s.4 is not unduly impaired, the declaratory nature of s.4 should be consistently emphasised – because s.4 is a tool that preserves parliamentary sovereignty.

Dr Elizabeth Adams

(Suggested citation: E. Adams, ‘The Judicial Approach to the Judicial Discretion under s.4 HRA in Secretary of State for Business and Trade v Mercer [2024] UKSC 12’, U.K. Const. L. Blog (28th October 2024) (available at https://ukconstitutionallaw.org/)