https://uklabourlawblog.com/2020/03/06/ ... abadjieva/
Date: March 6, 2020
Author: UK Labour Law Blog
The decision of Swift J in Royal Mail Group Ltd v Communication Workers Union (CWU)  EWHC 3200 (QB) of November last year is a notable addition to the growing list of cases in which employers have been able to obtain an interim injunction to block planned industrial action. The decision was unanimously upheld by the Court of Appeal in Royal Mail Group Ltd v Communication Workers Union  EWCA Civ 2150. The basis for the grant of the injunction was, as in a number of previous cases, a finding that the CWU had not complied with the provisions of the Trade Union Labour Relations (Consolidation) Act (TULRCA) 1992. The Act imposes a series of onerous procedural requirements with which the union must comply in order to take lawful industrial action. In other words, the Act confers immunity on unions from liability for economic torts committed in the organisation and conduct of strike action, but only where the requirements of the Act are met. This is in contrast to countries which have a constitutionally entrenched positive right to take collective action, such as France, Italy, Sweden, Greece and many other European states.
Here I provide a brief outline of the facts of the case, and comment on certain aspects of the reasoning of the judges at both instances which I consider to be problematic. I conclude that, even leaving aside criticism of the reasoning, the outcome of the case exposes the deficiencies of adopting a system of immunities to protect workers’ fundamental right to take collective action, rather than a positive right.
Facts of the case
Between September and October 2019 the CWU ran a ballot of its members in relation to industrial action, as required by s 226 TULRCA 1992. The union sought to undertake industrial action because it considered that Royal Mail had failed to apply the terms of an agreement concluded in 2018 and acted in ways which did not comply with that agreement. On a turnout of about 75%, 97.1% of the votes cast (110,292 CWU members) favoured industrial action. As such, an overwhelming majority of around 73 % of all members were supportive of the planned strike. Whilst notice of the strike had not yet been given, there was a possibility that the union could call it in the period before Christmas, which could cause serious disturbances to the heavy postal traffic in that period. At the time of the HC judgment, but not the CA judgment, there was a possibility that it could affect postal votes in the general election on 12 December.
Royal Mail challenged the validity of the ballot on the basis that the union had advised its members to intercept their own ballot papers whilst they were still at the Delivery Office of Royal Mail from where they would be delivered to their home addresses, and to fill them out with a ‘YES’ vote there before posting them back. As evidence, Royal Mail presented two Facebook posts, transcript of an address to the union members at one Delivery Office, and a number of photographs and videos showing union members filling out or posting their ballots, at times in ‘mass postings.’ Videos taken at only one office (Swansea) showed members filling out their papers in the canteen. Royal Mail argued that this amounted to a breach of a number of the requirements of s 230 TULRCA, namely of s 230(1)(a), which requires that every person entitled to vote must ‘be allowed to vote without interference from, or constraint imposed by, the union or any of its members, officials or employees;’ of s 230(2)(a), which requires that ‘so far as is reasonably practicable, every person who is entitled to vote in the ballot must … have a voting paper sent to him by post at his home address or any other address which he has requested the trade union in writing to treat as his postal address’; and of s 230(4)(a)which requires that ‘so far as reasonably practicable, those voting do so in secret.’
In the High Court, Swift J concluded that the actions of the union were likely an interference with the ballot process contrary to s 230(1): advising its members to pick up their own ballot papers and fill them out at work ‘was a form of subversion or undermining of the ballot process anticipated by section 230’ (para 23). This is because the premise of s 230 ‘is that each voter should receive his voting paper at his home address so as to have the chance to decide whether to vote or not, and if so, how to vote, away from the day-to-day routine and environment of his workplace’ (para 24). The CA decision substantially agreed with the view of Swift J. Considering this to be a separate point, the CA also agreed with Swift J that there was a failure to comply with s 230(2) because that provision ‘contemplates that ballot papers will be delivered to voters’ home addresses in the ordinary course of post’ (Males LJ, para 69). Swift J also ruled that the Swansea office videos showing members filling out their ballots was likely a violation of s 230(4)(a), but the main CA judgment delivered by Males LJ acknowledged that a potential breach of s 230(4)(a) might have been disregarded as de minimis. My commentary therefore focuses on the reasoning of the judges regarding s 230(1)(a) and s 230(2)(a).
There is much scope for commentary and criticism of the High Court and Court of Appeal judgments, particularly the former. However, here I limit myself to four observations. I argue that (1) the judges at HC and CA level construed the purpose of s 230 too narrowly; (2) it was not evident that the union’s actions amounted to interference and the judges should, in these circumstances, have given the union the benefit of the doubt; (3) both the HC and CA judgments pay insufficient attention to the human rights dimension of the case; and (4) the outcome of the case is that an employer is able to rely on provisions intended to protect workers to prevent industrial action which is widely supported by those workers, in a way which subverts the democratic rationale of the legislation.
a) The purpose of s 230
Key to the decision of Swift J was the conclusion that the premise of s 230 is that voters should receive their ballot papers at their home address (para 22). On appeal, Males LJ agreed with this view, arguing that ‘section 230 undoubtedly contemplates, and in my judgment requires, that … the ballot paper will be received by the employee at his or her home address’ (para 53). This was the basis for the conclusion that the actions of the union subverted the purpose of the provision, in a way contrary to s 230(1)(a), and s 230(2)(a).
It is questionable whether this is what s 230 ‘undoubtedly’ contemplates, in light of the wording of that section. The starting point for the analysis of Swift J and Males LJ—though this is not made explicit—is presumably the wording of s 230(2)(a), the provision imposing the obligation to post ballot papers. The goal of the provision must be that ballot papers are not merely sent but also received by individuals, so that they are able to vote. Whether they must necessarily be received at home is another matter. S 230(2)(a) envisages that ballot papers could be sent to ‘any other address which [the union member] has requested the trade union in writing to treat as his postal address.’ In other words, a member can choose to receive and consider their ballot paper at a place other than home, including the workplace. Both Swift J and Males LJ draw on the fact that an earlier amendment to the provision had removed the option of workplace ballots (para 22 and para 60 respectively). However, whilst the removal of workplace ballots indicates that these were not desirable—since there were concerns about abuses (Males LJ, para 60)—it does not necessarily follow that Parliament positively intended that postal ballots must only be received at home.
The proviso in s 230(2) shows that it was contemplated that workers can in principle receive their papers at another place of their choosing. A conclusion that ballot papers must necessarily be received at home therefore seems too narrow a construction of the goal of the provision. Such a construction does not follow from an ordinary reading of that provision, which is how the Act should be interpreted according to Elias LJ in RMT v Serco Ltd and ASLEF v London & Birmingham Railway  EWCA Civ 226 (para 9).
It is, however, not clear whether a broader construction—that members should receive ballot papers at home, or another place of their choosing—would have led to a different ruling. Members had not, after all, indicated an alternative address ‘in writing.’ However, at least such a construction would be a more accurate reflection of the wording of the statute, and a basis for a more nuanced discussion of whether the actions of the union complied with s 230(2)(a) and s 230(1)(a). It would be relevant to such a discussion that letters were addressed to members’ home addresses and in fact received by them; that taking their own letters from Delivery Offices seems to have been an ‘existing practice of RMG employees’ (Males LJ, para 51), and so they might have taken their ballot papers anyway; and that letters were intercepted by members themselves, who chose to take their own ballot paper before it reached their home. It is at least arguable that this did not go against the legislative purpose, and so complies with s 230(1)(a), and might comply with s 230(2)(a) on a generous reading of that provision.
The fact that this was a unique case which is unlikely to occur in another workplace, and that the strike was so widely supported, would support an argument for a somewhat less stringent approach in the application of the provisions. To this might be added the dicta of Smith LJ in British Airways Plc v Unite the Union (No 2)  EWCA Civ 669 that the Act is not intended ‘to create a series of traps or hurdles for the Union to negotiate‘ (para 152) – this would suggest that the union should not be penalised for assuming that the conduct described above complies with s 230(2)(a), given the plain wording of that provision.
b) The conclusion that there was interference contrary to s 230(1)(a)
Swift J is quick to arrive at the conclusion that encouraging members to take their own papers amounted to interference contrary to s 230(1)(a). There is not much by way of justification apart from the argument that what the CWU did subverted the process envisaged by s 230. The CA offers a somewhat fuller and clearer justification of this conclusion. Males LJ first considers the meaning of ‘interference’ in s 230(1)(a), and then assesses whether the union’s actions amounted to such interference.
Males LJ interpreted the word ‘interference’ in a straightforward way, as meaning simply ‘conduct … which has the effect of preventing or hindering the ordinary course of events with which the section is concerned’ (para 49; my emphasis). Conduct need not be ‘improper’, nor amount to ‘intimidation, coercion, fraud or the like.’ Sir Patrick Elias added at para 88 that this reading of ‘interference’—though it might be broader than the usual understanding of ‘interference in voting’—is appropriate to the statutory context of s 230. Whether one agrees with the point that the provision envisages that ballot papers are received at home, or with the broader reading I offer above which includes another place of the person’s choosing, s 230 certainly seeks to protect the ability of workers to have an opportunity to cast their vote freely. This includes choosing whether, how (i.e. whether to vote in favour or against the proposed action), where and when to vote. In light of this worker-protective objective—and the fact that intimidation, coercion and so on could be difficult for claimants to prove—this neutral reading of ‘interference’ seems fair.
Still, it is not evident that what the union did amounted to ‘interference.’ The aim of the union here was to ensure wide support for the strike and a high turnout, which is particularly important given the requirement of minimum 50% turnout for a ballot to be valid, introduced by the Trade Union Act 2016. It saw an easy and practical way of doing so, in suggesting that members take their own papers – something which was already existing practice and a logical step for members to take which would simplify voting. It is possible that many workers would have taken their own letters anyway, which—if they had done so unprompted—would not have been unlawful (see Sir Patrick Elias, para 82). In these circumstances, to suggest to members that they take their own letters from the Delivery Office is hardly to ‘prevent or hinder’ the normal course of events or to restrict members’ choice in how and where they vote.
The CA clearly considered that what the union did reached beyond merely suggesting this course of action. According to Males LJ, the CWU strongly encouraged its members to take their ballot papers, in ‘an environment in which it may not have been easy for anyone to stand out against the course which was being proposed’ (para 47). Certainly, the Facebook posts and transcript of the address at the Stockton, in combination with videos showing individuals taking their own post, and videos and photos of mass posting events at work could have created an environment in which workers felt under pressure to do as they are asked and as their colleagues do. In particular, this could be the case if voting did not take place in secret. It was argued by Royal Mail that the videos and photographs put forward as evidence suggested open voting (see Swift J, paras 31-34). Swift J accepted that voting did not take place in secret only at the Swansea office, where the videos showed members filling out their papers. However, subsequently some workers gave statements that they did so out of their own free will (Males LJ, para 71). In any event, the judgment of Males LJ suggests that he did not view the question of whether voting took place in secret as crucial to a finding of interference, but as a consequence of interference (para 72).
Whilst not unreasonable, the view expressed by the judges in the CA—which is based on hypothesis rather than concrete evidence that anybody actually felt under pressure or complained of the suggested course of conduct— is strongly concerned with the protection of the workers’ individual interest to the detriment of the collective interest in the planned action going ahead. A certain degree of collective pressure is inevitable, even required, in this context, in order to persuade workers to stand up together to the employer. Furthermore, it is the main task of the union to seek a high turnout and positive vote in the ballot – it is not a neutral association which merely organises the vote. The ‘strong encouragement’ to take their papers, evidenced by the Facebook posts and address to the Stockton Delivery Office, was framed in terms similar to the encouragement to vote ‘YES’ – without doubt something which the union is entitled to do. It is true that the ‘strong encouragement’ that members take their own papers and vote immediately goes a step beyond this, but it is questionable whether it goes so far as to ‘prevent or hinder’ the normal course of events. After all, it can be expected that members would read the words of the union in their context and understand that they are an aspect of the union’s pre-ballot campaign tactics, to which they are not required to adhere. In any event, these actions are not so far removed from admissible campaign tactics, and, on the basis of the presented evidence, it is by no means clear-cut that they constitute ‘interference.’
The judges chose to resolve this ambiguity in a way less favourable to the union, which had the consequence of blocking the proposed action. However, their assessment does not reflect the fact that there was overwhelming support for the action—evidenced also by the results of parallel ballots of employees of Parcelforce (part of RMG), one of which was concerned with the same dispute—and the fact that what was at stake was the exercise of the workers’ and union’s fundamental right to strike (discussed below). They ought to have taken into account the collective interest in the strike taking place, as well as the strong democratic mandate behind it, which would have pointed towards resolving the ambiguity in favour of the planned action going ahead.
c) The human rights dimension
There is a growing body of case-law of the European Court of Human Rights which confirms that the right to strike is protected by the Convention as an aspect of the right to freedom of association, which is a fundamental human right protected under Article 11 ECHR (see eg Enerji Yapi-Yol Sen v Turkey  ECHR 2251). The right to take collective action, including strike action, is also regarded as a fundamental right by the International Labour Organisation, and the European Social Charter, to which the UK is a state party. The outcome of the CWU case is that workers are denied the enjoyment of their right to strike, despite the fact that there was clearly huge momentum behind the planned industrial action. Moreover, the workers and union were denied this opportunity, at a time when they could have exerted significant pressure on the employer, because of actions by the union which were not obviously unlawful at the time and, as I have argued above, were a borderline case of ‘interference’.
It was argued by Lord Hendy QC, counsel for the union, that an injunction in these circumstances would be ‘an unnecessary and disproportionate interference with the right to strike which is protected by Article 11’ ECHR. Males LJ quickly rejected counsel’s proposition, arguing that the union could call a new ballot, which would cause expense for the union and loss of the opportunity to strike in the pre-Christmas period, but otherwise no prejudice. First, these are not insignificant losses, in particular the ability to take industrial action at a time when workers can put maximum pressure on the employer. Second, there are further costs in terms of the personal time and effort of individuals involved in the campaign and loss of momentum behind the strike. The decision that strike cannot lawfully go ahead in these circumstances is a serious interference with the union’s and workers’ Article 11 rights. It is true that in RMT v United Kingdom (2015) 60 EHRR 10, the ECtHR held a part of the claim inadmissible because, following an injunction, the union held another ballot and a strike took place. This aspect of the judgment can be criticised, amongst other things, as an easy way to avoid politically sensitive scrutiny of UK balloting requirements. In any event, key to the reasoning of the Court in the RMT case was that the strike had in fact already taken place (para 45), which is not the case here.
The High Court and Court of Appeal judgments devote little attention to the human rights aspect of this case, with little discussion of Article 11 and no attempt to justify the interference with the right to strike within the parameters of Article 11(2). Both this fact, and the outcome of the case itself, are reminiscent of the words of Maurice Kay LJ in Metrobus v Unite The Union  EWCA Civ 829 at para 118 that in the UK ‘the right to strike has never been much more than a slogan.’
Indeed, it is not clear how courts should properly take into account the fact that the right to strike is protected by the abovementioned international instruments. There are judgments which acknowledge that the right to strike is not a ‘mere slogan’ and show more sensitivity to the human rights dimension in cases involving labour injunctions, including Sir Patrick Elias himself in RMT v Serco (cited above). In that case, he held that the recognition of the right to strike requires that legislation be interpreted in a normal way, and not strictly against the union (para 9). However, even this might not be sufficient to ensure that Article 11 is properly taken into account, since that provision would suggest a presumption in favour of the exercise of the right, with the need for restrictions to be justified. In the present case, the judges at least propound to give the provisions their ‘ordinary’ meaning (eg Swift J at para 6, Males LJ, para 54), so the approach proposed by Elias LJ in RMT would probably not have made a difference. To ensure that cases such as this are Article 11-compliant, the judges’ assessment should place the human rights dimension front and centre, and include a more focused and detailed analysis of whether the grant of an injunction is a necessary and proportionate restriction of the right to strike.
d) Benefit to the employer
Finally, this case is a remarkable example of an employer successfully preventing industrial action on the basis of legal provisions which are intended to benefit workers. Many of the other notable cases, such as Metrobus, RMT v Serco or British Airways Plc v British Airline Pilots’ Association  EWCA Civ 1633 have concerned provisions (information and notice requirements) which are at least partially, if not entirely, intended to benefit the employer. As discussed above, the purpose of s 230 is to ensure that workers can cast their vote freely, and the purpose of ballots more generally is to ensure that industrial action has a democratic mandate. Here the employer is, effectively, able to use these provisions against workers, in a way which interferes much more deeply with their ability to choose whether and when to take industrial action than any action of the union. This paradox—which is not legally problematic—is perhaps the inevitable product of the fact that organising and going on strike is regarded in law as a prima facie wrongdoing against the employer, rather than as an expression of the fundamental human right of workers to defend their interests through collective action. So long as this is the case, the burden will remain on unions and workers seeking to defend their interest to show that they enjoy an immunity, on top of all the other challenges of organising a successful strike against ever more powerful and well-resourced employers; and the labour injunction will remain a key piece in the arsenal of employers keen to suppress collective action.
Even if we accept the reasoning of the judges at every stage—and I have criticised aspects of that reasoning above—the result in the case itself is difficult to accept. As Sir Patrick Elias says (para 89), given the high turnout and overwhelming ‘yes’ vote, it is difficult to believe that the result of the ballot would have been any different without the actions of the union in question. There was clearly wide support for the planned industrial action and a strong dissatisfaction amongst union members with the behaviour of the employer. Yet, workers were not able to exercise their right to take collective action, because the employer—hardly the hero of the story—was able to take advantage of provisions intended to protect workers against those workers. That such a scenario is possible under UK law is evidence of the fact that the system of immunities, as it stands at the moment, is simply not an adequate way of protecting workers’ and trade unions’ right to strike.