Hendy QC, J., & Novitz, T. A. (2018). The Holship Case. Industrial Law Journal, 47(2), 315-335. [dwy009]. https://doi.org/10.1093/indlaw/dwy009 Peer reviewed version Link to published version (if available): 10.1093/indlaw/dwy009 Link to publication record in Explore Bristol Research PDF-document This is the author accepted manuscript (AAM). The final published version (version of record) is available online via Oxford University Press at https://academic.oup.com/ilj/advance-article/doi/10.1093/indlaw/dwy009/5036770 . Please refer to any applicable terms of use of the publisher. University of Bristol - Explore Bristol Research General rights This document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/red/research-policy/pure/user-guides/ebr-terms/ The Holship case John Hendy QC* and Tonia Novitz** Background The Holship case1 is one of the most significant cases in European labour law so far this century and is of particular interest in the context of British exit (Brexit) of the European Union (EU). The litigation concerned Norway, a country which is not and has never been a member of the European Union (EU). Instead, Norway is a member of the European Economic Area (EEA) alongside other EU member States, but as a member of the European Free Trade Agreement (EFTA). This means that Norway (like other EFTA members, Iceland and Liechtenstein) is covered by standard EU internal market rules governing free movement of goods, services, establishment and workers, as well as competition law rules. EFTA members are not covered by other adjunct aspects of the EU, such as the Common Trade Policy or Common Foreign and Security Policy.2 The EEA operates under a two-tier structure, under which the EFTA Surveillance Authority can exercise powers equivalent to those of the European Commission and the EFTA Court can inter alia give advisory opinions on the appropriate * Old Square Chambers, hendyqc@oldsquare.co.uk University of Bristol, tonia.novitz@bristol.ac.uk 1 Holship Norge AS v Norsk Transportarbeiderforbund Case E-14/15, [2016] 4 C.M.L.R. 29 (Holship). 2 See for a full explanation, http://www.efta.int/eea/eea-agreement. ** 1 interpretation of the EEA Agreement.3 EFTA and EEA membership has been considered as an alternative to ‘hard Brexit’ for the UK and thus its contours and effects are very topical. There has been much speculation that one of the drivers of the referendum vote for Brexit was caused by working class antipathy to freedom of movement of workers, one of the four pillars of the EU Treaty.4 Though research has shown that EU immigration has had little effect on wage levels,5 the widespread perception of insecurity of jobs as a consequence of EU immigration (fostered by a xenophobic media) cannot be doubted. Yet the security of and quality of jobs in the UK is in our view more likely to be threatened by the other three pillars of the EU Treaty as by the free movement of workers from other States. The Holship case demonstrates this. The Holship case illustrates the ways in which the Treaty right to free movement in respect of establishment (a right only available to businesses) can be relied upon by an employer (under both EU law and the EEA Agreement) so as to undermine a longstanding collective agreement. That in question was introduced long ago to end 3 Explained at http://old.efta.int/sites/default/files/documents/eea/16-531-the-two-pillar-structure-surveillanceand-judicial-control.pdf and see R. Spano, ‘The EFTA Court and Fundamental rights’ [2017] European Constitutional L Rev 475. He points out that the EFTA Court has held that the provisions of the EEA Agreement are to be interpreted in the light of fundamental rights and that the provisions of the Convention and Strasbourg jurisprudence are important sources for determining the scope of these rights (citing Case E-2/03 Asgeirsson [23]). He also refers to the principle of homogeneity set out in Article 6 of the EEA Agreement which requires that that Agreement is to be interpreted in conformity with the relevant rulings of the Court of Justice of the European Union (CJEU) and Article 3(2) which requires that the EFTA Surveillance Authority and the EFTA Court shall pay due deference to the principles laid down by the relevant rulings of the CJEU. 4 M. Godwin and O. Heath, ‘Brexit Vote Explained: Poverty, Low Skills and Lack of Opportunities’ Joseph Rowntree Foundation (JRF), 31 August 2016, available at: https://www.jrf.org.uk/report/brexit-vote-explained-poverty-low-skills-and-lack-opportunities. 5 C. Dustmann and T. Frattini, ‘The Fiscal Effects of Immigration to the UK’ (2014) 124 The Economic Journal, F593–F643; confirming findings in C. Dustmann, T. Frattini and C. Halls, ‘Assessing the Fiscal Costs and Benefits of A8 Migration to the UK’ (2010) 31 Fiscal Studies 1. 2 the prevalence of casual hand-to-mouth labour on the Norwegian docks. It will become apparent that the findings of the EFTA Court gave the standard formalistic priority accorded to employer free movement rights (and competition law) in accordance with the approach established in the Viking and Laval cases.6 In applying the EFTA Court’s opinion, the majority of the Norwegian Supreme Court was persuaded by this (we believe flawed) analysis. In contrast, the minority in the Norwegian Supreme Court demonstrated their concern with the collective bargaining rights at stake and their social consequences. The stance taken by the minority demonstrates that it is not a jurisprudential necessity to take the line of the majority in Holship though the latter’s reasoning is evidence of an emerging trend in EU (and EEA) law. The case illustrates the dangers of internal market free movement rights when commandeered by employers to dismantle collective bargaining schemes without due consideration for standards long established by the International Labour Organisation (ILO). It also highlights the potential for resistance by an EFTA State like Norway (or perhaps, post 2019, the UK), which could be achieved with reference not only to ILO norms but also Article 11 of the European Convention on Human Rights.7 Ironically, the judgment comes at a time when there are signs that the neo-liberal antipathy to collective bargaining may no Case C-438/05, International Transport Workers’ Federation (ITF) and Finnish Seamen’s Union (FSU) v Viking Line [2007] ECR I-0779 (Viking); Case C-341/05, Laval un Partneri v Svenska Byggnadsarbetareförbundet [2007] ECR I-11767 (Laval). 7 Though the commitment of the European court of Human rights to trade union rights may be thought in doubt, at least for the moment, see K.D. Ewing and J Hendy, ‘The Strasbourg Court Treats Trade Unionists with Contempt: Svenska Transportarbetareförbundet and Seko v Sweden’ [2017] 46 ILJ 435, and K.D. Ewing and J. Hendy, ‘Article 11(3) of the European Convention on Human Rights’, [2017] EHRLR 356. 6 3 longer hold such a hegemonic place in capitalist thought than was formerly the case.8 The case was decided at a time when the political institutions of the EU have finally begun to grapple with the social effects of the 2007 Laval decision, which prioritised employers’ free movement of services. This is taking place after a decade of posted work leading to the curtailment of collective bargaining and collective action, which has wreaked havoc in the labour markets of member States. 9 Nevertheless, it is still doubtful whether the political manoeuvres in the EU will revert to efficacious protection of collective bargaining at the national level, despite commitments now made to amendment of the Posted Workers Directive.10 8 See, e.g., the 50 page chapter devoted to collective bargaining in the less than 200 pages of the OECD, ‘Employment Outlook, 2017’, OECD Publishing, Paris, 2017 and the footnoted references in chapter 2 of K.D. Ewing, J. Hendy, and C. Jones (eds), A Manifesto for Labour Law: Towards a comprehensive revision of workers’ rights (Institute of Employment Rights, 2016). 9 See chapter 7 of K.D. Ewing and J. Hendy, Reconstruction after the Crisis: A manifesto for collective bargaining (Institute of Employment Rights, 2013); I. Schömann, ‘Reforms of collective labour law in time of crisis: towards a new landscape for industrial relations in the European Union?’, in D. Brodie, N. Busby and R. Zahn, The Future Regulation of Work, New Concepts, New Paradigms (Palgrave, 2016) at 152. ‘…the reforms have resulted in a dramatic decline in collective bargaining coverage, a breakdown of collective bargaining, a strong downward pressure on wages leading to deflationary tendencies, downward wage competition and an overall reduction in the wage-setting power of trade unions’. See also E. Menegatti, ‘Challenging the EU Downward Pressure on National Wage Policy’ (2017) 33 Int J of Comp Labour Law and Ind Relations, 195-219; M. Martínez Lucio, A. Koukiadaki and I. Tavora, The Legacy of Thatcherism in European Labour Relations: The Impact of the Policies of Neo-liberalism and Austerity on Collective Bargaining in a Fragmented Europe (Institute of Employment Rights, 2017); and J. Hendy, ‘Britain, Trade Union Rights, The EU and free trade agreements’ in C. Jones et al, Europe, the EU and Britain: Workers’ Rights and Economic Democracy (Institute of Employment Rights, 2017). 10 Directive 96/71/EC of the European Parliament and of the Council of 16 December 1996 concerning the posting of workers in the framework of the provision of services [1997] OJ L18/1 (hereafter Posted Workers Directive);Proposal for a Directive of the European Parliament and of the Council amending Directive 96/71/EC of The European Parliament and of the Council of 16 December 1996 concerning the posting of workers in the framework of the provision of services COM/2016/0128 final ; General Approach agreed by the Council of the EU on 24 October 2017 13612/17 (‘Council General Approach 2017) and Joint Statement of February 2018: www.europarl.europa.eu/news/en/press-room/20180301IPR98901/joint-statement-on-the-revision-of-theposting-of-workers-directive. 4 In the UK there will be no support for any extension of protection for posted workers by reference to collective agreements from the Conservative Government: the 2017 BEIS ’Industrial Strategy’ White Paper, Cmd 9528 has but one mention of the words ‘trade union’ and none to collective bargaining. In contrast, the Labour Party in its June 2017 election manifesto, ‘For the Many not the Few’, promised to ‘roll out sectoral collective bargaining.’11 This could significantly alleviate the position of both posted workers and workers exercising their own freedom of movement to work in the UK. However, to achieve the roll out of sectoral collective bargaining, whether there is soft Brexit (or no Brexit), it will be important to seek amelioration of the limitation on trade union rights evident in this case. The facts The Norwegian Transport Workers' Union (NTF) is a constituent trade union of the Norwegian Confederation of Trade Unions (LO), the largest trade union federation in Norway. The system of industrial relations in Norwegian docks goes back at least to 1896 when dockworkers established the NTF and won a collective agreement which established the precedent for negotiated terms and conditions in Norway’s ports. Like dockers all over the world, dockers in Norway were casual, zero-hours workers with little job security and precarious earnings, hired for the shift by different employers. 11 Labour Party, London, 2017 at 47 and 51. 5 Against a background in which LO and the national employers' organisation, Næringslivets Hovedorganisation (NHO) had reached a National Agreement in 1935 (in parallel with the establishment of national collective bargaining arrangements elsewhere in the world as one of the mechanisms for emergence from the Great Depression of the 1930s12), a collective agreement for the dock industry was established through collective bargaining in 1939-1940 which gave priority of engagement in dock work to dockworkers. This too mirrored developments on the docks elsewhere. 13 With the advent of containerisation, roll-on, roll-off (ro-ro), increased mechanisation and automation, in 1973 the ILO adopted Convention No. 137 (CI37) ‘concerning the social repercussions of new methods of cargo handling in docks’. C137 addressed the issues of job security and irregular income which plagued dockwork (and which the UK had, at least partially, addressed in the National Dock Labour Scheme of 1947). Thus Article 2 of C137 provided for permanent or regular employment for dockworkers and for minimum periods of employment or a minimum income. Article 3 provided for registration of dockworkers and priority of engagement for them in dock work. Article 7 provided that the provisions of the Convention ‘shall, except in so far as they are otherwise made effective by means of collective 12 E.g the National Industrial Relations Act of 1933 in the USA, the 1936 Matignon Accords in France, the Conditions of Employment Act 1936 in Ireland and the Saltsjöben Agreement of 1938 in Sweden. 13 P. Turnbull and D. Sapsford, ‘Hitting the Bricks: An international comparative study of conflict on the waterfront’ (2001) 40(2) Industrial Relations: A Journal of Economy and Society 231; S. Davies, C.J., Davis, D, de Vries, L.H. van Voss,, L. Hesselink and K. Weinhauer, Dock Workers: International explorations in comparative labour history, 1790-1970 (Routledge, 2017). We are grateful for the insights into the dock industry and some of its literature provided to us by Prof Peter Turnbull. 6 agreements, arbitration awards or in such other manner as may be consistent with national practice, be given effect by national laws or regulations.’ C137 was ratified by Norway in 1974. LO and NHO agreed that there was no need for legislation, that the provisions of C137 had already been largely achieved in Norway by collective agreement and that they would enter (in accordance with the 1935 National Agreement, since amended) a ‘Framework Agreement’ for the dock industry in Norway. Accordingly, the 1976 Rammeavtale om fastlønnssystem for losse- og lastearbeidere (the Framework Agreement on Fixed Pay Scheme for Dockworkers - the ‘Framework Agreement’) secured for Norwegian dockworkers the benefits of C137 including the right to permanent employment and better pay. The Framework Agreement has been renewed by collective agreement between LO and NHO every other year since 1976, the latest amendment being in 2016. The essential features of the Framework Agreement (and C137) are to provide better protection of dockworkers' pay and conditions of work and to provide secure employment. This is achieved by predicting the variable demand for dockwork and dockworkers, giving priority of engagement to registered dockworkers for dock work and adjusting the size of the workforce to levels adapted to the needs of each of the thirteen largest ports in Norway, including the port of Drammen. The Framework Agreement established (by clause 3) an Administration Office (AO) in each port to ensure that all port users were given access to sufficient, trained and qualified registered dockworkers and to manage the dockworkers' priority of 7 engagement to perform the work in each port of discharging and loading of all ships of more than 50 tons dwt that called there. The Framework Agreement applied only from ship to quay and vice versa so goods handling outside the port was unaffected by its requirements. The AO in each port ran the scheme, maintained the Register, organised the rota of work, received the dues paid by each port user for the labour it used and paid the dockworkers. Each AO is a non-profit-making entity with independent legal status and its own board comprised (in Drammen) of three representatives of the local port employers and two representatives of the local dockworkers. ln Drammen there were six permanent dockworkers and a pool of on-call dockworkers. Holship Norge AS was an established user of Drammen port. It is a Norwegian company wholly owned by a Danish parent company. Holship’s main activity was the cleaning of fruit crates but it also handled a small but increasing amount of cargo transported by ship. Holship decided that it wished to provide its own labour to load and discharge its ships' cargo; a task previously understood to be the legitimate role of the registered dockworkers. NTF invited Holship to become a party to the Framework Agreement. Holship did not respond so NTF threatened industrial action in the form of a boycott to compel Holship to become a party and abide by the Framework Agreement. The EFTA advisory opinion 8 In Norway it is unlawful for a union to impose a boycott on an employer save in accordance with the Boycott Act of 5 December 1947. This requires the union to apply to the Court to sanction such industrial action as lawful. Section 2(a) of that Act provides that a boycott will be unlawful if its purpose is unlawful or if its object cannot be achieved without causing a breach of the law. Accordingly, NTF applied to the Drammen District Court for a declaration. The court held that the proposed boycott to enforce the collective agreement was lawful. Holship appealed to the Borgarting Court of Appeal which, in turn, also held that the proposed boycott was lawful. Holship then appealed to the Norwegian Supreme Court. The latter referred questions to the EFTA Court for an advisory opinion. The EFTA Court found, in an advisory judgment of 19 April 2016,14 that the purposes of the boycott went beyond 'the establishment or improvement of working conditions of the workers of the AO and go beyond the core object and elements of collective bargaining and its inherent effects on competition'.15 This was on the basis that the Drammen AO had a 'business objective' which was 'to preserve the market position of the Administration Office'.16 The EFTA Court assumed that because the NTF was a party to the Framework Agreement, it participated in the management of the AO. But it did not. The docker representatives on the AO board were elected by the dockworkers in the port to represent them. They were not appointed by the 14 Holship Norge AS v Norsk Transportarbeiderforbund Case E-14/15, [2016] 4 CMLR 29 (Holship). Holship, at [50]. 16 Holship, at [49]. 15 9 union to represent it. Yet, based on that erroneous holding and ignoring the social purpose of the AO in administering the collective agreement to protect the income and jobs of registered dockworkers, the EFTA Court concluded that the NTF and the AO had a ‘common interest to preserve the market position of the AO’. The NTF was thus understood to have a 'business objective' as well as its trade union functions.17 Correspondingly, the Court held that therefore the ‘boycott must therefore also be attributed to the AO’.18 It found that the Framework Agreement: and the creation of the AO appear therefore not to be limited to the establishment or improvement of working conditions of the workers of the AO and go beyond the core object and elements of collective bargaining and its inherent effects on competition.19 The EFTA Court further held that ‘the AO system in the present case protects only a limited group of workers to the detriment of other workers’ (i.e. the Holship workers). In particular, a boycott would detrimentally affect the situation of these other workers, who ‘are barred from performing the unloading and loading services and may even lose their employment if their employer affiliates to the Framework Agreement.’20. These findings led the EFTA Court to apply EEA competition law (Articles 53, 54 and 59 of the EEA, identical to the relevant provisions of the Treaty on the Functioning of the European Union (TFEU), Articles 101, 102 and 106). The Court 17 Ibid.. Ibid. 19 Holship, at [50]. 20 Holship, at [51]. 18 10 accepted the principle established in the Albany cases21 that collective bargaining over ‘measures to improve conditions of work and employment’22 was exempt from competition law.23 However, the EFTA Court held that by reason of the 'business objective' and the detriment to other workers, the Framework Agreement 'cannot generally be exempted' from EEA competition rules. It was for the national court to determine if the competition rules were broken.24 The EFTA Court also held that the proposed boycott aimed at procuring acceptance of a collective agreement which required priority to be given to employment of the registered dockworkers through the AO, was likely to discourage or prevent the establishment in Norway of companies from other EEA States and therefore constituted a restriction on Holship's freedom of establishment under Article 31 EEA (which is the exact equivalent of the Article 49 TFEU, the provision deployed in Viking).25 Whilst collective bargaining and collective action are fundamental rights so that the protection of workers may justify restriction on freedom of establishment,26 in the instant case the priority of employment given to the registered dockworkers and the creation of the AO were 'not limited to the establishment or improvement of working conditions' of the registered dockworkers and so were considered to 'go 21 Albany International v. Stichting Bedrijfspensioenfonds Textielindustrie (Case C-67/96), [1999] ECR I-5751; [2000] 4 CMLR 446; Joined Cases Brentjens' Handelsonderneming v. Stichting Bedrijfspensioenfonds Voor de Handel in Bouwmaterialen: (C 115–117/97), [1999] ECR I-6025; [2000] 4 CMLR 566 ; and Maatschappij Drijvende Bokken v. Stichting Pensioenfonds Voor de Vervoer- En Havenbedrijven (Case C-219/97), [1999] ECR I-6121; [2000] 4 CMLR 599. 22 Albany at [59]. Without an Albany type exception, competition law (as with the nineteenth century UK application of the doctrine of restraint of trade) necessarily prohibits agreements between suppliers of, inter alia, labour as to the prices and conditions on which they will work. 24 See Holship at [67]-[100]. 25 Holship at [120]. 26 Holship at [121]-[124]. 23 11 beyond the core object and elements of collective bargaining', protecting one group of workers to the detriment of others.27 This was an application (and perhaps an extension) of the rationale in the Viking and Laval line of authority (though only Viking was cited).28 The matter was to be determined by the national court which, as is evident, had been given the clearest steer as to what it ought to do. The thrust of the EFTA Court's judgment was that if the boycott had the sole (and genuine) purpose of protecting the interests of workers, that would be an overriding reason of public interest which could justify restrictions on freedom of establishment but, since the boycott served the independent business interests of the AO and the NTF in preserving the AO’s market position, such justification was lost. The judgment of the Norwegian Supreme Court The matter returned to the Norwegian Supreme Court, which gave judgment on 16 December 2016.29 Seventeen judges heard the case. They split ten to seven on the two issues in the case. As to competition law, the majority held that there were no sufficient grounds for overriding the EFTA Court's conclusion that the Framework Agreement was not 27 Holship at [126]. Holship at [108]-[114]. 29 HR-2016-2554-P, (Case No 2014/2089). Unofficial translation from Norwegian. (Holship NSC). 28 12 exempt from competition law but they held it was not necessary to explore the limits of the exemption for collective agreements because the NTF lost, in any event, on the application of the Viking rationale.30 The minority agreed with Justice Indreberg that the Framework Agreement should be exempt from competition law by reason of the Supreme Court’s earlier upholding of the lawfulness both of the Framework Agreement and of boycott action in support of it, in the Port of Sola case in 1997.31 Further, the minority found that the facts on which the EFTA Court had decided the Holship case were flawed – the NTF was not a member of AO’s Board; nor did the NTF and AO have any business interest in preserving AO’s market position.32 The majority found the factual error not significant. As to freedom of establishment, applying the Viking and Laval approach, the majority held that the boycott in pursuance of the Framework Agreement would constitute a restriction on Holship’s freedom of establishment under Article 31 EEA. Though the priority clause could be regarded 'as an overarching goal that safeguards the workers' interests' this was not conclusive.33 The priority clause would limit other operators' access to the market (loading and discharging cargo from ships) and protected AO from competition from them. Even though the AO does not operate 30 Holship NSC at [128]. Rt 1997 s 334. In that case the Supreme Court (of five judges) held, unanimously, that a boycott to achieve application of the Framework Agreement in Stavanger was lawful, see the extensive citation from the 1997 judgment in Holship at [73]. 32 Holship NSC at [151]. 33 Holship NSC at [101]. 31 13 for profit, it carries on a market activity in which other operators wish to participate. Prioritising the engagement of AO’s employees limits the access of other operators, gives AO employees an advantage over other workers and protects AO from competition. Protecting a business from competition, they held, cannot justify restrictions on freedom of establishment.34 Even if the overarching purpose of the notified boycott was to safeguard workers' interests, this cannot be accepted as a compelling, legitimate restriction on freedom of establishment.35 Though the right to boycott (which may be subject to greater restrictions than the right to strike) is protected as a human right under the various treaties, that right must nevertheless be subject to the test of proportionality and be balanced against freedom of establishment under Article 31.36 Undertaking that balancing exercise, the majority held that: The primary - and desired - goal of the boycott, is to restrict other operators' access to the loading and unloading services market. This means that the boycott has an extremely invasive effect on the freedom of establishment and that it will clash with other workers' interests.37 They held that Holship would create jobs which would be no less significant than those of the registered dockworkers, and that the purpose of the priority clause could be achieved by other (unspecified) means.38 Applying the balance therefore, the majority concluded that the priority clause ‘does not fulfil the requirement for a reasonable balance between the freedom of establishment and a possible 34 Holship NSC at [103]-[105]. Holship NSC at [109]. 36 Holship NSC at [110-[117], esp. at [116]-[117]. 37 Holship NSC at [118]. 38 Holship NSC at [118]. 35 14 fundamental right to boycott.’39 In effect, therefore, the requirements of C137 and the obligations to protect the right to strike and to bargain collectively in international treaties to which Norway was party were secondary to freedom of establishment under Article 31. The purpose of the boycott was therefore unlawful. The NFT and LO were ordered to pay 7,915,606 kroner (846,959.03 euros, £756,214) in costs. The minority followed the Supreme Court’s earlier judgment in the Port of Sola case which it summarised as concluding, in relation to the Framework Agreement, that:40 the collective agreement at issue was a distinctive one. But its sole purpose was to ensure the workers’ pay and working conditions. The minority did not accept that the AO had a business objective beyond the protection of its employees pay and working conditions; on the contrary: 'the real purpose of the establishment of administration offices is to strengthen dockworkers' pay and working conditions and this continues to be the only purpose.'41 Furthermore:42 For port users, the priority clause means, not only that they are required to use prioritised workers on loading and unloading operations, but also that, under the scheme which the NTF seeks to impose, they will have access to skilled labour. This takes place without any costly intermediaries since the AO is not allowed to make a profit. 43 The minority were thus able to pray in aid Viking and Laval, which emphasised that under the EU Treaty, the free movement of goods, persons, services and capital also required 'a high level of employment and social protection' and must be balanced 39 Holship NSC at [119]. Holship NSC at [157]. 41 Holship NSC at [169]. 42 Holship NSC at [170]. 43 Holship NSC at [170]. 40 15 against social policy objectives which included 'improved living and working conditions ... [and] proper social protection and dialogue between management and labour.'44 Consequently, and by reference to European Commission v Spain45 the minority held that both the actual purpose for the priority clause and its effect - thus also the boycott - were to promote safe and favourable conditions of work for the dockworkers, and ‘EU law permits restrictions on freedom of establishment on such grounds.' The minority concluded that the boycott was an appropriate means of enforcing the collective agreement,46 and the collective agreement was an appropriate means of protecting the pay and working conditions of dockworkers.47 The evidence was that there was no other practical way of achieving security and income for 44 Holship NSC at [171]-[172], citing [78]-[79] of Viking reiterated in [104]-[105] in Laval. Holship NSC at [173]-[174] and [194]). In Case C-576/13 European Commission v Kingdom of Spain, the CJEU held that a similar port labour scheme was aimed at protecting workers and ensuring the regularity, continuity and quality of cargo handling and these two purposes were within the overriding reasons of public interest which can justify restrictions on freedom of establishment. The scheme in that case only failed to pass the hurdle of justification because the party seeking to do so (the Spanish government) had made no attempt to show that the restriction was necessary or proportionate and that less invasive alternatives were not available. The CJEU appeared to consider the creation of a pool of dockworkers was legitimate (at [195]). The minority (at [172]) also found nothing in the Charter of Fundamental Rights of the EU had weakened the significance of the ‘social policy considerations’ referred to in Viking and Laval, and noted (at [176]) that many other countries in the EU had adopted dock labour schemes in accordance with ILO C137. They observed that the European Parliament had rejected a Directive proposed by the European Commission because it would have weakened dockworkers’ priority of engagement. The minority noted that a later proposal from the Commission was intended to give protection to dockworkers’ interests (at [177]-[179]). 46 Holship NSC at [182] citing Viking and Laval which both concerned a boycott. Indeed, they could have cited the judgment in the UK Court of Appeal in Govia Thameslink Railway Ltd v Associated Society of Locomotive Engineers and Firemen [2016] EWCA Civ 1309; [2017] 2 CMLR 24; [2017] ICR 497; [2017] IRLR 246 which, analysing the Viking and Laval line of authority (and, ironically, Holship in the EFTA court) pointed out (at [28], [30], [35] and [36]) that it was the objective sought to be achieved by the industrial action and not the damage caused by the action itself which might unjustifiably interfere with freedom of establishment or of service under the TFEU: ‘it is the object or purpose of the industrial action and not the damage caused by the action itself which renders it potentially subject to the freedom of movement provisions’ (at [39]). 47 Holship NSC at [183]-[136]. 45 16 dockworkers.48 The boycott did not exceed what was necessary to achieve its purpose and therefore the consequential restriction of freedom of establishment was justified. Implications of the judgment The effect of the Holship judgment is that the Framework Agreement is unlawful in every Norwegian port in which it formerly operated. It is consequentially unenforceable by boycott or other industrial action. It has been, in effect, annulled as was the collective agreement in Demir and Baykara v Turkey.49 The same is true in every EU Member State which has adopted a similar scheme with a similar purpose, whether or not they have ratified C137. Any port employer in such a State which cares to establish under its control a company in another EU State to undertake its port operations (whether directly or through a local end-user company) may, like Holship, rely on the 'Holship loophole' to be free of a similar dock labour scheme. Such an employer will then have no restraint in using casual dock labour, hired by the day or the shift. Indeed, there is compelling evidence that the European Commission is encouraging challenge of such schemes throughout EU Member States.50 Holship NSC at [187]-[193]; and the minority noted (at [188]) that the European Parliament’s rejection of the Commission’s proposal to weaken protection for dockworkers meant that ILO C137 was not regarded as outdated and was regarded as still relevant by the ILO Committee of Experts in 2002. 49 [2009] 48 EHRR 54. See See K.D. Ewing and J. Hendy, ’The Dramatic Implications of Demir and Baykara’ (2010), 39 ILJ 2. 50 European Commission, ‘Impact assessment accompanying the proposal for a regulation of the European Parliament and of the Council establishing a framework on market access to port services and financial transparency of ports’ SWD(2013) 181 final. See discussion in R. Thomas and P. Turnbull, ‘Talking up a storm? Using language to activate adherents and demobilize detractors of European Commission policy frames’ (2016) Journal of European Public Policy, available at: http://dx.doi.org/10.1080/13501763.2016.1162831 48 17 Though the case gives prominence in its justification to the business role it artificially attributed to the AO, it is yet a further demonstration of the destructive force of EU law operating through the Viking and Laval line of authority (in neither of which was a joint enterprise a necessary ingredient).51 As the minority judgment concluded:52 the employment situation for the permanently employed dockworkers with the Administration Office would become a lot less secure if the priority clause were not observed. The basis for permanent employment may disappear. Furthermore, for the intermittent hours they are employed, such casualised workers will be paid, not at the hourly rates set by the AO but at rates as low as the market will stand. It is here that perhaps the greatest direct impact of the case may be seen. For the employer exploiting the Holship loophole will be able (as in Viking and Laval) to use workers from another EU State (or even the host State) to carry out the work at a cheaper rate than the collective agreement formerly provided.53 The incentive for port employers to exploit the Holship loophole either to retain competitive advantage or to avoid being undercut on labour costs is self-evidently 51 The judicial innovation of the Viking, Laval and Holship line may also pose a threat under the Free Trade Agreements (‘FTAs’) which the UK is likely to be obliged to enter to replace trade with the EU (and by which it is already bound, e.g. the Comprehensive Economic Trade Agreement between Canada and the EU). Where these FTAs include an investment chapter, this will invariably provide for an investor state dispute settlement procedure (‘ISDS’) which allows corporations to bring legal proceedings against States for, amongst other things, failure to ensure Fair and Equitable Treatment (‘FET’). The rationales for each of these judgments could, we think, equally unfortunately found a complaint of lack of FET under a relevant FTA. 52 Holship NSC at [193]. 53 For the growth of such extreme cost-cutting practices, see P. Turnbull and V. Wass, ‘Defending Dock Workers Globalization and Labor Relations in the World's Ports’ (2007) 46(3) Industrial Relations: A Journal of Economy and Society 582. 18 likely to become irresistible. The protections that all three international parties (governments, employers and workers) of the civilised nations of the world sought to provide for dockworkers in 1973 in enacting ILO C137 will be, at least in Europe, swept away and a return to brutal casualised conditions in the docks appears to be presaged. The autonomy of the Norwegian State in ratifying C137 in 1974 and by the NHO and LO in making the Framework Agreement 40 years ago and maintaining it ever since has been negated by the EFTA Court and the Norwegian Supreme Court. Indeed, the very system of industrial relations in Norway is put at risk by these judgments, as they are in other EU States where such relations are primarily determined by collective agreement. The reversion to free competition on workers' terms and conditions (against which the Albany case permitted collective agreements to protect) will inevitably infringe the ostensible social policy objectives of, and the rule of law in, the EU. We return below to the question whether the European Commission or Council has evinced much concern regarding such issues.54 It is notable that one aspect of the case is that, in order for a company to take the benefit of the Holship loophole, it must have a national identity which is not that of the host country (and is within the 27 EU States or further EFTA membership). Any company the nationality of which is exclusively that of the host country is barred from this benefit and is thus treated less favourably on grounds of nationality. The discriminating difference in treatment between the foreign owned and domestic See K.D. Ewing and J. Hendy, ‘The Eclipse of the Rule of Law: Trade Union Rights and the EU’ (2015) 4 Revista Derecho Social y Empresa 80. 54 19 companies in the Holship case is not a point that seems to have troubled the Norwegian Supreme Court.55 Application to the European Court of Human Rights (ECtHR) In response to the judgment the unions have made an application to the ECtHR asserting, principally, breach of Article 11 of the European Convention on Human Rights (ECHR).56 This is essentially on the ground that Norway has now violated their right to take industrial action to enforce the terms of a collective agreement. The outcome of Holship placed restrictions on both the right to bargain collectively and the right to strike, which are guaranteed under Article 11(1) ECHR.57 The application asserts that the restrictions cannot be justified under Article 11(2) ECHR. It prays in aid the proposition that the very existence of C137 negates any suggestion that restrictions contrary to that Convention could be regarded as ‘necessary in a 55 Neither did the EFTA Court or the Norwegian Supreme Court have any hesitation in piercing the corporate veil to hold that the Danish parent company could displace its Norwegian subsidiary so as to take advantage of freedom of establishment. Workers denied, in most jurisdictions, the right to claim against a foreign parent company when their rights cannot be enforced against the domestic subsidiary employer may feel justifiably aggrieved. 56 There is also a complaint of breach of Art.13 ECHR, a failure to provide a domestic remedy that allows the competent national authority to address the substance of a complaint under the ECHR and to provide appropriate relief. The ground here is that the Supreme Court has, by subordinating the right to bargain collectively under Art.11 ECHR to the right of establishment under Art.31 of the EEA Agreement, deprived the unions of their right to remedy since the subordination precluded any assessment of whether the restriction based on Art.31 EEA was necessary in a democratic society as required by Art.11(2) ECHR (this is shown in the majority judgement at [86] which makes no distinction between whether the court’s analysis starts with one or the other right). The application points out that the obligations in the EEA Agreement are not binding as are those under the EU Treaties and the former does not purport to protect human rights. The balancing operation is therefore different (taking note of Bosphorus Hava Yolları Turizm ve Ticaret Anonim Şirketi v Ireland (2006) 42 EHRR 1 at [160]-[165]). The application also asserts a breach of Art.14 ECHR by discrimination on grounds of nationality in that only a company claiming foreign nationality and not that of the host country can take advantage of the Holship loophole. 57 Demir and Baykara v Turkey (2009) 48 EHRR 54 and RMT v UK (2015) 60 EHRR 10; Hrvatski Liječnički Sindikat v Croatia (App. No.36701/09, judgment of 27 February 2015) and Veniamin Tymoshenko v Ukraine (App. No.48408/12, judgment of 2 January 2015). 20 democratic society’ and points out that the ILO Committee of Experts,58 and the European Committee on Social Rights (ECSR),59 have held that the Viking and Laval decisions are inconsistent with ILO Convention No. 87 and the European Social Charter respectively.60 The ECSR made clear that the EU freedoms of cross border establishment and provision of services cannot be treated, from the point of view of the system of values, principles and fundamental rights embodied in the Charter, as having a greater a priori value than core labour rights, including the right to make use of collective action to demand further and better protection of the economic and social rights and interests of workers.61 The danger, of course, is the precedent set by the RMT v UK case,62 in which the ECtHR upheld in principle protection of the right to strike under Article 11, but refused ultimately to follow the supervisory findings of the ILO Committee of Experts or the ECSR on the status of secondary action.63 It is to be hoped that the Strasbourg Court will be bolder in this instance when the ‘core’ of the right is clearly ILO Committee of Experts, Report III (Part 1A) (2010) (UK): ‘With respect to the matter raised by BALPA [British Air Line Pilots’ Association], the Committee wishes to make clear that its task is not to judge the correctness of the ECJ’s holdings in Viking and Laval as they set out an interpretation of the European Union law, based on varying and distinct rights in the Treaty of the European Community, but rather to examine whether the impact of these decisions at national level are such as to deny workers’ freedom of association rights under Convention No 87. … The Committee observes with serious concern the practical limitations on the effective exercise of the right to strike of the BALPA workers in this case. The Committee takes the view that the omnipresent threat of an action for damages that could bankrupt the union, possible now in the light of the Viking and Laval judgements, creates a situation where the rights under the Convention cannot be exercised’. 59 LO and TCO v Sweden, Complaint 85/2012, 5 February 2014. The Committee pointed out that ‘excluding or limiting the right to collective bargaining or action with respect to foreign undertakings, for the sake of enhancing free cross border movement of services and advantages in terms of competition within a common market zone, constitutes, according to the Charter, discriminatory treatment on the ground of nationality of the workers, on the basis that it determines, in the host State, lower protection and more limited economic and social rights for posted foreign workers, in comparison with the protection and rights guaranteed to all other workers’ ([141]). 60 In relation to the European Social Charter, the Committee of Ministers of the Council of Europe, in noting the report of the ECSR, concluded that the decision of the latter ‘raises complex issues in relation to the obligation of member States to respect EU law and the obligation to respect the Charter’: Council of Europe, Committee of Ministers, 5 February 2014 9CM/ResChS (2014)1. 61 LO and TCO v Sweden (ibid) at [122]. 62 (2015) 60 EHRR 10. 63 Ibid at [106]. 58 21 at stake, the case involves the de facto annulment of a collective agreement as in Demir, and the respondent State is not in a position to argue either that it considered the restriction necessary in a democratic society or that it exercised its margin of appreciation in imposing that restriction.64 Furthermore, the affront to the ILO here is not to its jurisprudence but to the very substance of a widely ratified Convention. The ECtHR, it is to be hoped, will return to a more ‘integrated’ approach,65 especially after the careful reasoning (and intense criticism of the Court’s findings in RMT) by Judge Pinto De Albuquerque in the Hrvatski case.66 However, the political realities of the Convention and the Court which arguably forged the outcome in RMT remain, namely the UK threat to abandon its Convention obligations.67 Although the case concerns Norway, not the UK, the Court will be concerned about its impact in the UK. The UK Conservative Party (currently in government) has repeatedly placed in its manifesto a pledge to leave the Convention;68 although the voting public has been informed that such a measure will not be taken until after Brexit (of whatever kind) is complete.69 64 Though there is a margin of appreciation for the Norwegian State in relation to these rights, it has not (thus far) sought to rely on that discretion. It ratified C137 43 years ago and has not denounced it. Norway has since the beginning of the 20th century, accepted that employment relations are regulated by the national social partners, and for 40 years it has been content that the regulation of employment relations in the docks of Norway should be in accordance with the Framework Agreement. The argument is that the judgment in Holship expresses a change of policy which stems not from the exercise of discretion by the State but from law imposed on Norway by EFTA and its Court which has given primacy to certain economic freedoms of companies in EU States with no or little regard for the deference required to be paid to the human rights of citizens of the States of the Council of Europe. 65 See V. Mantouvalou, ‘Labour Rights in the European Convention on Human Rights: An Intellectual Justification for an Integrated Approach to Interpretation’ (2013) 13 Human Rights Law Review 529 66 See Hrvatski above. 67 See the two articles by K.D. Ewing and J. Hendy cited in n.7 above, and A. Bogg and K. D. Ewing, ‘The Implications of the RMT Case’ (2014) 43 ILJ 221. 68 See the 2010 and 2015 Conservative Party Election Manifestos, 69 See https://www.conservativehome.com/thetorydiary/2017/10/remaining-in-the-echr-is-the-price-for-leavingthe-eu.html. 22 Discussion The first point to note is that the judgment of the EFTA (and majority in the Norwegian Supreme Court) in Holship is jurisprudentially unnecessary. It is a sop to employer demands in the EU and symptomatic of an ongoing preference for commercial interests over those of workers, following case law which emphasises under Article 16 the freedom to conduct a business.70 Such an interpretation of both EU (and EEA) competition and free movement law relating to establishment was previously considered unwarranted, as acknowledged by the minority in the Norwegian Supreme Court. It also places EU law in a contradictory position. On the one hand, sectoral bargaining is the only way at present in which binding norms can currently be set for regulating the labour standards of posted workers (after Laval), but here it is deemed anti-competitive when in the interests of achieving stable access to safe work for a particular group of workers. This is likely to cause all sorts of difficulties around the informal demarcation of collective agreements nationally and sectorally. 70 See Case C-426/11 Alemo-Herron and others v Parkwood Leisure Ltd [2013] IRLR 744, Judgment of 18 July 2013; E. Gill Pedro, ‘Freedom to Conduct Business in EU Law: Freedom from Interference or Freedom from Domination?’ 23 European Journal of Legal Studies available at http://www.ejls.eu/23/241UK.htm; see also for a more skeptical view of the influence of Article 16 see J. Prassl, 'Freedom of Contract as a General Principle of EU Law? Transfers of Undertakings and the Protection of Employer Rights in EU Labour Law (Case C-426/11 Alemo Herron v Parkwood Leisure)' [2013] 42 ILJ 434 and J. Prassl, ‘Business Freedoms and Employment Rights in the European Union’ (2015) Cambridge Yearbook of European Legal Studies 189 and note the reassuring outcome in Case C-201/15 AGET Iraklis, despite the Opinion of AG Wahl. The force of Alemo-Herron is only slightly diluted by C-328/13 Osterreichischer Gewerkschaftsbund v Wirtschaftskammer Osterreich- Fachverband Autobus-, Luftfahrt- und Schifffahrtsunternehmungen [2014] ICR 1152 and C-680/15 Asklepios Kliniken LangenSeligenstadt GmbH v Felja [2017] IRLR 653. 23 Moreover, the minority of the Norwegian Supreme Court found good reason to dispute the proposition that Holship’s workers would need to lose their jobs; indeed, to preserve the collective agreement would rather prevent wage and job insecurity.71 The majority judgment was predominantly an attempt to legitmate Holship making its workers do more for less cost. Portraying the outcome as in the interests of those workers is entirely artificial, as is demonstrated by the political agreement (reflective of the wider interests of other workers, of other employers and the State) previously reached on ILO C137. The Holship judgment would seem to represent an underhanded attack on the very kind of collective bargaining that was residually, and in theory, to be permitted. The restoration of sectoral collective bargaining in the UK to which, as noted above, the Labour Party is committed and which we and others have promoted elsewhere72 should establish minimum terms and conditions (and forms of engagement) which will ensure a level playing field in each industry so that the national origin of the worker gives no advantage, holds no detriment and provides no competitive advantage to the employer. It will be recalled that sectoral collective bargaining was introduced and extended in the past in the UK and in Norway with the active support of employers precisely in order to prevent competition on labour costs, and to direct competition towards investment, efficiency and productivity. Of course, 71 Holship NSC at [156] K.D. Ewing and J. Hendy, Reconstruction after the Crisis: A manifesto for collective bargaining (Institute of Employment Rights, 2013); K.D. Ewing, J. Hendy and C. Jones (eds), A Manifesto for Labour Law: towards a comprehensive revision of workers’ rights (Institute of Employment Rights, 2016); K.D. Ewing and J. Hendy, ‘New Perspectives on Collective Labour Law: Trade Union Recognition and Collective Bargaining’ (2017) 46(1) ILJ 23; T. Novitz, ‘Collective Bargaining, Equality and Migration: The Journey to and from Brexit’ (2017) 46(1) ILJ 109. 72 24 restoration of the coverage of sectoral collective bargaining even to the levels before the neo-liberal agenda began its destruction in 1979 will take some time but the industries with the most precarity should be the first candidates in which national joint councils should be established.73 However, this approach by the EFTA Court and the majority in the Norwegian Supreme Court demonstrates deep suspicion of any apparatus set up to administer such a scheme and a complete lack of sympathy for the benefits to workers, social justice and the economy that the inevitable constraints of collective bargaining place on individual employers to the benefit of employers as a whole.74 It is notable that the Laval case has had profound repercussions across Europe, including undercutting and depression of wages, health and safety concerns and loss of access to jobs for home state workers.75 Collective agreements were only permitted to set labour standards for posted workers if given statutory effect or of general application nationally or in a particular sector, or within a specified geographical area.76 The decline of sectoral bargaining pushed by deregulatory austerity measures during the financial crisis led to a lowering of labour standards and employers began pulling in temporary migrant workers in ever increasing efforts to cut labour costs and improve profit margins.77 Many States were complicit in these tactics so as to promote foreign direct investment and thereby enhance their 73 E.g., L. Hayes, 8 Good Reasons why Adult Social Care needs Sectoral Collective Bargaining (Institute of Employment Rights, 2017). 74 Not least by raising wages and so increasing demand. For a fuller discussion with references see A Manifesto for Labour Law, op cit. 75 European Commission, Commission Staff Working Document, Impact Assessment, Strasbourg, 8.3.2016 SWD(2016) 52 final, at 13 and 36; see also the extensive research conducted by J. Cremers, In Search of Cheap Labour in Europe: Working and Living Conditions of Posted Workers (CLR Studies: European Institute for Construction Labour Research, 2011). 76 See the interpretation of Article 3(1) and (8) of the PWD in Laval at [80]. See also Case C-346/06 Rüffert v Land Niedersachsen (Rüffert) [2008] ECR I-1989 at [27]. 77 See ns 9 and 10 above. 25 gross domestic product (GDP) which had all sorts of implications for IMF and Troika lending.78 The result was the loss of jobs for local host State workers, an increase in temporary posting and a decline of wages often by up to 50% for those doing the work. This has proven politically unacceptable, spurring nationalism and potential fracturing of the European project as perceived in the White Paper on the Future of Europe: Reflections and scenarios for the EU27 by 2025 issued by Jean-Claude Juncker on 1 March 2017.79 . Posting has in this way become an urgent matter of political attention. As a result, there are signs of change. For example, in Sähköalojen ammattiliitto ry v Elektrobudowa Spółka Akcyjn, 80employers again sought to prioritise their free movement entitlement (this time of services) over workers’ entitlements to fair wages and representation. However, the Court of Justice of the European Union (CJEU) responded by recognising that a host State trade union should be able to recover unpaid wages for posted workers with their explicit consent. In some circumstances trade union representation can be permitted a role. Moreover, the failure to apply collectively agreed wages was acknowledged to be part of the problem by the resultant Commission proposal for revision of the PWD. Perhaps the most encouraging of the Commission’s proposals in this respect was in relation to ‘subcontracting chains’, whereby Member States would have the option to apply remuneration established at company level and other applicable collective 78 For e.g. under IMF Stand-By and Extended Arrangements, a member can borrow up to 145 percent of its quota annually and 435 percent cumulatively; 50% of its quota being determined by GDP. 79 European Commission COM(2017)2025 of 1 March 2017 at 12 for e.g. ‘a questioning of trust and legitimacy’ is acknowledged. 80 Case C 396/13 Sähköalojen ammattiliitto ry v Elektrobudowa Spółka Akcyjn (ESA), judgment of 12 February 2015, unrep. Available at: http://curia.europa.eu/juris/liste.jsf?num=C-396/13. 26 agreements to any subcontractor even where posting of workers is envisaged.81 This would give ‘the faculty to Member States to oblige undertakings to subcontract only to undertakings that grant workers certain conditions on remuneration applicable to the contractor, including those resulting from non-universally applicable collective agreements.’82 That capacity would remain subject to proportionality tests and nondiscrimination requirements. And presumably it would not apply to work done by workers from another EU State working in the host country by reason of exercise of their own freedom of movement rather than by being posted by an employer. Nor would it apply to work done in another State.83 Yet, even so, such a measure would partially address the realities of the labour markets of EU States under neoliberalism and, in particular, the incidence of subcontracting core functions (avoiding enterprise-based collective agreements) and the out-sourcing of non-core functions (undermining sectoral collective agreements).84 The importance of the free movement of services and the posting issue is evident from the fact that, despite the ‘red card’ shown to one Commission proposal, the Council is now determined to proceed with amendment of the Posted Workers Directive to address this crisis.85 However, in this context, in the Council redraft explicit reference to collective bargaining (and supply chains) was lost, expunged 81 See COM(2016) 128 final at 7-8. Ibid. 83 In Case C-549/13 Bundesdruckerei GmbH v Stadt Dortmund, 2014, the CJEU held that where work is done in one EU State to exploit lower labour costs with the purpose of supplying the goods or services to another with higher labour standards, the latter cannot insist (because of the freedom to supply services) on its higher standards being met by the employer in the State where the work was done. 84 See text accompanying n10 above. 85 See Council of the European Union 2016/0070 (COD) 13612/17, 24 October 2017 General Approach on the Commission Proposal amending Directive 96/721/EC. 82 27 from Article 1(2)(b). What will make a dramatic change is the entitlement of a posted worker to all applicable terms and conditions of employment in the Member State where the work is carried out after 12 months by virtue of a new insertion (aa). More generous than the Commission proposal of 24 months, this indicates a clear entitlement to terms under applicable collective agreements under national law. So key EU institutions have now jointly recognised the need to reform treatment of employers’ free movement of services (with the most recent agreement in principle by the European Parliament as of February 2018), but not (as yet) the destabilising approach of the EFTA Court and the majority of the Norwegian Supreme Court in Holship. Lessons? What do these developments tell us? First, that there is urgent need to recalibrate the balancing exercise as performed in Viking, Laval and Holship for reform in the EU around treatment of employers’ free movement rights and their relative weighting in relation to protection of the right to industrial action and the right to bargain collectively (a right from which the right to industrial action cannot be severed). Thankfully, there are now signs of recognition of this need. Secondly, the crude importation into labour law of laws prohibiting anti-competitive behaviour by entrepreneurs (which underpins employers’ freedoms of establishment and the provision of services) continues to haunt labour law like a ghoul. Holship 28 shows that the battle fought in the nineteenth century between competition law (then called ‘restraint of trade’) and trade union rights is being refought in Europe.86 In this connection it is worth recalling that notwithstanding that the CJEU recognised that competition law could not be permitted to negate collective bargaining in the Albany cases,87 it was resurrected to prohibit collective bargaining for the self-employed,88 with, more recently, a relaxation for a tiny sub-set of selfemployed workers.89 Thirdly, it is evident that there are options for interpretation of current EU and EEA laws which depart from the conventional wisdom of the EFTA Court and the majority Norwegian Supreme Court judgment in Holship. In this, the minority judgment offers a genuine alternative, which could be pursued by the CJEU or the UK courts90 - or even, in the event of ‘soft’ Brexit (which entailed EEA membership) 86 This is, no doubt, because neo-liberalism in the twentieth and twenty first century is no more than a more sophisticated restatement of nineteenth century laissez-faire capitalism. In the UK, s.3 of the Trade Union Act 1871 prevented the contract of membership of a trade union being rendered void because its purposes (making and enforcing collective agreements) were necessarily in ‘restraint of trade’ (see the line of cases: Hornby v Close (1867) 19 Cox CC 393; Hilton v Eckersley (1855) 6 E&B 47; Osborne v ASRS [1909] 1 Ch 163 at 189;. the principle was reiterated as recently as 1994 in Boddington v Lawson [1994] ICR 478 Ch D). Other European countries required protection to similar effect although different in form. In the USA trade union protection against competition law was established in the Clayton Act 1914, the Norris-LaGuardia Act 1932 and non-statutorily in Apex Hosiery v Leader 310 US 469, 60 S Ct 982, L Ed 1311 (1940). 87 See n.21 above. 88 Pavlov and Others v Stichting Pensioenfonds Medische Specialisten (Joined Cases C180–184/98) [2001] 4 C.M.L.R. 1, 89 An illogical exemption for self-employed workers undertaking the same activity as employees of the same employer: FNV Kunsten Informatie en Media v Staat der Nederlanden, Case C-413/13. The restriction on collective bargaining by the self-employed is now currently being challenged by a collective complaint submitted by the Irish Congress of Trade Unions to the European Committee on Social Rights (123/2016) which was found to be admissible on 23 June 2017. 90 Though the attitude of the UK courts to Article 11 is not encouraging: Gate Gourmet London Ltd v T&GWU [2005] IRLR 881; R (National Union of Journalists) v Central Arbitration Committee [2006] ICR 1; Metrobus Ltd v Unite the Union [2010] ICR 173 (esp at [35]); British Airways v Unite (No.1) [2010] IRLR 423; Elias LJ in London & Birmingham Railway Ltd (trading as London Midland) v ASLEF; Serco Ltd v RMT [2011] ICR 848, at [8] and [9]; Secretary of State for Education v NUT [2016] IRLR 512, at [77]; Thames Cleaning and Support Services Ltd v United Voices of the World [2016] EWHC 1310 (QB); Netjets Management Ltd v Central Arbitration Committee [2013] 1 All ER 288 (at [41]-[42]); R (Boots Management Services Ltd) v CAC [2017] 29 at a later date. Such judicial boldness would be reminiscent of the earlier refusal of the Norwegian Supreme Court in the ‘wharfs case’91 to be bound by the EFTA Court’s limited view of what collective agreements were permissible under the Posted Workers Directive. The ECtHR could be part of that normative correction, if the Court is willing to assert the rights to collective bargaining once articulated in the Demir case, which now seem to be eclipsed by its political hesitation on such matters. The Labour Party is committed to negotiating a satisfactory deal with the EU which protects workers’ rights.92 Amongst the other things which it must negotiate93 is a solution to the Viking, Laval and Holship problem. If able to do so, this might save the institution of collective bargaining across the continent and would do an enormous favour for the whole European trade union movement (and indeed for European workers and employers generally). It is also essential for the Labour Party to do so if it is to have the legal space to ‘roll out sectoral collective bargaining’ which will otherwise be under challenge, without doubt, by reference to EU competition law IRLR 355; and London Borough Wandsworth v Secretary of State for Business, Innovation and Skills [2017] EWCA Civ 1092. 91 2012/1447 ‘Verftsdommen’ available at: https://www.domstol.no/globalassets/upload/hret/decisions-inenglish-translation/2012-1447-engelsk.pdf. See also S. Evju, ‘Safeguarding National Interests: Norwegian Responses to Free Movement of Services, Posting of Workers and the Services Directive’ in Stein Evju (ed.), Cross-Border Services, Posting of Workers, and Multilevel Governance. Institutt for privatretts skriftserie 193/2013. Oslo. Privatrettsfondet. 2013, 259; and E. Rogstad, ‘From Rush-Portuguesa to Laval and the WharfCase: How the posting of workers in the EU and Norway has been shaped by the Courts’ (Trondheim, Masters thesis, 2013). 92 For the Many not the Few, n.11 above, and see K. Starmer, Theresa May’s Brexit red lines were reckless. Now she has to cross them, The Guardian, 5 December 2017 at https://www.theguardian.com/commentisfree/2017/dec/05/theresa-may-brexit-red-lines-reckless-hostage-duppromises-cant-keep. 93 Amongst others it needs a way round the constraints imposed by the Treaty on the Functioning of the European Union (Articles 107, 113, 119(2), 170(2), in particular) so as to enable it, without Commission challenge, to renationalise and nurture the NHS, Royal Mail, the railways, energy, water, gas and so on. 30 and the business freedoms which constitute three of the four pillars of the EU Treaty. In the meantime, the campaign for sectoral bargaining must go on despite the threat that Holship presents. We have never needed sectoral national level collective bargaining more. 31