Along with Mark O’Brien and several other authors in this journal, I am anxious to understand what lay behind the dramatic drop in UK strike activity from the late 1970s to the early 1990s and the extremely low level ever since.1 Mark’s approach, in his article last year, was broad brush, with limited evidence to back up his claims, and he was often mistaken about the law on strikes. So I produced an alternative account with sources that could be checked.2 His response has been to restate his position.
In his own words: the “Tories’ anti trade union laws” were “the result of deeper changes in policy orientations (industrial and social) that, while driven by Conservative strategists, were accepted and incorporated by a Labour government by the late 1970s”. This shift toward legalism “fundamentally altered the modus operandi of national trade union leaderships”. “During the early 1980s, and especially after the 1984-5 miners’ strike, unions internalised the principle of legalism”, which “represents the fundamental change that properly explains the low level of strikes in Britain today”.3 Union leaderships became “averse to leading prolonged and open-ended strikes” for fear of members “compromising the union legally”, so they now operate a “policy of extreme restraint…well within and very far from the limits of the law, even to the point of walking away from what should have been easy victories”.4
This brief comment on Mark’s response can only tackle some of our profound differences of approach. It will focus on arguments against the thesis expounded in the previous paragraph. There is insufficient space to engage with his objections to my own explanation of strike decline.
Mark rightly says I did not discuss “the socio-legal framework” of welfarism and working class accommodation, and its subsequent “dismantling”, which he now claims was “the core” of his analysis, nor the “social protections for workers” that was “half” his analysis. He argues that, by focusing on the law (rather than the wider socio-legal framework), I miss his larger framing, particularly (since the 1970s) “the raised levels of risk associated with industrial action for trade unions and for workers”.5 But I went into much more detail than Mark on the legislative changes and potential (and actual) effects for unions taking strike action. Regarding the socio-legal framework, Mark’s short account of “the inter-war social settlement” focused mainly on social housing and poor relief. Unwittingly, perhaps, it painted a picture of a benign state and ignored, to name but two exceptions, the persistence of slum housing post-war and the infamous “means test” of the 1930s. Mark’s post-1970s “loss of social protection” covered the decline of social housing, rise of private renting and “attrition” of the welfare state, resulting in “the social quid pro quo that once mitigated the risks” of striking being eroded.6 That is true, but individual strikers had always been vulnerable, well before the 1970s. They could be dismissed (for breach of contract); and strikers have been (since 1875) potentially liable for (limited) damages for breach of contract.7 Even before the welfare state’s attrition, strikers were barred from claiming unemployment benefit (since its origin in the 1911 National Insurance Act) or national assistance (later, supplementary benefit) because of the “trade dispute disqualification”.8 Dismissed strikers were always penalised (initially for up to six weeks) when claiming unemployment benefit, their status being likely categorised as “voluntary unemployment” or dismissal for “misconduct”.
A key argument in Mark’s first article was Labour’s embrace of legalism in the late 1970s. He now claims that “the use of physical force…by a Labour government” against pickets at Grunwick in 1977 was “a historical turning point for class relations”. But he provides no evidence for this statement (or for his claim that Labour at this time “accepted and incorporated…deeper changes in policy orientations”). Instead we are told: “I find it difficult not to see [prime minister Jim] Callaghan’s worries about scenes redolent of ‘Saltley Gates’…as significant”.9 But, for Callaghan (as befits a former home secretary), Grunwick became a public order issue.10 Mark then jumps from Callaghan in 1977 to Tony Blair in 1997 and ignores my timeline detailing when and how Labour policy changed.11 Tony Cliff and Donny Gluckstein, writing at the time, observed that it was not until September 1986 that Labour approved strike ballots (but not other Tory changes) and they quoted the Financial Times on “this break with Labour’s own history”.12 And it was only at the 1989 Labour Party conference, after its fateful wide-ranging “Policy Review”, that Labour secured what Lewis Minkin (the chronicler of Labour-union relations) called a “profound break” and a “historic shift” in its policy on strike law—accepting several Tory legal changes, including union liability for “unlawful” strikes. It “transcended ninety years of history and overcame what seemed an impossible barrier in the form of 90 percent of the Party Conference votes”.13
Mark’s explanation of his key term “legalism” is also unsatisfactory: “The type of state and court intervention in the operations and behaviour of trade unions that emerged within a new socio-legal framework”.14 “State” intervention was the series of statutes, starting in September 1980, enacted by Tory governments, particularly changing the lawfulness of strikes. Court intervention followed as employers were now able to seek injunctions against unions (from December 1982).15 Yet collective agreements in the UK (unlike many other countries) are not legally binding on the parties to the agreement,16 and this continues to keep the courts out of much trade union activity. Mark’s specific allegation is that “trade unions were hobbled by statutory ballots to stay within the law, and so were then compromised in every other aspect of the conduct of their industrial action. Balloting ahead of strikes [“the most important anchor of legalism”] became the norm during this period, the principle of self-regulation being abandoned”.17
This last point warrants further exploration: unions, which, by the 1970s, “had operated for a century, as self-regulating bodies”, according to Mark, “could no longer [by 1997] be said to be self-regulating”.18 But they never were totally self-regulating. Achieving lawful status in 1871, unions that registered were granted tax relief (from 1893) on their provident funds (friendly benefits). Registered unions had to deposit rulebooks (covering certain prescribed minimum topics), rule changes and annual accounts with the registrar.19 Laws governing rules for union mergers were enacted in 1876, 1917, 1940 and 1964, and all involved ballots. Following the “Osborne judgment” and subsequent court cases by disaffected members against their unions, the Liberal government passed the Trade Union Act 1913, which required any union wishing to spend money on political objects to hold a ballot to set up a political fund, which was itself subject to further rules. There is also a long history of court cases of unions pursuing members for embezzlement of funds, and members accusing unions of breaching the rulebook.20 Full self-regulation never existed.
Mark takes me to task for contesting his view that the 1984-5 miners’ strike was the “last major example of an industrial battle fought against the legalistic terms upon which such disputes were now to be conducted”. He clarifies this by now arguing that he was not stressing that they were the last group to defy the new laws, instead that their “behaviour was that of a self-regulating organisation, determining their action not according to a statutory ballot”.21 But the strike started in early March 1984, six months before statutory industrial action ballots were introduced (on 26 September 1984), so it was not affected by this legal change and Mark’s argument on this point falls. I then cited News International in 1986 and P&O European Ferries in 1988 as disputes where unions did defy the law. Mark complains that the print unions balloted before striking; yes, but once their strike started they were then buried under a deluge of injunctions against unlawful secondary action—their main recourse once all strikers had been sacked and News International started printing newspapers with a scab workforce.
Regarding P&O, Mark cites an injunction against the seafarers’ union (the NUS), preventing it from balloting, as evidence of the union’s “intention to conduct the dispute within the law”.22 But this is facile, given the circumstances. An unballoted national strike of ferry workers, in support of sacked Isle of Man seafarers, on 31 January 1988, led to a slew of injunctions against the union for not holding ballots and for unlawful secondary action. Striking workers at Dover then balloted quickly on their own dispute with P&O which began on 3 February. Mark’s point concerns court action against a national ballot of ferry workers, in late March, which had to be suspended. In May the union was separately fined £150,000 (and its assets sequestrated) for “flagrantly, repeatedly and gravely” continuing secondary picketing against Sealink, the other main cross-channel ferry company. Unofficial occupations of ferry boats and oil rig supply vessels, in solidarity, took place away from Dover. Further fines were inflicted for continuing disobedience of injunctions.23 Mark dismisses such resistance: “strikes in this era were confused affairs, characterised by rhetorical defiance of the law followed by climb-downs in the face of injunctions and sequestrations of union assets”. I return to Mark’s view that “union leaders are unprepared to defy the law, and so are utterly reluctant to provide any type of effective leadership”,24 but even if the first statement is true does the second statement follow? Mark has (retrospectively) set an impossible standard for unions by his reading of the experience of the 1971 Industrial Relations Act, against which he claimed (wrongly) that trade union opposition was “absolute”.25
What was the general experience of ballots? Between 26 September 1984 (under the Trade Union Act 1984) and 30 August 1993 (when postal ballots and notice periods became compulsory), strike ballots could be conducted at the workplace. These could be quick, especially where only one workplace was involved, and action could follow immediately. Even then, active campaigning improved ballot results. In the late 1980s workplace ballots took place typically after workplace meetings; and, from 1993, meetings before or during postal ballots were common. Evidence from 1994 showed that turnout in postal votes was high where up to about 500 members were involved, after which it fell sharply; but some 90 percent of ballots involved less than 500 trade unionists.26
On average, 150 strikes annually were made official in advance through the 1960s and 1970s; perhaps another 200 or so were “officialised” retrospectively each year.27 By contrast, trade union surveys and then Electoral Reform Services figures suggest that, from the mid-1990s to 2008, there were around 1,000 official strike ballots per year. Some unions (public and private sector) could have hundreds of ballots in a single year.28 The CWU postal union alone recorded over 1,500 between 1996 and 2001 (at a time also of significant unofficial action in Royal Mail).29 Generally only a small proportion of successful ballots have resulted in strikes, because it is a cheap way of demonstrating workplace strength—a regular finding and one given particular force in this journal by Richard Morgan’s account of private-sector disputes.30
Yet Mark insists that, by the early 1990s, “the door to collective official industrial action had closed considerably”. Unions found an alternative, he says, when the 1992 Trade Union and Labour Relations (Consolidation) Act (TULR(C)A) introduced statutory consultation on redundancy. It “altered the register of trade unionism, especially at the local level” and “opened up a new terrain of legal action as the alternative for trade union bureaucracies, now able to trumpet occasional victories”.31 The problem with this argument is that the 1992 Act, as I pointed out before, consolidated existing legislation and did not create new law.32 Redundancy consultation was brought in by Labour in the mid-1970s (responding to a European directive).33 Mark’s invention of a “new model of statutory consultation” in the 1990s is another example of his jumping to conclusions without firm evidence.
To summarise so far, the component parts of Mark’s argument are found wanting. Individual strikers have always been vulnerable. Labour did not start accepting the Tory legislation until the late 1980s by which time most unions were conforming, having shown a lot more than rhetorical defiance of the law. Unions are only slightly less self-regulating as a result of the Tory laws. Far from ballots “hobbling” industrial action, there have been a large number of ballots most years. To adapt Carl von Clausewitz’s most famous saying, strikes are the continuation of collective bargaining by different means, while strike ballots are now often a stage in negotiations (as strike threats, and some unofficial strikes, were before 1984).
Finally, Mark’s claim that unions have shown extreme restraint since they “internalised legalism” ignores my point that as civil service and local government white-collar unions developed strike policies, in 1969-70, they ruled out indefinite national action. This followed the practice, from the 1950s onwards, of private sector unions where industry-wide agreements covered many individual employers.34 As regards unions “walking away from…easy victories”, I have already dealt, elsewhere in this journal, with Mark’s “great example” of the public service pensions dispute of 2011.35 But, more generally, union negotiators are often reluctant to exploit to the full any advantage in a strike situation if they fear that this will damage long-term bargaining relationships with the employer. After all, as Mark himself admits, unions are “organisations of class compromise”. There are plenty of instances of such behaviour from before Mark’s “era of legalism”, which were therefore not contingent on union leaders’ “determination…to operate within the legalist domain” for fear of members “compromising the union legally”.36
Since 1980 the lawfulness of industrial action has been greatly constrained. Simultaneously (but not generally as a result of the law, as I showed in my last article) the level of strike activity has fallen dramatically. Despite Mark’s claims of “fundamental” changes in trade unionism, there are striking continuities with earlier periods in unions’ negotiating tactics today.
Dave Lyddon is a retired industrial relations lecturer and a long-standing SWP member.
1 Latest figures (14 May 2019) show only 67 stoppages beginning in 2018, identical to the all-time low of 2017—ONS, 2019a. Working days “lost” and workers involved in stoppages were similar to 2017 figures—ONS, 2019b.
2 O’Brien, 2018; Lyddon, 2018.
3 O’Brien, 2019, pp136, 130, 137
4 O’Brien, 2019, pp130, 138, 134.
5 O’Brien, 2019, pp131-132. Mark argues that the socio-legal framework is one “upon which…[trade unions] develop a symbiotic though conflictual dependency”, but I am unclear what this means.
6 O’Brien, 2018, p165.
7 Wedderburn, 1971, p115. This was little used outside coal mining but persisted after nationalisation.
8 Ewing, 1991, pp63-117.
9 O’Brien, 2019, pp132-133.
10 McGowan, 2008, p392.
11 Lyddon, 2018, pp203-205.
12 Cliff and Gluckstein, 1988, pp359-360.
13 Minkin, 1991, pp623-624; McIlroy, 1991, pp205-225.
14 O’Brien, 2019, pp129-130.
15 Dates of operation of the 1980-93 Acts given in Lyddon, 2018, p207.
16 In February 2014 the CWU and Royal Mail signed a legally binding “contract”, which has been updated since, but this is very unusual.
17 O’Brien, 2019, pp134, 136.
18 O’Brien, 2018, pp165, 162.
19 Originally the Registrar of Friendly Societies, now the Certification Officer.
20 Hickling, 1967.
21 O’Brien, 2018, p164; O’Brien, 2019, p133. The latter claim is repeated, the miners struck “outside of the statutory definition of how a union should decide to act and without any attempt to seek legal sanction for their industrial action”—O’Brien, 2019, p134.
22 O’Brien, 2019, p133.
23 Marsh and Ryan, 1989, pp228-247; Berlin, 2006, pp67-73; McIlroy, 1991, pp143-149; and contemporary newspaper coverage.
24 O’Brien, 2019, p133; O’Brien, 2018, p167.
25 O’Brien, 2019, p132. On the IR Act, O’Brien, 2018, p160; and Lyddon, 2018, pp197-199, for a rebuttal.
26 Undy, Fosh, Morris, Smith and Martin, 1996, pp197-232.
27 For details, see Lyddon, 2018, p210.
28 TUC, 1995-1997; TUC, 1998-2001; ONS, 2017. The TUC surveys were never comprehensive; ONS figures (which stopped in 2016) come from Electoral Reform Services (not the only independent scrutineer), and exclude ballots of under 50 members where a union chose (lawfully) to supervise the ballot itself.
29 Gall, 2003, p64.
30 Morgan, 2016.
31 O’Brien, 2019, pp136-137.
32 Lyddon, 2018, p200, n51. All subsequent trade union legislation amends this Act; see the text of the Trade Union Act 2016, for example.
33 Sections 188 and 189 of TULR(C)A 1992 (as amended), which Mark cites, derive from sections 99 and 101 of the Employment Protection Act 1975.
34 Lyddon, 1998, cited in Lyddon, 2018, p212.
35 O’Brien, 2019, p134; Lyddon, 2015, pp154-164.
36 O’Brien, 2019, pp132, 145, 130.