Union purpose and power: Regulating the fissured workplace

Abstract

Notwithstanding long membership decline and fresh threats in workplaces, there is evidence that unions can still exercise power even in the unlikely setting of fissured workplaces. The question is: in what particular ways do unions exercise power and for what purpose? This study of union success in a fissured supply chain argues that, to answer this question, considerations of union power must be tied more explicitly to strategic choices, union purpose, and to how conditions are regulated. In so doing, it challenges the argument in ‘power resources theory’ that power is chiefly measured between actors, and suggests that there is a necessary prior step: considering the power resources available to a specific actor. Similarly, it undermines the idea that power needs to be transformed to be effective and instead suggests that it is only through the interplay of power resources available to unions that any individual power becomes effective.

Introduction

Over the past 30 years there has been a decline in employment standards and income equality along with enhanced employer discretion and union membership decline (Visser, 2019). Among the many drivers of these trends have been changes in the employment relationship, notably the ‘fissuring’ of the workplace, be it through the ‘gig economy’, casual work or ever more intricate supply chains (Weil, 2014). The fissuring of the employment relationship poses several difficulties for workers, unions and other regulatory parties by, among other things, introducing a complexity about who the ‘true’ or ‘controlling’ employer is. This leads to questions about which actors can best undertake workplace regulatory roles. This conundrum is the foundation of this research. Historically, workers’ associations, chiefly unions, have taken on this role, representing workers and enforcing labour standards. However, as unions and state-based institutions seem to be regulating an ever-diminishing set of workplace relationships, the question of how these increasingly fissured employment relationships are regulated becomes more complex and pressing.

Can unions act as non-state regulators or has the fissured workplace made that impossible? Unions in Australia have seen their formal regulatory role decline through successive decades of neoliberal policy-making (Ellem and Cooper, 2020). Nonetheless, we argue that, despite this state hostility and membership decline, unions can still exercise power, representing members and defending their conditions. The question is: how? While there has been significant scholarship on the erosion of labour standards and the role of unions as countervailing forces to labour market changes (Weil, 1994), there has been little theorising of how unions might build power to complete this task. The revival of power resource theory may be a fruitful way ahead.

Our starting point in developing this argument is the challenge set out by Refslund and Arnholtz (2022) in their comprehensive assessment of the power resource approach in this journal. They point to a number of limitations of the approach but also to its potential, urging scholars to embrace ‘a more nuanced understanding of power resources and how these are embedded and institutionalised in society’ (Refslund and Arnholtz, 2022: 1973). They also emphasise that it is not enough to insist that unions can exercise power; it is vital to understand the specific sources of power and how power is (or is not) exercised (Refslund and Arnholtz, 2022: 1970–1971). This article responds to Refslund and Arnholtz by developing one line of argument as we examine unionism in an Australian supermarket supply chain: that the ability of unions to exercise power in fissured workplaces is enhanced by measuring one power resource relative to the others available to a union under specific circumstances and not relative to the power held by other actors.

This article starts with arguments about modes of re-theorising union power and moves on to provide an overview of the empirical context and the research methodology. It then turns to a case study of union strategy in the supermarket supply chain where, in the face of significant difficulty, a union succeeded in organising the workers in a particular supplier at the bottom of the supply chain and winning improved conditions. This is followed by a conclusion drawing out the key theoretical findings and, in line with the stated political purpose of power resource theory, assessing the implications of the case study for union strategy.

Reconceptualising union power in the fissured workplace

In this section, we set out an argument shaped by Refslund and Arnholtz’s view that a consideration of the purpose of the exercise of union power is missing from much existing analysis of union strategy. A focus on specifying union purpose means that our analysis overcomes the problem of the power resources approach using only ‘crude indicators’ (Refslund and Arnholtz, 2022: 1959) of union power. In this case study, we define that purpose chiefly in terms of the enforcement of the conditions under which union members work, in this instance, in a fissured supply chain. Rather than simply review the literature on power resources, then, this section weaves this discussion into a consideration of studies of enforcement, that is, it brings together arguments about power and purpose.

Faced with union decline and changes in the nature of work, there is now at least a generation of scholarship in the ‘union renewal’ literature that focuses on how unions might rebuild (for an overview, see Ibsen and Tapia, 2017; also Holgate, 2021, on union purpose and organising). In the last four years or so, much of this work has been framed through a revival of interest in ‘power resource theory’ (following Schmalz and Dorre, 2018; Schmalz et al., 2018). Most recently, this framework has been used to propose a path for union renewal (Holgate, 2021) and to analyse union power and strategy, in many national settings (Ellem et al., 2020; Hui, 2022; Nowak, 2022; Preminger, 2020; Trif et al., 2021; for important earlier work see Brookes, 2013; Perrone et al., 1984). These studies add both empirical richness and theoretical insights to earlier studies framed by power resource theory.

The power resources approach has a more complex lineage than is commonly acknowledged: Refslund and Arnholtz (2022) helpfully trace it back to Korpi (1978), who notes that ‘[w]hile the collective organisation of labour could take on other forms of organisation, power resource theory perceives unions as the primary collective actor on behalf of the working class’ (Refslund and Arnholtz, 2022: 1960, see also Atzeni, 2021; Nowak, 2022). For scholars who are interested in explaining the trajectory of unionism, this framework has been powerful in theory development and, potentially, in union practice.

This article draws on (and adapts) Refslund and Arnholtz’s overview of the theory which identifies five sources of power: structural, associational, institutional, ideational and coalitional. The power resources approach builds on Wright’s (2000) class power/interest approach to argue that workers’ power is derived from specific sources. The foundational elements are the ‘structural’ power that workers hold because of the location of their work in the economy and the ‘associational’ power they can exercise through collective activity. Unions may also have ‘institutional’ power from being recognised actors in state-based tribunals and processes (Schmalz et al., 2018). Unions may also access ‘ideational’ power (Preminger, 2020) through aligning their work to the morality of meta-narratives and casting some acts as moral imperatives and the acts of others as morally reprehensible. Refslund and Arnholtz (2022: 1962, 1965–1966) also identify a fifth source, coalitional power, which is similar to the concept of ‘societal power’ in the work of Schmalz and Dorre (2018), pertaining to unions networking to build their legitimacy. We briefly explain each of these forms of power, showing at the same time that understanding the nature of the workplace and regulation is essential to an assessment of effective union power in pursuit of a given purpose, and examining the conceptual links between power and regulation.

Returning to what in some senses is the foundational power resource, structural power, we build on other scholarship which shows that it varies between sectors, and within supply chains, as well as over time. In the fissured workplace, where the employment relationship has been fragmented by outsourcing (often into supply chains), casualisation and ‘on-demand’ or ‘gig’ work (Anner et al., 2021; Weil, 2014), unions’ structural power is challenged. In several studies of supply chains, a pattern becomes apparent: the further down the chain a firm is, the lower their financial strength, or control over the work. The capacity of both the firm to bargain with the lead-firm and in turn of the employees to bargain with their firm is also lower (Baglioni, 2022; Fine and Bartley, 2019; Hardy and Howe, 2015; James et al., 2015; Nossar et al., 2017; Warren, 2019). This means that although employees may have the proximity to the person doing the work, control sits with the lead-firm. Fissuring forces a re-examination of assumed knowledge about practical elements of the employment relationship, notably the definition of an ‘employee’ and in turn an ‘employer’. This is a theme picked up in scholarship that suggests that the fissured workplace renders state-based systems of regulation ineffective because such regulation applies to an increasingly limited set of direct employment relationships (Helfen et al., 2018). Nonetheless, fissured workplaces like others are still of necessity regulated (Brody, 1993; Buchanan and Callus, 1993). Familiar questions therefore remain: who does the regulating – employers, the state, workers or unions – and how do they do it?

How do unions exercise structural power to regulate fissured workplaces? Although the fissuring of workplaces has reduced the structural power of workers and their unions by disrupting markets or sectors, some workers might still exercise power because of their structural location in particular supply chains (Alimahomed-Wilson and Ness, 2018). For example, ports workers may hold up goods entering or leaving a country, and workers with specialist skills that require long training periods – or where there are limited training places – such as pilots or licensed aircraft maintenance engineers may also exercise structural power (Nowak, 2022).

These points of leverage can be thought of as particular sites of structural power, described in other studies as ‘choke-points’ (Moody, 2017). For example, in a supermarket supply chain, a choke-point may be a factory’s dispatch office. What does this mean for union power and strategy? It may be that a union does not need to build power across a whole factory or in all transport operators servicing the factory but rather the power held by, for example, dispatch operators could be exercised to enforce standards along the chain (Alimahomed-Wilson and Ness, 2018). Initial scholarship (and some union practice) was optimistic about choke-point strategies (Alimahomed-Wilson and Ness, 2018). Thinking in terms of the power resources approach, however, it appears that the effectiveness of this kind of strategy is dependent upon other factors, critically, the specific market and political context (Nowak, 2022). This problem is neatly illuminated in Nowak’s (2022) account of a national strike by Brazilian truck drivers where, despite significant union member participation, the gains made were limited and transitory because a combination of structural and associational power was not sufficient to secure ongoing gains along the transport supply chain.

Labour markets and workplaces are not the only sources of power for employers and workers. The importance of institutional power has become clear across Western economies through state-led institutional strategies which have enhanced managerial power (Baccaro and Howell, 2017) and facilitated the downward pressure that fissuring places on labour standards (Fine and Bartley, 2019; Wright and Kaine, 2015). Doellgast et al. (2018) find that the key to regulating standards for precarious workers is for unions to develop more worker-friendly industrial relations institutions rather than seeing unions themselves as a potential regulator. Our assessment of institutional power goes beyond this narrative to argue that we need to understand the power that workers and unions can exercise to access and effectively utilise those institutions – whether they are friendly or not.

How, then, might unions exercise power through potentially hostile institutions? After all, institutions may do more than provide a basis for power; they may be the medium through which unions exercise power to enforce labour standards. This matter is addressed here by adding another theoretical strand, a consideration of models of enforcement, to assessments of union power. The most common proposal for institutional power in the fissured workplace is ‘strategic enforcement’ (Weil, 2018). Strategic enforcement in the fissured workplace is enforcement against one firm in the hope that it functions as a deterrent to other firms within the supply chain, a practice made more important with fewer state resources allocated to enforcement at the very time that workplaces have become more complex sites. To argue that institutional power can still be exercised in the same form as in the past, merely at a different scale, ignores, however, the unwieldy structure of the fissured workplace. National regulatory institutions are also, in many cases, ineffective in fissured workplaces because the scale of the span of action available to the parties is incommensurate. Lead-firms operate at a market or sectoral level while the workers operate at the level of the enterprise or even workplace (Anner et al., 2021). This means that the scale at which decisions are being made and the scale of the action by a single group of workers within an enterprise – that is the exercise of associational power – are unlikely to meet in any meaningful way (Helfen et al., 2018; Herod, 1997; Johnstone et al., 2012).

For our argument, this discussion takes us directly to a consideration of ideational power. While strategic enforcement appears to be an example of institutional power being applied to regulation, it also relies on ideational power, that is, a meta-narrative about good and bad, morality and justice (Preminger, 2020). For our purposes, the question is: can unions exercise this kind of power? It is worth noting that ideational power is not simply discursive power (cf. Lévesque and Murray, 2013). Ideational power goes beyond the ‘argument’ being led by the union to connect the union’s conduct to their purpose with meta-narratives about morality (particularly ideas of justice). Nonetheless, ideational power has its own difficulties, not least that it can be difficult to build in a timely way within one campaign.

The very structure of the fissured workplace appears to undercut the effectiveness of ideational power as a driver of strategic enforcement or union effectiveness (Fine and Bartley, 2019; Hardy and Howe, 2015). Because supply chains often traverse the manufacturing, horticulture, logistics and services sectors, it is not obvious who a firm will see as a market competitor or moral leader. We nonetheless agree that unions need to ‘engage in [an] overall discursive battle about visions for society and workers’ self-perception’ (Refslund and Arnholtz, 2022: 1965; see also Lévesque and Murray, 2013). This understanding goes beyond the need for unions to articulate – as they must – visions in relation to a particular dispute or one area of enforcement, or mount an argument. Ideational power seeks to define the values of a dispute in order to reshape the power dynamics between the parties (Preminger, 2020).

How, then, might unions exercise ‘societal power’ (Schmalz and Dorre, 2018: 7) or ‘coalitional power’ (Refslund and Arnholtz, 2022: 1965)? How can this source of power be tied to union capacity to regulate fissured workplaces? Some scholars propose that the focus of regulation should shift from contest to cooperation – and unions might exercise power in fresh ways. The cooperative approach to enforcement refers to a series of loosely linked relational strategies that range from employee ‘hotlines’ to formal partnerships (Ackers, 2015) but it does rely on employees and unions being seen as legitimate civil society actors able to hold other parties accountable (Simms, 2015). Refslund and Arnholtz advance the debate by suggesting that the key question of coalition effectiveness is one of the scale at which power is exercised, one being social, the other sectional. This seems to be problematic in competitive sectors and fissured workplaces in particular where often power is being exercised at multiple levels simultaneously. ‘Coalitional power’ was not a primary resource exercised by the union in this case, but it is important to understand all the potential resources available to see the choices that determined the resources that the union did use.

All these forms of power shape and are shaped by the associational power of workers, the power seen by researchers from Wright (2000) on as the most obvious expression of their actual, as opposed to latent, power. Associational power is the process by which membership numbers are transformed into a resource (Ganz, 2009). Associational power has been understood in the traditional employment relationship as a collective exercise that is immediate, both conceptual and physical, exercised through control over the way that labour is performed and when labour is withdrawn (Ellem et al., 2020; Herod, 1997; Wright, 2000).

Associational power is the manifestation of the relative money, people, authority and resources of the actors (Hahn et al., 2021). Importantly, there is an implicit spatiality to associational power (Herod, 1997) which can be undercut by fissuring. There is both a structural distance between the parties because of the economic system that fissuring imposes and a physical distance. The parties are no longer required to assess their position in the contested space (Herod et al., 2007). This means that the actual decision-makers are not forced to confront the physical discomfort of associational power (for example, walking through a picket line) but instead receive reports on the consequences of that power (for example, a drop in output). The exercise of associational power must be assessed, we will argue, in the context of the (interrelated) options available to a union rather than simply how it exists in relation to the employers’ power.

When power resource theories are combined with more attention to union purpose and regulation, we can identify the sources of power and the relative strength of the actors. This is important strategically as well as conceptually: unions need to be able to measure the interplay of the strength of their various power resources before assessing them against the relative power of other actors. Combining these literatures therefore allows us to overcome one of the conceptual weaknesses of current approaches to union power in supply chains: the prior step of understanding the sources of power available to an actor before seeking to understand the contest of powers between actors. We argue that by doing this and by locating power resource theory within literature on fissuring and regulation, we can move towards the ‘more nuanced understanding of power resources’ that Refslund and Arnholtz (2022: 1973) are calling for. This reframing allows us to address our central question – about how unions exercise power in fissured workplaces – in ways that speak to the fundamental questions raised by power resources theory.

Research methods

A project such as this, which draws out multiple and perhaps inherently conflictual responses to understand in what precise ways union power manifests itself in a fissured workplace, is best served by a case study design. More specifically, the opportunity to observe, ask questions, test responses and refine conclusions is only available through a field research methodology. Understanding the interplay of power sources as unions become powerful in a fissured workplace would be difficult to do simply using a prescribed set of survey responses or numeric data. A case study like this is suited to theory-building and to the examination of under-explored phenomena because it ‘investigates a contemporary phenomenon within its real-life context; when the boundaries between phenomenon and context are not clearly evident’ (Yin, 2003: 13). It is difficult to imagine an analysis of the exercise of the multiple dimensions of union power without being able to contextualise each insight within the specifics of the supply chain being studied.

As with most industrial relations case study research, interviewing of those unionists closely associated with a particular campaign was central to the research. Participants were recruited using purposive sampling (Creswell, 2014) to ensure that they had useful knowledge of the research issues. After initial enquiries and informal discussions, 12 people were invited to participate in the interviews, 10 of whom responded and participated. The sample included five staff members of the relevant union and global union federation, and five union members who had been chosen by their peers to have a leadership role in this campaign. Interviewees were selected on the basis of initial campaign documents provided by the union, and interviews were all conducted in 2021.

Interview participants were approached by email or SMS (as most appropriate for them) and invited to participate in the study. Due to Covid-19 stay-at-home orders, none of the interviews was conducted face-to-face. The five officials were interviewed independently using video-conferencing or telephone. The five workplace representatives were interviewed independently via telephone as they did not have access to, or familiarity with, the video-conferencing platforms. Interviews were between 20 and 60 minutes in length depending on the campaign role of the participant. Interview questions were developed from the literature (Bryman and Bell, 2015), focusing on the interviewees’ understanding of union purpose and power.

Although a sample size of 10 participants is relatively small, it is methodologically sound in this instance for three reasons. First, interviews were supplementary to written materials; that is, they were not the sole source for research. Second, 10 interviews can be sufficient to achieve saturation. Morgan et al. (2001) found that 80–92% of new themes were found in the first 8 to 10 interviews (see also Hagaman and Wutich, 2017; Namey et al., 2016). So it proved in this particular study, where the first six interviews brought to the fore the issues of ‘power within’ before ‘power between’ and the remaining interviews confirmed this finding from each of the different key campaign roles. This sample size also reflects the specialised nature of the campaign. The officials interviewed encompass the totality of the roles involved in planning a campaign from this specific organisation. Similarly, the workers interviewed represent a spread of active members from across roles and shifts at the De Costi plant. ¹

Document analysis was used to triangulate (Myers, 2013) the interview data. There were five key areas of written data. The first was Australia’s Fair Work Act 2009, the primary legislation regulating workplaces. The second set of data was drawn from the workplace agreements setting the terms and conditions at the specific worksite (Tassal, 2021a) and other Tassal Group sites (Tassal, 2018, 2020, 2021b) as well as from the ‘Award’ which, as for all workers, prescribes minimum wages and conditions for an industry (FWC, 2020). The third set of data available was extensive campaign materials. These allowed for a contemporaneous assessment (Myers, 2013) of the interviewees’ reasoning about the sources of power available to them and the relative strengths of these powers. Importantly, these documents provide a synchronous history of when shifts between the type of power being exercised were planned and when such shifts were reactions to external circumstances (Nowak, 2022). The fourth set of data was a series of leaflets prepared by the union for both the worksite and for its public campaigning. The worksite material was largely focused on industrial trends such as comparative wages and conditions (AWU, 2019), whereas the public campaigning material was focused upon the moral imperative for the employer not to engage in ‘wage theft’ (AWU, 2020). The final data set comprised media and business analyst reports about the nature of the industry and the company in particular. These reports were important in ensuring that the assessments of the setting being offered by the union were consistent with other publicly available information.

Any research has limitations, and in this case the study focuses on a small set of union activities. These limitations are countered by emphasising that the research question is an exploratory one that seeks to develop theory that can be tested in other contexts. Conclusions can be reached without overstating the significance of the results (Myers, 2013). This tension between achieving a breadth and a depth of data is an inherent characteristic of case study research (Clibborn, 2019). Whilst a case study comparing all unions undertaking this type of campaigning might have provided a larger sample size, the robustness that comes from being able to interrogate the participants’ experiences under very specific circumstances would have been lost. The overarching purpose of the article is to contribute to theoretical debate about how unions exercise power. That is the object of the article; the case study is a means to that end.

The research in context

Union strategy, power and purpose are of course shaped by the structural and institutional power of other actors. Union capacity to organise and bargain is, as elsewhere, shaped by the regulatory regime. This national regime consists of both high levels of state intervention and less support for unions and collective bargaining than was the case for most of the twentieth century (Ellem and Cooper, 2020). In Australia, minimum labour standards are prescribed by the state and between employers and employees through bargaining. The state’s mechanisms aim to cover all employees. This is done through the national parliament setting ‘National Employment Standards’ and the national labour tribunal, the Fair Work Commission, setting industry-specific minima through ‘Modern Awards’. Between them, these regulations cover about 20 items including wages, ordinary hours of work, various forms of leave, flexibility arrangements, superannuation (pensions) and dispute-settling procedures A system of ‘enterprise level’ collective bargaining sits atop these foundations in some though by no means all workplaces. This system aims to encourage worksite specific efficiencies for employers and (therefore) higher wages for employees than the state-provided minima. In this national setting, then, ‘minimum labour standards’ has a very specific meaning: the conditions prescribed by the relevant parts of this tripod of regulation. Who (or what) enforces these standards remains a fraught question, as our theoretical framing has suggested. Historically, this was the responsibility of unions but their decline makes this problematical while state bodies are underfunded.

The bargaining regime of the Fair Work Act 2009 is highly prescriptive, and, apart from ‘good faith’ clauses, it appears to have been developed absent a consideration of the powers of the respective actors. Bargaining must be initiated using a specific set of words issued by an employer to employees but not necessarily their union (Fair Work Act, 2009: ss 173–174): non-union collective bargaining is a distinct possibility (Bray et al., 2020). An employer can choose to commence bargaining at any time with a group of employees if there is no enterprise agreement in place but employees wanting to bargain must win a Majority Support Determination (Fair Work Act, 2009: ss 236–237). If employees wish to take industrial action, they can only lawfully do so in limited circumstances: their enterprise agreement must have passed its nominal expiry date, they must all be union members, they must adhere to the prescribed ballot process, secure 51% of eligible voters to participate, and 51% of participants must vote in favour of action (Fair Work Act, 2009: s 181). In comparison, an employer seeking to make an agreement only has to provide it to employees for consideration for seven days and provide information on the content. The vote will be decided by a simple majority of those participating (Fair Work Act, 2009: s 181). If that vote concludes in the affirmative, an agreement is made. Once such agreements are in place, no worker can be asked to work for lesser conditions than written in the Agreement: that is, it overrides the relevant Award.

The institutional and structural power of Australian unions has been undermined for the best part of a generation through legislative change, employer strategy and changes in labour markets. Despite its stated aims, the Fair Work Act has not reversed the decline in collective bargaining. Associational power has been attenuated; membership density is only 14.3% (ABS, 2020) down from well over 60% at its peak (Bain and Price, 1980: 121–125). Nonetheless, unions remain significant institutional actors. Although collective bargaining coverage has fallen, and while unions are not the agents in the making of Modern Awards, collective bargaining and award coverage provide collective regulation for 58.9% of workers (ABS, 2019).

In this national setting, the case study examined a supermarket supply chain, an area of work where, typically, union power is even more circumscribed than in the aggregate. Supermarkets have attracted a great deal of research because of their buying power and capacity to control their supply chain while employing very large numbers of workers. Likewise, their suppliers have attracted much research (Hardy and Howe, 2015; McKinsey & Company, 2022).

In Australia at the time of the research, retail accounted for 62.4% of the purchases in the seafood processing sector and 28.4% of that work was performed by this particular processor (IBIS World, 2021). The structure of these supply chain arrangements is based on supermarkets using direct purchase arrangements with suppliers as opposed to working through a wholesaler (IBIS World, 2021). Seafood processors achieve their market share through vertical integration from farming through to retail (IBIS World, 2021). The priority of these processors is scale in the domestic market (how many canned products they can sell) and sale of premium products in the Asian market (IBIS World, 2021). Although the processor has a degree of value to the supermarkets based on the volume of product they provide, that also makes the supermarket their single most important client.

This study focused on Tassal, an interstate seafood processor with a plant in Lidcombe, a suburb in the city of Sydney in the State of New South Wales. The Lidcombe plant’s most important buyer in this study was Coles, one of the two largest supermarket chains in the country. The level of control over the Lidcombe plant exercised by the corporate structure of the Tassal Group was high. Reflecting on how decisions are made, one participant stated that ‘we worked with the union and then each part of our company, and then the main person from Tasmania’ (Worker 5). Official 3 added: ‘they’re expanding into everything. They have complete, like what I mentioned at the start, vertical integration of everything’ (Official 3).

At the time of the study, the Lidcombe site employed approximately 200 workers. The particular union entitled by labour law to cover the site was the Australian Workers’ Union (AWU), a large, general union of blue-collar workers. The factory became an organising priority for the union because it wanted to ensure consistency with other sites in the same company where there were higher levels of membership, well-developed workplace union structures and higher wages. Official 1 explained the focus on the site in these words:

Tassal had bought [the Lidcombe site] from De Costi roughly 12 months or few years prior, that we [the union] hadn’t identified in Sydney, knowing that as a national union, we have some good organisation and structure down in their main facilities in their farms and processing facilities in Tasmania. (Official 1)

The campaign was also a question of market significance reflecting the firm’s structural power: ‘They’re the largest seafood company in the country. And they’re expanding into everything’ (Official 3).

In the Lidcombe plant, between 70 and 95% of the workforce did not speak English as their first language (Official 3; Worker 5); many had temporary (that is, short-stay) visas (Officials 3, 4; Worker 6); a significant proportion of workers were engaged casually (Officials 3, 4). In short, this was a workforce that, at first sight, lacked power. Indeed, participants noted that ‘95% of the workforce there speak English as a second language. Predominantly Nepalese and Vietnamese, some Pacific Islanders. They didn’t know what a union was. They didn’t know what their rights were’ (Official 3).

Workers had limited institutional power resources to enforce the Modern Award’s prescribed minimum industry conditions. As low-wage, culturally and linguistically diverse workers, casually employed, on temporary visas, they also lacked structural power resources. At first, they had no union structure or other form of associational power. This also made it difficult for these workers to exercise societal or ideational power because the same factors also made it hard for them to engage in coalition-building or storytelling within (let alone outside) their immediate workplace. The result was that the union and workers were forced to consider each of the power resources normally available to them relative to the other power resources (or lack of resources) available to them to achieve their purpose before considering the relative value of that specific power resource to each actor.

Union purpose and the exercise of power

There were five stages of the AWU’s campaign, unfolding against the (supposedly guaranteed) minima set by the Seafood Processing Award 2010 and then 2020. The first stage was the attempt to build associational power at the workplace beginning in 2019 with no members and is ongoing today albeit with significant growth in union density in that time (Official 3). The second was the underpayment of shift loading for workers who began work as early as 1:30 a.m. That dispute began in 2019 and a Federal Circuit Court decision in favour of the workforce was handed down in 2022 (after the research was concluded). The third was the decision to commence enterprise agreement negotiations at the site and in turn protected action preparations (Official 1). This stage included outreach by the global union federation to the Tassal Group (Official 10). The fourth was a decision to undertake public actions at the Sydney Fishmarkets, Coles’ distribution centres and Coles’ retail stores to force the lead-firms to act on labour standards at the Lidcombe plant (Official 1). The fifth was the approval of the enterprise agreement in May 2021 (Tassal, 2021a) and the AWU’s subsequent work to use this win to engage other structurally or geographically similar workers online (Official 4).

The majority of the workers at the Lidcombe plant were union members after this campaign (Official 1, 3; Workers 6, 7) despite structural power seeming to be against them. The specific outcomes to the campaign were significant in terms of the goals that had been set. First, the union secured an enterprise agreement with 2.25% pay increases (at the time, above inflation) and improvements to shift loadings and allowances (Tassal, 2021a). Second, the agreement contained delegates’ rights and training provisions (Tassal, 2021a). Third, there was a favourable decision in the Federal Circuit Court.

The campaign began with union trade union officials visiting the Lidcombe plant. They recognised that in the context in which they were working – with apparently limited structural power and institutional power resources – building associational power was both a strategic goal and an organisational necessity: ‘long-term we want to build capacity in the workplace. So the EA [enterprise agreement] and having delegate structures and stuff like that is much more important’ (Official 4).

It is clear from all the interview participants that building associational power was difficult. As Official 10 notes: ‘it’s an industry which is hard to organise’. There were limited members at the site and the organisers were tasked to recruit workers in six languages (Official 1): ‘It was difficult to sign people up at the start because most of them thought that if they joined the union, that the company will sack them, even the permanent employees’ (Official 3). Worker 9 expands on this: ‘a lot of them joined the union and then they left, and then some of them re-joined. Because I don’t think they realised what the union can do for them.’ A recurring theme from all the interviewees was the level of fear that workers had that unionising would lead to them losing work hours: ‘I think maybe it’s the age, that they have the fear factor’ (Worker 9).

Through employer-facilitated mass meetings during bargaining for a new agreement, membership growth did occur. Interestingly, trade union officials attribute that growth to a shift from offering an individual a means by which to access their institutional power to raising expectations of what associational power could achieve:

When bargaining occurred, it was a different kind of conversation. It was well, ‘now you’re operating as a real collective here. The strength of what we have is the strength of what we will get in negotiations, and this is your real opportunity.’ (Official 1)

This is a variation of an emerging theme from the literature about the need to transform sources of power for unions to be effective on an ongoing basis (Ellem et al., 2020; Hui, 2022; Nowak, 2022). The comment by the official suggests that moving to associational power is the goal and institutional power a tool to achieve that. Usually in the literature the objectives are cast as the reverse of this. Implicit in the existing analysis is that unions need to transform societal and associational power resources into institutionally protected rights (Chen, 2003; Ellem et al., 2020; Hui, 2022; Schmalz et al., 2018). The problem with this argument is that it assumes a superior benefit to exercising institutional power resources without recognising the very real risks and limitations of institutional reliance.

There are two potential outcomes of transforming associational power into institutional power. The first is that institutions are changed because the parties engage with equivalent levels of power and so ‘institutional conversion’ (Chen, 2003) occurs. This is where institutions become a site of contest and resistance. Alternatively, the transformation can lead to a loss of associational power without rebalancing access to institutional power because those institutions become a mechanism by which the rights of employees are not enhanced but, rather, obscured, confused and removed by complex processes that sit outside the influence or understanding of workers (Peetz and Pocock, 2009).

A useful observation on the limitations of this approach of focusing on institutional power was made by one of the union officials who had overseen the campaign:

It seemed like the company was saying ‘we’ve done this for a very long time’. But it was a non-union site. I mean, employers can run and do things for a long time that aren’t lawful, if there’s no one to hold them [the employers] to account. (Official 4)

Official 4 makes a critical observation. These workers theoretically had institutional power (the Award legally mandated their additional payments). However, it was only when they developed associational power that they were able to give effect to their institutional power and engage with those institutions on an equivalent basis (legal counsel and so forth) as their employer. Institutional power resources without a functioning enforcement mechanism have limited practical use to regulate labour standards in a fissured (or any) workplace. Implicit in the comment that ‘there’s no one to hold them to account’ (Official 4) is a repudiation of the assumption in much of the literature that institutional power resources can function as an autonomous arbiter of labour standards despite the consistent underfunding of industrial relations institutions (Weil, 2014), the narrow priorities of those institutions (Kaine and Rawling, 2019) and the shifting sets of institutions, actors and interests able to be engaged (Weil and Mallo, 2007).

The union was also concerned that the legislated Award minimum standards were not being properly paid for work commencing as early as 1:30 a.m. Workers were being paid ordinary hourly rates, but the union believed they should be being paid overtime rates. Underpayment is not uncommon in supermarket supply chains, where contracting practices leave a small margin for labour costs (Wright and Kaine, 2015). The interview participants had a consistent view that this was a deliberate undercutting of accepted standards, summarised as: ‘It was systemic. It was not an oversight. It was pretty deliberate. You don’t accidentally underpay AU$7 million’ (Official 2).

Despite the workers’ emerging associational power at the Lidcombe plant, management’s engagement in the institutionally prescribed collective bargaining processes was viewed by trade union officials and workers as perfunctory and a way to avoid meaningfully addressing the issue of Award underpayment. ‘Our log of claims was well . . . look . . . they’ve got to abide by good faith bargaining so it was tokenistically looked at and some things we got across the line’ (Official 1). This was a source of frustration for the workforce as ‘they [the employer] were making money; they just weren’t giving it to the workers’ (Worker 7). Instead ‘they were just trying to hide or just not responding as quick to the union and then to us as well’ (Worker 5). These comments suggest that these institutional rules are difficult to enforce without strong associational power resources. Coupled with the lack of supply chain or vicarious liability provisions in the relevant Award (FWC, 2020), the cumulative effect is an incongruity between employer structure and regulatory institution reach. The impact of this fragile institutional framework at the Lidcombe plant was observed succinctly by one of the participants:

This whole good faith resolution is very few and far between. Companies have business models based on knowing what they have implemented is immoral and is not how laws are written. But if they can legally justify it, then they’re fine doing that. (Official 1)

The status of these workers as culturally and linguistically diverse migrants also created a practical impediment to exercising associational power. As one of the trade union officials notes: ‘There are about six major nationalities identified, and some of those activists really were sowing seeds for many months’ (Official 1). This breadth of languages made it difficult for workers to build connections and unity outside their language group. When trying to build an organising committee one interviewee notes:

It’s hard to just explain [to] them, and then [it’s] just broken English, and then we use that other type of English. We make a group and then we just try to convey what we want to pull from the committee, and then he or she convey[s] the messages to others, just like that. (Worker 5)

As so many of the workers at the site were in Australia on temporary visa arrangements, their view that it was inevitable that their work would be low-wage, insecure and poorly paid was reinforced (Official 1; Worker 8).

Compounding their precarious structural position were the workers’ employment arrangements at the site where most were casual (Officials 1, 3; Workers 5, 6, 7, 8, 9): ‘most of them, they believe that if they are the union member or they do anything against the company, maybe they won’t call them for long, or for longer work’ (Worker 8). This was confirmed by other worker interviewees: ‘I was scared they would reduce my hours or point the finger at us, you know’ (Worker 6).

Despite the workforce having low structural power resources due to their characteristics and limited institutional power resources due to the gaps in the legal framework to regulate fissured workplaces, these same characteristics did now present an opportunity to use ‘ideational power’. The trade union officials were clear that there was a ‘moral’ (Official 1) story to be told in this campaign. As Official 2 sets out:

What we were kind of wanting to explain was they were basically exploited and taken advantage of. They were migrant workers, who didn’t have a great command of English, and essentially, I don’t think you would have been able to treat Australian workers like that, if you know what I mean, and get away with it. And so that was an exploitation of people who were not aware of their rights. And when they were made aware of their rights and they tried to go to Tassal about it, Tassal were like . . . ‘We’re not interested’. (Official 2)

In embracing the lack of structural power resources available to these workers (Schmalz et al., 2018) the union was able to recast the negotiations from being a contest that should be balanced to an injustice that it was morally imperative to fix. This finding reflects the need to analyse power not as a contest between parties but first as an interplay between power resources available to actors, in this case the union.

The trade union officials report that they are now looking at opportunities to use online tools to undertake this work at scale, targeting workers with a similar experience of (lacking) institutional power to those in the case study:

We also did some online advertising and just a little bit of taking that content we’d already put together and kind of boosting it. We kind of targeted the areas where we knew that those members were working and living. (Official 2)

This section has explored the data from the primary research in the context of some of the key stages of the campaign. By analysing the data through the conceptual framework of the power resources approach, this research is able to move beyond the familiar discussions on the power of industrial relations institutions to start to consider the power that workers and unions have to utilise those institutions effectively, be they supportive or hostile, by considering those power resources as a part of a suite of options available to an actor before considering the relative value of that specific power resource to each actor.

Conclusion

The decline of state support for union-based bargaining and the fissuring of the workplace have often eroded working conditions and undermined workers’ power. We have tried to rethink what this set of problems means for unions trying to represent workers in fissured workplaces by revisiting power resource theory to locate the exercise of union power more explicitly in terms of union purpose and scholarship on regulation. We apply this analysis to a site typical of these forms of work, a supplier to a major supermarket, and to workers who were at first sight typical of the ‘unorganisable’. Our findings in both empirical and conceptual terms are surprising. They open the way for further research and, as most power resource theorists seek to do, for fresh assessments of union strategy.

In this particular case, the union’s purpose was threefold: to enforce existing rights, improve pay and set up worker enforcement mechanisms through a delegate structure. Although Refslund and Arnholtz (2022) identify five sources of power (structural, associational, institutional, ideational and coalitional) this article has focused on the first four as they were the power resources utilised by the union that was the focus of the case study. In examining how the union chose which power resources to use and how that power was exercised to achieve those purposes, three key findings emerge.

First, conceptually, the assumption that power is measured relative to other actors misses the necessary prior step that unions must have an understanding of the various power resources available to them that are within their control. This leads to the second finding about how those power resources are utilised. Through the case presented in the article, it is shown that power resources are not deployed in isolation. These workers had been experiencing a breach of their legally protected labour standards for many years, but only when they built their associational power did they have a practical mechanism to access the institutions able to address this issue. In a strategically similar practice, the workers lacked structural power, but, through their union, they turned their vulnerable status into a powerful form of ideational power based on meta-narratives of injustice. Our third finding follows from this: the need to utilise multiple power resources repudiates the power resources orthodoxy that unions need to transform associational and societal power into institutional rights in order to use that power.

In setting out to address an important question for industrial relations scholars and union activists and leaders about how unions exercise power in fissured workplaces, the research took us to the challenge posed by Refslund and Arnholtz (2022: 1973) to provide a more nuanced and critical analysis of union power in general. Our three major findings rest on refining the power resources approach, particularly in terms of how we think about the power resources available to unions and how they exercise power to achieve specific purposes. Union powers are rarely used in isolation: in particular, institutional power can only be built if strong associational power and structural power can be asserted through the mobilisation of ideational power. In the spirit of much of the scholarship in this field, we have shown that this rethinking is not merely theorising for its own sake. Rather, it has the potential to speak to union practice and to questions of union purpose and power more generally.