• Specific Year
    Any

Orchiston, Alice; Creighton, Breen; Denvir, Catrina; Johnstone, Richard; McCrystal, Shae --- "Pre-Strike Ballots and Enterprise Bargaining Dynamics: An Empirical Analysis" [2019] MelbULawRw 7; (2019) 42(2) Melbourne University Law Review 593

PRE-STRIKE BALLOTS AND ENTERPRISE BARGAINING DYNAMICS: AN EMPIRICAL ANALYSIS

Alice Orchiston,[1]* Breen Creighton,[1]** Catrina Denvir,[1]† Richard Johnstone[1]

and Shae McCrystal[1]‡‡

Under the model of enterprise bargaining enshrined in the Fair Work Act 2009 (Cth), the primary mechanism for employees to exercise industrial power during negotiations in respect of future wages and conditions is protected industrial action. Access to protected industrial action is contingent upon the employees and their bargaining representatives having complied with a series of statutory prerequisites. The most significant of these is the requirement to hold a pre-strike ballot of the relevant employees to authorise the proposed action. Without approval in a ballot, any industrial action undertaken will be unprotected, leaving the union open to liability under common law and statute, and its members potentially subject to dismissal from employment. This article explores the effect of pre-strike ballots on union decision-making and enterprise bargaining. Drawing on statistical analysis of a cross section of protected industrial action ballot applications made to the Fair Work Commission over a period of 12 months, and grounded theory analysis of interviews with ballot applicants, employer respondents and key stakeholders, this article evaluates whether the protected action ballot order regime in practice provides a ‘fair, effective and simple’ mechanism for ascertaining employee support for industrial action, as suggested when the provisions were first introduced. The discussion explores how the introduction of pre-strike ballots has created new opportunities for unions to exert pressure in collective bargaining but at a considerable cost in terms of the use of union resources and in providing opportunities for employers to delay access to industrial action. It also considers whether there is a better process for allowing union members to indicate their support for industrial action, in order to remove the administrative burden imposed on the Fair Work Commission, employers and unions by the existing model of pre-strike ballot regulation.

Contents

I Introduction

Under the model of enterprise bargaining enshrined in the Fair Work Act 2009 (Cth) (‘FW Act’), the primary mechanism for employees to exert industrial leverage during negotiations in respect of future wages and conditions is ‘protected industrial action’.[1] Access to such action is contingent upon the employees and their bargaining representatives having complied with a series of statutory prerequisites. The most significant of these is the requirement to hold a pre-strike ballot of the relevant employees to authorise the proposed action.[2] Without approval in a ballot, any industrial action undertaken would be unprotected, leaving the union open to liability under common law and statute, and its members potentially subject to dismissal from employment.[3]

The requirement to hold a pre-strike ballot was introduced into the federal enterprise bargaining system in 2006.[4] Prior to this, a decision by a trade union to take protected industrial action was made in accordance with the rules of the union itself.[5] However, when mandatory pre-balloting was introduced, an entirely new regulatory step was added to the existing enterprise bargaining structure. Instead of being a matter for the union to resolve, the legislation required that it seek permission to ballot from the federal industrial tribunal and, if such permission was granted in a ‘protected action ballot order’ (‘PABO’), to engage an independent ballot agent to conduct the ballot.[6] From 2006 onwards, the application to ballot, the ballot process and the ballot results have been public,[7] and employers have been able to mount legal challenges to the grant of ballot applications.

This shift from private to public regulation of trade union decision-making altered the context in which unions could access protected industrial action. This altered the dynamics within which enterprise bargaining takes place, entailing significant changes to: the processes that apply where a trade union seeks to exercise coercive pressure; the information available to employers about union decision-making; and the ability of the union to decide to take or not to take industrial action. Such changes must inevitably have a significant impact upon negotiations over working conditions.

This article empirically explores the effect of pre-strike ballots on union decision-making and enterprise bargaining, drawing on statistical analysis of a cross-section of protected industrial action ballot applications made to the Fair Work Commission (‘FWC’) over a period of 12 months, and grounded theory analysis of interviews with ballot applicants, employer respondents and key stakeholders. The empirical findings are used to evaluate whether the PABO regime in practice provides a ‘fair, effective and simple’ mechanism for ascertaining employee support for industrial action, as was suggested by the government of the day when the provisions were first proposed.[8] The article demonstrates that the introduction of pre-strike ballots has created new opportunities for unions to exert pressure in collective bargaining, but at a considerable cost in terms of the use of union resources and in providing opportunities for employers to delay access to industrial action.

The article begins by outlining the existing literature concerning the impact of pre-strike ballots on enterprise bargaining dynamics and the relevant legislative context. The methods of the study are then set out. This is followed by an analysis of the effects of pre-strike ballots in respect of bargaining leverage, internal union democracy, administrative and resource burdens, and employer responses to PABO applications. Finally, the primary findings of the study are outlined and discussed.

II Pre-Strike Ballots and Enterprise Bargaining Dynamics

There is a considerable body of literature examining the dynamics of enterprise bargaining in Australia and elsewhere. This shows that a range of factors can influence power relations between the parties, such as competitive pressures impacting upon the employer, the history of the negotiations, and whether or not the relationship between the parties is cooperative or antagonistic in character.[9] Studies of the Australian system have explained how the structure of the regulatory regime, the enterprise focus, and the shift to employer control over initiation of bargaining and agreement-making processes, shape the power dynamic impacting decision-making.[10] Further, there is a substantial literature on the economic impact of industrial action,[11] and patterns of industrial action activity in Australia.[12] The focus of this article, however, is on how mandatory pre-strike ballot requirements shape bargaining dynamics — irrespective of whether the parties actually engage in industrial action.

There is extensive literature exploring the public policy motivations for the introduction of mandatory pre-strike ballots in collective bargaining regimes.[13] This literature is particularly well-developed in the United Kingdom (‘UK’), where pre-strike ballots were introduced in 1984, and in relation to which compliance has progressively been made more complex and challenging.[14] In particular, the literature contrasts the public rhetoric that usually accompanies the introduction of mandatory balloting regimes focusing on making unions more ‘democratic’ and ‘accountable to their members’, with implicit underlying policy motivations of reducing industrial action and eroding the collective basis of union membership.[15]

In contrast, there has been limited examination of how pre-strike ballot regulation impacts upon both internal union processes and enterprise bargaining dynamics.

In both the UK and Australia, unions are required to hold pre-strike ballots before engaging in lawful industrial action. In the UK, the union itself conducts the ballot in accordance with stringent statutory rules and the involvement of independent scrutineers,[16] whereas in Australia, the ballot is authorised by a regulator and is conducted by an independent third party.[17] Research on how the UK requirements have shaped trade union behaviour suggests that UK unions have effectively internalised those requirements, and generally accept that industrial action ‘should be preceded by a ballot’.[18]

Early UK research drew some positive conclusions concerning the response of unions to the ballot requirement, and the impact on bargaining dynamics. It showed, for example, that the pre-strike ballot requirement had driven efficiencies within unions, leading to the maintenance of better member records and the adoption of more streamlined decision-making processes.[19] In addition, the push by unions to ensure voter turnout in ballots led to better lines of consultation and communication between union officials and members.[20] The evidence also suggests that mandatory pre-strike ballots had the effect of legitimating trade union decisions to take industrial action, countering suggestions from employers or other stakeholders that unions make strike decisions without consultation with members.[21] This had the flow-on effect of enabling trade unions to use the threat of industrial action more credibly in bargaining negotiations, where such threat was backed up by the demonstrated resolve of the membership.[22] However, from a union perspective, some negative consequences were also observed — in particular, the imposition of the costs and delay associated with balloting, the increased opportunities for employers to challenge the legitimacy of pre-strike ballots,[23] and the adverse impact on a union’s bargaining position of failing to approve proposed industrial action.[24]

More recent research has suggested that following the expansion of pre-strike ballot requirements in the UK, union compliance with the legal provisions has become more uncertain and the ballots provisions have remained ‘fertile ground’ for employer litigation.[25] Furthermore, Elgar and Simpson have suggested that the ability of unions to operate under the UK pre-strike ballots regime has come ‘at a considerable cost of diversion of resources in order to ensure that they are able to stay within the parameters of legality which have been persistently redrawn’.[26] This suggests that some of the positive benefits for unions of the mandatory balloting requirements may have been offset in practice by the diversion of resources to ensure compliance with the ever-changing regulatory scheme.

In Australia, there has been only limited research into how the FW Act ballot requirements shape union decision-making and bargaining dynamics.[27] The 2012 review of the FW Act suggested that the ballot provisions ‘have a significant impact on employees and their union representatives’,[28] but did not focus on the process with sufficient precision to make it possible to draw more detailed conclusions. The Productivity Commission inquiry in 2015 also examined the ballot process,[29] and although it suggested some avenues for simplification,[30] the scale of its analysis was limited.

Early doctrinal research on the introduction of pre-strike ballots in 2007 by Orr and Murugesan concluded that ‘mandatory secret ballots add cost and delay ... and provide an additional avenue for employers to intervene and object’.[31] The authors noted that the ‘chief practical effect’ of the pre-strike ballot regime would be to ‘give employers additional notice of likely industrial action, and to enhance enforcement of a one-sided, quasi-good faith bargaining regime on unions’.[32] At the same time, the authors suggested that ballots were advantageous to unions because ‘merely initiating the ballot process may in itself be a bargaining tactic’, due to the fact that ‘a strong “yes” vote signals the employees’ determination’.[33]

Overall, the research in both the UK and Australia suggests that, from a trade union perspective, legislated ballot requirements have both positive and negative effects on enterprise bargaining dynamics and union decision-making. This is borne out by the examination in this article of the ways in which the pre-strike balloting requirements in the FW Act have shaped bargaining dynamics in enterprise bargaining, how trade unions have responded to the increased regulatory burden imposed by the legislation, and the extent to which employers have resisted union access to protected industrial action through utilising legal channels to oppose pre-strike ballots.

III Legislative Context

The mandatory pre-strike ballot requirement in the federal system was originally proposed in a 1998 discussion paper released by the then Minister for Workplace Relations and Small Business, Peter Reith. According to the Minister, mandatory secret ballots

provide a fair, effective and simple process for determining whether a group of employees in an enterprise want to take industrial action ... can help to improve the arrangements for bargaining; strengthen the accountability and responsiveness of unions to their members; and minimise unnecessary industrial disputation.[34]

When the Coalition gained control of the Senate in 2005, the Howard Government was able to secure passage of the Workplace Relations Amendment (Work Choices) Act 2005 (Cth) (‘Work Choices’), which included a mandatory ballot requirement.[35] The Australian Labor Party (‘ALP’) took office in 2007, committed to the repeal of Work Choices. However, despite the ALP having opposed the imposition of mandatory pre-strike ballots while in opposition,[36] the Rudd Government’s Fair Work Bill 2008 (Cth) retained the Coalition’s provisions in a largely unaltered form.[37]

Those provisions are now set out in pt 3-3 div 8 of the FW Act. Employees and their bargaining representatives are able to engage in ‘industrial action’, which is defined in s 19(1) to include strike action and lesser forms of action including bans, etc,[38] provided the statutory prerequisites have been met

and the action is taken in support of negotiations for a single-enterprise agreement.[39]

As noted earlier, engaging in industrial action without approval through a balloting process would expose those who organise or participate in the action to liability at either or both of common law and statute.[40]

The pre-strike ballot requirements involve a two-stage process. First, a bargaining representative of the employees who will be covered by the proposed enterprise agreement must apply to the FWC for a PABO.[41] Such application may not be made more than 30 days before the nominal expiry date of any existing enterprise agreement applicable to the relevant employees.[42] The FWC must make such an order if the application has been made in accordance with the FW Act,[43] bargaining for the agreement has commenced,[44] and the bargaining representative has been, and is, genuinely trying to reach an agreement with the employer of the employees to

be balloted.[45]

The second stage involves the conduct of the ballot. Once a PABO has been made by the FWC, the secret ballot must be conducted by an independent ballot agent. This functionary may be either a person or body, selected by the bargaining representative and approved by the FWC or the Australian Electoral Commission (‘AEC’).[46] The mode of ballot may be postal, attendance or electronic.[47] The conduct of the ballot, and the preparation of the electoral roll, is the responsibility of the ballot agent.[48] Where the ballot agent is the AEC, the bargaining representative may have input into the mode of ballot used;[49] however, the final decision rests with the AEC. Where a ballot agent other than the AEC is used, the FWC must provide the ballot agent with detailed instructions on a range of matters, including the ballot mode.[50]

For proposed industrial action to be approved, a quorum of at least 50% of those eligible to vote must do so, and at least 50% of those who vote must cast a valid ballot and vote in favour of the proposed action.[51] Where a ballot seeks to authorise more than one form of industrial action, each form of action must be separately approved in the ballot. This means that it is possible for a ballot to propose a range of forms of industrial action, but for only some forms of action to be approved. Once approved in the ballot, the balloted employees may take protected industrial action, provided that they comply with other statutory provisions relating to notice, form of action and timing.[52] Importantly, if action is approved in the ballot, it must be commenced within 30 days (or up to 60 days with the approval of the FWC).[53]

IV Methodology

This study seeks to investigate the effect of pre-strike ballot regulation on union decision-making and on the dynamics of enterprise bargaining negotiations. The analysis addresses the following research questions:

1 Does the PABO regime confer opportunities for unions or employers to gain leverage in bargaining negotiations?

2 How does pre-strike ballot regulation influence internal union decision-making processes?

3 What administrative requirements does the ballot regime impose on unions, and how onerous are these requirements?

4 How frequently do employers oppose PABO applications, and on what grounds (technical or substantive)?

In order to answer these questions, this article draws on mixed methods empirical research, involving analysis of qualitative semi-structured interviews with PABO participants and other stakeholders, and analysis of quantitative data drawn from a database of PABO applications. The empirical findings are then used as a basis for evaluating whether the PABO regime is, in practice, a ‘fair, effective and simple’ means of ascertaining employee support for industrial action.[54]

To address the questions relating to leverage, internal union decision-making and the administrative burden imposed by the PABO requirements, qualitative data was drawn from 74 in-depth, semi-structured interviews undertaken with applicants and respondents to PABO applications submitted during the study period.[55] The 74 PABO participant interviews consisted of 16 pairs (where both one union and one employer were interviewed in relation to the same negotiation), one triple (in which two unions and one employer were interviewed in relation to the same negotiation), nine single employer interviews and 30 single union interviews.

Participants were asked to share their experiences of engaging with the regulatory framework. The questions were designed to explore:

• factors influencing decisions to apply for PABOs;

• perceptions of the state of negotiations at the time of the application;

• perceptions of why applications and consequential pre-action strike ballots are successful or unsuccessful; and

• factors influencing the time between an application to the FWC, the conduct of a pre-action ballot and any subsequent industrial action.

The 74 interviews conducted related to approximately 713 PABO applications (54.76%) submitted during the study period. Interviewees typically spoke ‘globally’ about negotiations, rather than confining their comments to specific applications. In total, the interviews covered 55 sets of negotiations

(or 13.78% of the total number of negotiations) for which one or more

PABO applications were submitted during the July 2015 – June 2016 period. Participants were not individually tracked after the conclusion of the data collection period to ascertain how many had reached agreement.

Additional qualitative data were drawn from 11 interviews with stakeholders including representatives of leading employer and employee associations, members of the FWC and senior AEC staff with

responsibility for industrial elections. The interview questions were designed to explore issues relating to the PABO regime, drawing on respondents’

specific expertise.

Participants were contacted via email in the first instance. Where permission was granted (as occurred in all but three cases), interviews were recorded via digital voice recorder. All audio recordings were later transcribed in full for analysis. Interviews were generally of approximately 40 minutes

in duration.

The interview data was coded thematically using a modified grounded theory approach. Grounded theory is an inductive method that uses the research data itself as a starting point for generating themes through which the data can then be coded and analysed.[56] Traditionally, grounded theory research starts with ‘open coding’ — a process where the researcher keeps an open mind as to ‘all possible meanings’ in the data.[57] Since this study considered the relationship between a specific area of regulation and human behaviour, it was necessary to adapt an iterative approach to reflect this from the outset. To do this, initial coding categories were developed with reference to general features of the PABO process such as the application form, the ballot mode used, employer response, and communication with members. These broad categories were then progressively refined into sub-categories, and ultimately extrapolated into overarching conceptual categories.[58]

To determine how frequently employers object to PABO applications,

and associated grounds for objecting, quantitative data was drawn

from a database developed by the authors capturing information

relating to all PABO applications submitted to the FWC during

the period 1 July 2015 – 30 June 2016 (n = 1,302) (‘the study period’). The database was hierarchical in nature with PABO applications nested

within ‘negotiations’ for single-enterprise agreements. The 1,302 PABO

applications pertained to negotiations in respect of 399 enterprises or ventures.[59] Of these:

• 1,254 (96.4%) were approved;

• 6 (0.5%) were not approved; and

• 42 (3.2%) were withdrawn by the applicant before an order was made.[60]

Of the 1,254 applications that were approved, 1,204 proceeded to ballot.[61]

The remainder of the article discusses the research findings in relation to how mandatory pre-strike ballots requirements shape bargaining dynamics and union decision-making. Lastly, the implications of the research findings in relation to policymaking are discussed.

V Findings

A Ballots and Bargaining Power

Under the logic of voluntary collective bargaining, access to strike action provides unions and their members with industrial leverage or bargaining power as a counterweight to the inherent advantages employers derive from managerial prerogative and control of private property. As Weiler observes, it is ‘the strike that determines which side will find it more painful to disagree, which party will be forced to make the major moves toward compromise’.[62] However, what is clear from the analysis of the interview data in this study is that mandating a series of publicly accountable steps before strike action can occur creates new opportunities for unions to exert pressure in bargaining.

Our analysis found that a PABO application can be an effective means for unions to increase leverage (via escalating the ‘threat’ of industrial action), without necessarily having to resort to industrial action. This is because a successful ballot outcome signals industrial strength and accelerates the ‘countdown’ towards industrial action.

In general, the union representatives interviewed perceived the decision to apply for a PABO (with a subsequent ballot), and the decision to take industrial action, as separate and distinct steps in escalating pressure in negotiations. However, respondents expressed a range of different views as to the degree of leverage that a PABO application and/or subsequent industrial action would provide.

Some union representatives explained the power of a PABO application with reference to the risk of ‘disruption’ that employers would face from any industrial action that could logically follow a successful ballot.[63] Other representatives described a successful ballot result as a strategic signal of collective strength. For example:

It’s factored as a message ... the conversation I had with the members is, ‘look to be honest even if you disagree with taking industrial action have a think about voting yes for it anyway because it is the message that counts’ ... in the formal process, the employer sees employees vote ‘yes’ because they will be looking at that and seeing how strong you are.[64]

For some union representatives, the prospect of media coverage, either of the ballot outcome or subsequent industrial action, and the opportunity to ‘get the employer’s name in the paper’, were seen as further potential advantages of invoking the PABO process.[65] For example, one union interviewee said that once the ballot results were announced, ‘it actually made the papers’, and this helped to shift the negotiations in the union’s favour because of the employer’s desire to avoid ‘bad press’.[66] These sentiments were echoed by the employer representative from the same negotiation:

We have an employer brand. We don’t want to be splashed all over the

media ... we don’t want that out there because I don’t think it’s a positive look.[67]

The threat of media coverage can create pressure on employers where industrial action could disrupt production or service delivery. For instance, one employer representative referred to a PABO application that was timed to coincide with negotiations for a major business contract, and indicated that if the customer found out about the threat of industrial action, they would ‘probably cease negotiations’ with the company.[68]

Interviews revealed that some unions are able successfully to threaten industrial action (via the application for a PABO), even where they have no intention of actually taking action. As one union representative explained:

I have to confess one of the greatest things the Liberals ever did for us soft unions was that really lengthy and painful process to take a protected action ballot. That’s actually been one of the greatest advantages for us ... in the olden days you couldn’t bluff about industrial action; you took it or you didn’t take it. You were out the gate or you were not out the gate. This lengthy process of both and application and objection and things like that, it’s just been a godsend for us because you can bluff the whole way through it. Even if there’s no intention of taking action you can look like you’re going to take action. Whereas before the ballot process and all the lengthy things, the first thing the company knew about action was people were marching past the bloody manager’s office out

the gate.[69]

The ability to ‘bluff’ in this manner is highly sector and negotiation specific. For instance, one bargaining representative said that making a PABO application did not have the same effect as it had ‘a few years ago’ due to a ‘downturn’ in the business sector.[70] It also relies on the union being able to get its members to vote for industrial action in the ballot, even if they would not be prepared to take the action. Another union interviewee said that a PABO application was not an effective threat because the employer could easily call the union’s bluff: ‘[I]f you’re using it as a bargaining chip, a bluff, you’ll be found out eventually.’[71]

For the majority of the union interviewees, the PABO process in itself was insufficient to bring the bargaining process to an end or gain a major concession from the employer. Instead, it was perceived as a step in ‘ramping up’ pressure, and a prelude to industrial action:

[Applying for a PABO is a] stage of taking industrial action ... when faced with the threat of taking industrial action, each of those steps [including a PABO application] ramps up the pressure, and ... certainly there has been

occasions where the threat of an application for [a PABO] has been enough to move the employer.[72]

Accordingly, many interviewees observed that they would not pursue a ballot application unless they were confident that the members were committed to actually taking industrial action. For example:

[T]he conversations that we have with the [bargaining] team ... is that they need to make sure that ... even before we go to the PABO application itself, they have people who have committed to take that particular action.[73]

In summary, sometimes a PABO application in itself was enough to break a deadlock in the negotiations or achieve a beneficial outcome for the union. This was also the case for the conduct of the ballot, particularly where there was strong endorsement for the proposed action. However, there were also situations where there was no change in the employer’s bargaining behaviour until after a union engaged in industrial action. Of course, as with industrial action, the effect of a PABO application and ballot will depend on the prevailing economic circumstances, along with the capacity and willingness of any specific employer to resist such pressure.[74]

B Union Tactical Considerations

While a PABO application is a necessary precondition of taking protected industrial action or escalating the threat of industrial action, engaging with the regulatory regime carries with it inherent risks for applicants. There was a general consensus among interviewees that an unsuccessful ballot could damage the union’s bargaining position and undermine its credibility. As one union representative put it, the harm caused by a failed ballot is ‘[g]uaranteed, because it shows ... that the [workers] aren’t prepared to take any

strike action’.[75]

Importantly, interviewees did not distinguish between ballots that were unsuccessful due to a failure to meet the quorum requirement and ballots that failed because the proposed action was not approved. In both instances, there was a perceived adverse impact upon the union’s bargaining position. However, the extent of such impact depended on the circumstances.

In some contexts, a failed ballot may completely undermine the position of the bargaining representative, leading to an employer breaking off negotiations and putting their preferred agreement to a vote. For example, one union representative described the significance of a failed ballot in the following terms: ‘It destroys the credibility. I don’t know how you pick yourself up from that.’[76] Several employer respondents stated that a failed ballot signalled that the membership had lost faith in the union’s bargaining team and indicated that employees might be ready to ‘vote up’ the employer’s agreement, instead of holding out. For example:

If the employees don’t vote for a protected industrial action, then you’d have to say that this was an action driven entirely by the bargaining representatives. That then means they’re not what the employees sought. And what we try then is you would have to think that maybe our agreement, if we put it back with ... a revised offer, in fact, may get voted up as well.[77]

However, in some negotiations, a failed ballot simply delayed protected industrial action. For example, one union representative described an unsuccessful ballot as ‘annoying’ because ‘everything works on the assumption that the ballot will get up’, but then added that it ‘wasn’t a huge problem’ as their bargaining strategy could be revised post-ballot, and that negotiations with the employer were ongoing.[78]

A second tactical consideration is that if the ballot succeeds, the smaller the level of voter turnout or support for industrial action, the less persuasive the ballot will be in escalating pressure on the employer. This came across strongly in both union and employer interviews, and is illustrated by the following comments:

[T]he speech I give is, ‘Guys, if this is going to be anywhere near close 50/50, don’t bother. The employer knows you’re not interested in industrial action, we know we’re not interested in industrial action so why would we bother [applying for a] PAB?’[79]

***

I think the results of the ballot [were around] 57%, and that equates to 25% of our total number of staff voted for industrial action. That tells us there’s not much heat out there on these issues ... If it was that bad, then you would have 80% of the staff ... voting to go on strike but we haven’t got that.[80]

A third consideration is that the ballot reveals union density in the workplace to the employer. The results of the ballot show how many employees were balloted, which equates to the number of employees who would both be covered by the proposed agreement and are members of that union. This plays an important role in determining whether unions pursue a PABO. As one union interviewee explained:

And the other biggest thing is that it exposes your density. So as part of the ballot process, you have to provide a list to the Australian Electoral Commission — the standard ballotter — and they [the employer] provide a list of employees. And they don’t get the list of members, which is really good, so there’s no identities given, but they get the numbers. So they know that they have 54 employees, and that at the time that the ballot was held, there was 31 who were able to vote. So you don’t take the decision lightly because density, and [the] employer not knowing density, can sometimes very much be in your favour. So if, for example, we had 30% density at the site, I doubt we would take the measure, because then the employer knows, well, you don’t really have much pull here.[81]

Several union representatives stated that they would not even consider a PABO application unless they had a certain level of density at that enterprise. For example, one union interviewee said: ‘I always say to the organisers,

“don’t give me a protected action ballot to make, or an order application to me if you haven’t got 70%”.’[82]

C Internal Union Democracy

Our analysis indicates that the PABO regime has had a normative effect on ‘intraorganizational bargaining’[83] — the internal negotiations which take place between the bargaining team, the union leadership and union members around the decision to apply for a PABO, the types of industrial action to take, and when to take it. These internal union processes generally operate in an essentially democratic manner.

Most union bargaining representatives framed the decision to apply for a PABO as a decision of the membership, rather than of the bargaining team or the leadership. As one interviewee explained: ‘[It] shouldn’t ever be a decision of a union official ... it’s a decision of the members who work in an enterprise.’[84] The process is typically initiated by the bargaining team and the leadership, who strategically decide when to apply for a PABO based on signals from the employer, and then seek the approval of the membership to proceed with the application. One union interviewee described this in terms of ‘advising’ the membership, adding that the members (generally) ‘take the [union’s] advice’.[85]

Unions used different methods to assess members’ support for making a PABO application. In smaller workplaces, polls were typically by way of a show of hands at a members’ meeting, or less formal methods of gaining ‘feedback’ from employees. For instance, one union representative stated that he would have one-on-one conversations with members at their worksite to ascertain their views.[86] In larger workplaces, or where members were spread across different geographical locations, some unions used online polling, or asked members to sign a written ‘petition’ expressing support for the PABO application and support for the specific forms of industrial action proposed. In some instances, unions conducted their own secret ballot to assess the level of support.[87]

While it tended to be the bargaining team who initiated discussions around the proposed industrial action, several union representatives recounted situations where groups of union members agitated for industrial action at an early stage in the negotiations. In these circumstances, the bargaining representatives dissuaded the members from pursuing industrial action, and instead ‘recommended’ that they continue to negotiate with

the employer:

They were wanting to go on strike or take action or whatever and I said but you’re jumping the gun here. You really need to — we’re still negotiating, they’re still willing to talk. So my recommendation to them was that we continue to talk.[88]

However, while all of the union interviewees described an essentially democratic process underpinning the decision to apply for a PABO, two employer interviewees asserted that union bargaining representatives had applied for a PABO and then approached their members for support, without having first consulted the membership on the decision to apply.[89] However, the union representative involved in one of those negotiations, who was also interviewed, described engaging in a consensus-based decision-making process prior to applying for a PABO, directly contradicting what the employer representative said about failure to consult the membership.[90]

D Administrative Burden Incurred by Unions

The PABO regime constitutes a significant administrative burden for unions seeking to engage in protected industrial action. However, it is not the PABO application itself that is particularly onerous. As one union interviewee noted, ‘it’s a pretty well-trodden path for us, you know, our affidavits ... the rest of the material that’s in the application that we need to put to the Commission is all essentially a template’.[91] Instead, the administrative and practical challenges for most unions, where an employer is not engaging in litigious delaying tactics, relate to ensuring that quorum is achieved, ensuring that the ballot is successful and strategising around an uncertain timeframe.

It is not enough for the bargaining team to consult members at the initial stage the PABO is applied for, and then make an application. Instead, the legislation effectively requires unions to dedicate resources to ensuring that the membership: (i) understand the nature of the ballot and the importance of turning out to vote; (ii) participate in the vote; (iii) vote in favour of industrial action; and (iv) ensure that their membership records are up-to-date.

In relation to the first point, unions must explain the legislative process and significance of the ballot itself, as distinct from the decision to take industrial action. Otherwise, members may fail to recognise the importance of participating in the vote or to draw a distinction between voting to approve the option of taking industrial action and voting actually to take industrial action. As one union representative explained:

[O]ur members predominately, unless they’re very, very highly educated in the bargaining framework (because they’ve done it many, many times), you know, for them if they want to take industrial action they feel like, well they’ve made that decision, they can do it. They don’t see the clear connection or the importance of the ballot as a process. The ballot is a legal technical thing that they have to do ... The meeting that authorises the union to make the protected action ballot application is really important because people are actively making a decision. The ballot is kind of seen as a fait accompli — it’s just a process we have to go through — and that’s what people are a bit apathetic about ... Even though [we] try and drill it into people, ‘Look, it’s really important that you vote, we have to do this’. I think people go, ‘Oh well, it’s just a technicality, if I don’t vote it doesn’t matter’.[92]

To ensure that members vote, unions generally engage in an education process to explain the legislative requirements. The larger and more diffuse the membership, the longer this process is likely to take. For instance, in one negotiation involving a large number of employees spread across different locations statewide, it took the union ‘a couple of months’ to ‘go and educate the members about what is a PABO and why we do this’ before the union could ‘then go back out and seek endorsement from the members to do

the PABO’.[93]

In this respect, arguably the ‘delay’ that the ballot requirements impose on unions comprises both the statutorily imposed time it takes to apply for a PABO and to organise and conduct a ballot, and the time it takes for unions to fulfil the de facto requirement of educating members and conducting

an internal poll or other process to ascertain member support. The larger

the number of employees to be balloted and the more geographically dispersed those employees are, the longer the lead time required to pursue industrial action.

Further, unions normally undertake a member engagement strategy during the period between the initial decision to apply for a PABO and the conduct of the vote. In particular, the requirement that half of all enrolled voters cast a vote means that unions must invest resources in ensuring that members remain committed and motivated to participate in the ballot.[94] As one bargaining representative explained, with reference to the outcome of a failed ballot:

We’ve sent out more communications and we’ve made sure an official is at the facility at the time of the ballot, which we didn’t do last time — we just had trust in our leaders at the workplace to make sure that people would vote ... that didn’t work last time, so this time we’ve made sure we have at least one official there during the voting hours to be in the lunchroom telling people, ‘Remember, we’ve got to vote now, [have] you voted?’ That kind of stuff. And that will mean the turnout will be a lot higher.[95]

The quorum requirement influences union messaging. Union representatives described frequently having to ‘reassure’ members that a further vote will be conducted later, should the negotiations fail to progress after the ballot has been conducted. The following quotes from two different bargaining representatives exemplify the kind of messaging that is typically used:

We’ve got an inbound and outgoing call centre and I actually put together a script for them and sent it and they called most of the members. And the basis of the script was: we’re still negotiating, they’ve really not come at anything. We’ve ... put this out to a ballot which was approved by members, and now to make the biggest possible impact and to give the company the biggest possible scare, we need to all vote ‘yes’ for everything. And another reminder that a vote could help to access protected action, not a vote to take it, and a reassurance that taking action would be a separate vote that everyone could contribute to at a later date should negotiations not progress and should we get the

actions through.[96]

***

[O]ur messaging usually is: ‘Really, really important that you vote and you vote “yes” on all questions. Whether or not we take action, well we’ll make that decision after the ballot. So, this ballot is about your right to take action.’ So, we want to be really, really clear with people that the two things are separate. So when we vote in this ballot, it doesn’t mean we’re going out on strike or taking any kind of industrial action, it just means that we give ourselves that right and then we’ll be democratic about any decisions after that. And so ... people hopefully understand that. So, the ballot’s about having the right to do that and we always say, ‘Look, it’s really, really important that you vote “yes” and it’s a really, really strong turnout and a really, really strong “yes” vote because that shows everyone that we’re united and acting together.’[97]

However, different messaging may be used depending on the circumstances, as one union interviewee explained:

Oh look, it probably depends on the workplace. I mean, sometimes obviously protected action ballots are just used as a tool to put a little bit more pressure on the company. So, it probably depends on the mood of the workers at the time. Sometimes, you know to be honest, we’ll say ‘vote on it now, we’ll decide on action later,’ sometimes it’s upfront, you know, ‘we intend to take action’ or the members endorse taking that sort of action and, you know, it’s made clear from the get-go. So, I think it varies based on the site.[98]

The legislative requirements in relation to voting method and ballot agent also influence union bargaining strategy and resource allocation. If a PABO applicant uses the AEC to conduct the ballot, the cost is borne by the state.[99] If another ballot agent is used, the ballot applicant bears the full cost.[100] This pushes ballot applicants towards the AEC,[101] a choice which has consequences in terms of the voting method used.

While the AEC is cost-effective, the AEC does not guarantee ballot applicants their preferred choice of ballot mode, and does not presently offer electronic voting,[102] an option which is available under the FW Act.[103] This creates a dilemma for unions because postal ballots are associated with a lower chance of reaching quorum than attendance and electronic ballots.[104] According to the AEC officials who participated in the study, AEC officers try to accommodate the bargaining representatives’ ballot mode preferences, but cannot always do so.[105] Characteristics impacting on the practicality of an attendance ballot include matters such as the number of employees, their location, whether they are participating in shift work, and the attitude

of the employer.

These constraints impose a burden on unions. The fact that the AEC does not offer electronic balloting means that in order to use electronic voting, unions must pay the full cost of the ballot. A similar problem arises if unions do not want to take the risk of failing to achieve quorum in a postal ballot, and this is all that the AEC can offer in the circumstances. Here, the union can either pay for an attendance ballot, or accept a postal ballot through the AEC and take steps to mitigate against the risk of failure to attain quorum. In each case, the union must expend resources to maximise the prospects of successfully navigating the ballot process.

Most ballot applicants elected to use the AEC in preference to paying someone else to conduct the ballot, and some devised methods to

address the perceived shortcomings of postal balloting, as the following interviewee disclosed:

What I have done is said to the lads, once you receive your papers, bring them into work, sit with the reps, all sit together, fill them out and then post them off. So we then can tick them off to see who’s received one, who hasn’t and where they are.[106]

Finally, due to the quorum requirement, coupled with the fact that the ballot result will reveal union density, it is incumbent on unions to ensure that the ballot agent has access to a complete list of all of the members who are eligible to participate in the ballot, and that the address and contact details for those members are up-to-date for the purposes of preparing the roll and notifying eligible employees of their right to participate.

It should be noted that unions are unable to check the accuracy of the AEC electoral roll. Instead, individual voters must ensure that the AEC has their correct details. Inaccuracies in the roll are material because abstentions effectively count as ‘no’ votes. To overcome this, unions must prompt members to update their details in advance of the roll being prepared.[107]

Given the complexity of the PABO regime, the regime favours ‘repeat players’ who are well-resourced and skilled in navigating the regulatory requirements. The delay created by the PABO process, and associated resource burden created hardship for some union interviewees:

It’s made so difficult and tiresome, this is the process. It is hard and

long — whereas years ago we could walk in, show of hands, yep, bang, go in and see the boss and say right, they’re ready for protected action. They’re going to take action tomorrow ... now this process has come in and it ties you

all up.[108]

However, not all union respondents found the ballot process disadvantageous. One bargaining representative stated: ‘[I]n the general sense we like the ballot. Like there’s a lot of unions who don’t like the ballot, we like the ballot, we think it’s a great campaign opportunity for us.’[109]

Ultimately, ballot applicants in certain industries or sectors may find it easier to go through the PABO process than their counterparts in other industries or sectors due to the relative likelihood of an employer taking steps to undermine subsequent industrial action:

[In this industry] for us to do a four-hour stoppage, it might literally cost them hundreds of thousands of dollars ... they’re subcontracted to [companies] up the chain, so there’s a lot of pressure on them from the contractors to avoid any stoppages of work because it literally costs so much money ... these are employers who have deep pockets to fund significant legal challenges to the PABO. When I was at [a different union] we might be talking about [redacted] workers wearing stickers that say ‘I’m low paid’. That was the kind of industrial action that we’d take. So the employers, they had less money, they had less of an interest in stopping the action potentially, it was more reputational damage rather than commercial damage.[110]

E Employer Responses

Employer responses to PABO applications feed into tactical considerations and the regulatory burden, and provide insight into how the legislation is operating from the perspective of the party whom it was designed to shield from ‘unnecessary’ industrial action.[111] The interviews also reveal how employers have adapted to the regime and how they have responded to maximise their own advantage.

Employer responses to a PABO application vary according to the likely benefits of the outcome, and external factors such as vulnerability to industrial action, public image and resources. Three specific employer strategies deployed in response to PABO applications were identified by the analysis: forestalling industrial action; undermining momentum in order to frustrate the union’s ability to use protected industrial action at a specific time; and creating fatigue so as to reduce employees’ appetite for industrial action and force concession.

1 Opposing the Ballot Application

Employers may contest the making of a PABO by the FWC on the ground that the applicant has failed to satisfy the statutory tests in s 443 of the FW Act outlined above.[112] Employers may also challenge aspects of the proposed order, for example, by arguing that the proposed industrial action was not sufficiently clearly stated,[113] or seeking an extension of the notice period for taking protected industrial action up to the statutory maximum of seven working days.[114]

While most PABO applications are unopposed, some employers object on technical grounds to slow the union’s momentum and delay access to industrial action. As an Australian Council of Trade Unions (‘ACTU’) representative explained: ‘[E]very time you go to the Commission, there’s an opportunity for the employers, with their resources, to bring the lawyers along and pick arcane, technical fights with you.’[115] Another union representative described this as a ‘frustrating’ side effect of the ballot approval process:

I find the approval process, if the employer wants to challenge, really quite irritating. Often I find that often it’s misunderstood what grounds you can challenge on and so we’ll go to hearings and end up listening to submissions about health and safety impacts and this kind of thing. Often with their legal reps knowing full well they’re not grounds, but it’s just a way to stall the process and it’s very frustrating. I feel like the Commission should ask the grounds or for submissions in advance, or ask the grounds upon which they are challenging in advance.[116]

One union representative commented that the employer’s ‘delaying tactic’ was not only designed to forestall industrial action, but was intended to ‘provoke’ the union members to ‘do something silly’, such as walking off the job, so that the employer could then pursue the union for taking unprotected

industrial action.[117]

Several employer representatives described taking steps to stall industrial action. For example:

Last time we did think there was a heightened threat of industrial action because the fever of where negotiations were at and the militancy was very different and so it was basically, you know, if we can delay this occurring and damaging our business even by 48 hours, and 48 hours, and 48 hours, because they have to reapply, then it gets, you know, goes back through the registry and all that sort of stuff, then that’s protecting our business. So that was one of the key things last time, was well, let’s make sure that they dot their i’s and cross their t’s and hold them to account because that buys us more time as

an employer.[118]

Another employer representative explicitly acknowledged using technicalities to obtain a bargaining advantage:

Well, the request to have an extension [of the notice period for industrial action] or to have the period of time extended the ballot’s opened for is simply a stall tactic, it buys us more time to be able to continue to have negotiations ... So initially, you know, requests, and I can’t think, one of them might have been 20 days, but, you know, we got 28 days and in fact [the union] have been in discussions with the manager to [extend to] 35 days, so that simply means that obviously the period before they could take protected action is delayed, so that’s in our interests.[119]

Employer opposition to a PABO application may stem from a misunderstanding of the legislative requirements, rather than a calculated strategy to stall. One union representative recalled a PABO application where the employer opposed the order because ‘they wanted to keep talking’.[120] The union representative did not perceive this to be a ‘delaying tactic’, and said: ‘I just think they didn’t know ... they’ve got a history of doing things like that ... they don’t fully understand processes all the time’, and added: ‘they always oppose it’.[121]

Irrespective of an employer’s motive for opposing a PABO application, the associated delay for the ballot applicant can be greater than the 48 hours allowed in the legislation,[122] because a Commissioner may reserve their decision and/or the matter may go to hearing. For example, an employer interviewee recalled opposing a ballot on the basis that the union had not correctly described the entities and there were ‘irregularities’ with the application; the Commissioner reserved their decision for two weeks and then made the PABO.[123]

Deciding whether to oppose a PABO application is ultimately a cost-benefit exercise for employers. Most employer representatives indicated that they did not routinely oppose PABO applications because they had no valid basis on which to do so, or because the ballot would likely be granted even if it were opposed. For example, one employer representative stated that after weighing up the ‘potential outcomes’, they resolved not to oppose the application: ‘[O]ur experience was that ... they generally get through ... [even if] you’re picking at a certain piece of it that you might have some success on.’[124] Keeping on the ‘good side’ of the relevant FWC Member was a guiding consideration for another employer:

We also wanted to keep in the Commissioner’s good graces because if you run an argument that you know is likely to fail you’re going to consume a lot of a very busy Commissioner’s time; so what we said was that we’ll save that time with the Commission and if the vote gets up, instead of asking for three days’ notice, we’ll ask for seven [days’ notice] of industrial action. So, that was our company strategy.[125]

Our statistical analysis of the PABO applications confirms that most are not contested. Applications were contested in 279 (21.4%) cases, not contested in 312 (24%) cases and assumed not contested in 668 (51.3%) cases.[126] For a further 43 applications (3.3%), it could not be determined whether or not the ballot was contested, as the order or decision relating to the application could not be located or was not made (eg due to withdrawal by the applicant).

Further, while only 21.4% of PABO applications in the sample were contested, analysis of contested applications reveals that frequently such contests were of a technical nature. Two hundred and twenty-eight of the 279 employer contests involved an assertion by the employer that the application failed adequately to specify the group to be balloted. Only 14 were contested on the substantive ground that the FWC could not be satisfied that the applicant was genuinely trying to reach agreement.

Another issue that employers frequently raise in response to PABO applications is a request for an extension of the notice period that a bargaining representative must give before taking protected industrial action. In 27.5% of PABO applications in the sample (n = 358), employers indicated that they would not challenge the PABO application if the notice period was changed by consent, and such changes were agreed by the applicant in all but eight instances. In total, in the sample only 14 cases involved active contest on the part of the employer with respect to extension of the notice period.

2 Undermining Support for Industrial Action

Some employers utilise the time period between the PABO application and the vote taking place to discourage support for industrial action. For example, by flooding union members with communications designed to scare the employees off taking industrial action:

Our members are getting flooded with toolbox meetings and letters from the boss and they’re here all the time, so they really ramp up the propaganda of our membership when this sort of stuff happens. And you know, when you’ve got the boss in your ear 24/7, sometimes you start to believe them.[127]

Several employers alluded to doing this. For instance, one employer representative described using the following messaging in order to diminish employee support and undermine the likely success of the ballot:

I also need to flag if you go down this path of protected action it’s absolutely fine— that’s your prerogative, no problem whatsoever — however, we will need to know in advance. We’ll have to have you taken off the after-hours roster because you can’t go and support our customers seven days a week, 24 hours a day if you’re not prepared to work overtime, and there’ll also be an impact on [client X]. Now, [client X] for us is an important service customer, it’s not our biggest service customer, but all the work we do at [client X is] very lucrative for them. So as soon as they started hearing, oh my goodness, that means I’m not going to get my 200-odd dollars a week if I’m on the roster for after-hours, I won’t be able to work at [client X], that’s a problem.[128]

The legislation in effect creates a ‘moving’ timeline which makes it more complex for unions to plan when to use industrial action. The PABO regime provides opportunities for employers to throw off the union’s timeline, undermining momentum and member support for the proposed course of action. This contributes to the administrative burden that the legislation imposes on unions because ensuring participation and support for industrial action becomes more difficult the longer the process takes. A decision to pursue a PABO may be entirely democratic, but when faced with interruptions and tactics to induce fatigue, or simply the length and complexity of a large postal ballot, support may wane, and members may disengage. An ACTU representative explained this as follows:

Number one, the delay is fatal because you can expect the employer to actively campaign against what the union is trying to organise. And the longer the process takes, the more opportunity there is for the ‘no’ campaign, the counterinsurgency, to be run and won. So unions want it to be quick, is one consideration.[129]

3 Deterring Ballot Participation

As observed earlier, the method of voting impacts on ballot outcome, because attendance ballots are more likely to reach quorum than postal ballots.[130] AEC staff confirmed that the AEC does not guarantee ballot applicants their preferred choice of ballot mode, and acknowledged that the AEC must walk a ‘fine line’ between satisfying the applicant’s preferences and minimising disruption to the employer.[131] In practice, this means that there is potential

for employers to object to the conduct of an attendance ballot at the work premises.[132]

According to union representatives, some employers are cognisant of this and try to make it more difficult for the union to achieve a successful ballot outcome by opposing an attendance vote:

Well a lot of companies will go postal votes hoping that the lads haven’t updated their addresses and stuff because a lot of people move around from house to house and stuff. Attendance votes I like because they’re the blokes actually on the job.[133]

One union interviewee described the reverse situation, where on two occasions employers had urged that the ballot be an attendance ballot, but then sent the employees to be balloted to work at another location so that they would be absent when the ballot took place:

[S]ome companies have turned around and said the vote’s got to be here: ‘sorry I’ve got a job away; I just sent 10 or 15 blokes interstate or into a remote area’. That happened a couple of times and we got caught out. So ... we’d say that all these people that are on the list must attend this date. The boss doesn’t have the right to send them away somewhere else. If he does he has to notify the AEC say probably two hours before he makes his decision, at least offer the guys to either ring in or put in a postal vote.[134]

Although it was clear from many employer interviews that employer representatives are aware that ballot mode impacts on quorum, none of the employer representatives interviewed in this study explicitly referred to taking steps to deter employees from participating in ballots, whether by opposing an attendance ballot, or by directing eligible voters to work at another location where they could not be present for an attendance ballot.

VI Conclusion

This article has examined the role that pre-strike ballot regulation plays in shaping bargaining behaviour and decision-making. The findings reveal that instead of functioning simply as a ‘neutral access gate’ that ensures a democratic process for determining whether a group of employees wish to take industrial action, the pre-strike ballot regime structures bargaining dynamics and influences bargaining behaviour.

First, the analysis confirms Orr and Murugesen’s 2007 hypothesis: the ballot regime in effect delays protected industrial action. In its practical effect, the requirement to engage in the PABO process serves to slow employees’ access to protected industrial action because of the need to apply for a PABO, obtain an order, schedule and conduct the ballot, await results and then give the employer notice of when the action will take place. But, more importantly, the timeframe structures the negotiations by adding considerable uncertainty for unions, who are not always able accurately to predict when they will be able lawfully to take action.

Second, the analysis confirms that applying for a PABO, and the conduct of the subsequent ballot, can be an effective means for unions to increase leverage by escalating the ‘threat’ of industrial action, without necessarily having to resort to industrial action to gain that leverage.

A successful ballot outcome signals industrial strength and accelerates the ‘countdown’ towards industrial action. At the same time, however, the findings show that an unsuccessful ballot outcome, whether by a failure to achieve quorum or due to insufficient voter support for taking industrial action, signals weakness, significantly de-escalates the threat of industrial action, and shifts leverage in the negotiations to the employer. In order to maximise the strategic ‘benefit’ of a PABO, unions must invest time and resources in ensuring that a sufficient number of employees vote in the ballot, and that those employees vote in favour of taking industrial action. In some instances, this is likely to offset the signalling effect that Orr and Murugesan saw as the ‘positive’ of the ballot process for unions.[135] Further, some unions will avoid balloting altogether if the outcome may reveal low levels of membership density to an employer, or where they are concerned that they may not achieve quorum.

Third, the analysis reveals that some unions are better able to navigate the ballot requirements than others. For example, achieving the quorum requirement may be more difficult for a ballot applicant that has a large number of employees to be balloted and those employees are spread across multiple worksites, requiring additional communications. For some unions, there is a need to devote significant effort to demystifying and explaining the ballot process and ensuring that members remain engaged as the process unwinds. This is more complicated and requires a concerted effort to obtain membership approval than if the union could manage its own ballot

process and then take industrial action if a majority agreed, in accordance with the rules.

Fourth, the research confirms that some employers are using the pre-strike ballot regime to their strategic advantage. They are, for instance, exercising opportunities to ‘intervene and object’ to the initial ballot application in order to slow the progress of the ballot even though there are no substantive grounds for the objection. The analysis also reveals that employers have other means of delaying the ballot and adding uncertainty to the timeframe, for example, by objecting to the ballot questions, seeking longer notice periods, or for the ballot to be conducted over a longer period. In addition, some employers take steps to deter eligible employees from voting, to reduce the likelihood that the ballot will reach quorum.

This suggests that some employers use the statutory provisions to:

(i) forestall industrial action; (ii) undermine momentum in order to frustrate the union’s ability to use protected industrial action at a specific time; and

(iii) create ‘fatigue’ so as to reduce employees’ appetite for industrial action.

These findings about the practical operation of the ballot requirements have significant policy implications.

As identified above, the introduction of the mandatory pre-strike ballot requirement was justified by reference to democratic imperatives, specifically the need to implement a ‘fair, effective and simple process for determining if a group of employees in an enterprise want to take industrial action’.[136] In practice, mandatory pre-strike ballots have had a positive effect on intra-organisational communication and internal union decision-making around industrial action. The analysis revealed that, on the whole, the unions in the study implemented democratic processes for member consultation and engagement over every step of the PABO process, including the decision to apply for a PABO, the ballot itself and the subsequent decision to take industrial action. These decisions were almost universally referred to by union interviewees as ‘member’ decisions subject to internal union processes with high levels of member engagement.

Although there was strong evidence of the normative effect of the pre-strike ballot regime on intra-organisational communication and decision-making, this outcome came at the cost of the diversion of significant union resources to explaining the statutory processes rather than discussion of the decision to take industrial action. Almost all union interviewees described the continued need to engage in a process of education with respect to the PABO application, the ballot vote and the signalling effect of the vote, and any subsequent decision to take industrial action. The layers of complexity built into the regime hinder its ability to provide a mechanism for employees to authorise industrial action in a fair and simple manner.

The ballot requirements, in effect, impose a significant barrier to unions and their members exercising their right lawfully to take industrial action. This is consistent with the hypothesis that the ‘real’ reason for the imposition of mandatory pre-strike ballot requirements is to constrain access to protected industrial action by making it harder for unions to strike.[137] While the PABO regime has produced unexpected benefits for unions, in particular the ability to leverage a successful PABO application or ballot outcome to obtain concessions in bargaining, these benefits come at the cost of revealing levels of union density or the potentially devastating impact on union bargaining power that may result from a failed ballot result.

The interviews with union representatives revealed that the PABO regime has a ‘shadow’ effect on bargaining and access to the right to strike, because some unions choose not to go to ballot due to lost opacity through exposure of density or due to the danger of losing. Even where industrial action is approved, low voter turnout or a narrow majority may signal weakness to an employer. The constraints of this study meant that it was not possible to quantify this effect due to the fact that only union representatives who had actually made a PABO application were interviewed. To assess accurately the shadow effect would be methodologically difficult, and would require interviews with union officials who had decided not to apply for a PABO. It seems reasonable to suppose, however, that such an impact would be significant. This is borne out by recent research by Stanford on the decline of industrial disputation in Australia, which suggests a direct link between increased regulation of the right to strike, decreased industrial disputation and wage stagnation.[138] The potential deterrent effect of the PABO regime on unions seeking to test member support for industrial action may form part of this picture.

Research participants expressed a variety of views on whether the current regulatory approach is working effectively. In light of the findings described above, it is suggested that a better balance could be struck between the public interest in ensuring democratic decision-making and the legitimate interests of organised labour. This could occur by allowing unions to conduct their own ballot process within an appropriate regulatory framework without having first to apply for a PABO, face a potential delay whilst the tribunal deliberates, and then wait for an independent ballot agent such as the AEC to conduct a ballot.

It is suggested that adopting this approach would be more likely in practice to achieve the stated objectives of the FW Act.[139] Union resources and communications would no longer need to be directed to explaining to members elaborate statutory processes and emphasising the signalling role of PABOs and ballots. Member support could be gauged without some of the attendant risks of compromising unions’ bargaining positions. This, in turn, could be expected to make unions more likely to use internal democratic processes in order to gauge levels of member support for taking industrial action. This could both facilitate democratic decision-making and remove the administrative burden imposed on the FWC, employers and unions by the provisions currently enshrined in pt 3-3 div 8 of the FW Act.


* LLB (Hons) (QUT), LLM (Syd), PhD (Syd); Lecturer, Faculty of Law, University of New South Wales.

** LLB (Belfast), LLB (Hons), (Melb) PhD (Camb); Honorary Professor, Graduate School of Business and Law, RMIT University.

† BA, BPPM (Hons) (Melb), BA (Law) (Hons) (Camb), PhD (Law) (UCL); Research Fellow, Director of Ulster Legal Innovation Centre, School of Law, Ulster University.

‡ BBusSci (Hons) (UCT), LLB (Hons), PhD (Melb); Professor, Faculty of Law, Queensland University of Technology.

‡‡ BA LLB (Hons), PhD (Tas); Professor of Labour Law, Sydney Law School, The University of Sydney.

The authors would like to thank the anonymous referees for helpful feedback on an earlier version of this article. This research is supported by a grant from the Australian Research Council: ARC DP140100902, ‘Protected Action Ballots and Protected Industrial Action under the Fair Work Act: The Impact of Ballot Procedures on Enterprise Bargaining Processes’.

[1] Fair Work Act 2009 (Cth) pt 3-3 div 2 (‘FW Act’).

[2] Ibid pt 3-3 div 8.

[3] As to these potential exposures, see Andrew Stewart et al, Creighton and Stewart’s Labour Law (Federation Press, 6th ed, 2016) ch 26.

[4] The changes were introduced by the Workplace Relations Amendment (Work Choices) Act 2005 (Cth) sch 1 (‘Work Choices’), amending Workplace Relations Act 1996 (Cth) pt 9 div 4 (‘Workplace Relations Act’).

[5] Although there was a rarely used procedure where union members could apply to the then Australian Industrial Relations Commission for an order for the conduct of a secret ballot: Industrial Relations Act 1988 (Cth) s 136. See also Breen Creighton, Catrina Denvir and Shae McCrystal, ‘Strike Ballots and the Law in Australia’ (2016) 29(2) Australian Journal of Labour Law 154, 156.

[6] Work Choices (n 4) pt 9 div 4. However, the term ‘protected action ballot order’ was first used in the FW Act (n 1) s 437.

[7] FW Act (n 1) s 457.

[8] Peter Reith, Pre-Industrial Action Secret Ballots (Ministerial Discussion Paper, August

1998) 1.

[9] See generally Andreas Pekarek et al, ‘Old Game, New Rules? The Dynamics of Enterprise Bargaining under the Fair Work Act(2017) 59(1) Journal of Industrial Relations 44.

[10] See, eg, Keith Townsend, Adrian Wilkinson and John Burgess, ‘Is Enterprise Bargaining Still a Better Way of Working?’ (2013) 55(1) Journal of Industrial Relations 100, 113; Breen Creighton, ‘Impacts of Enterprise Bargaining on Unions and Employers: Discussants Comments’ (2012) 22(3) Labour and Industry 275, 283; David Peetz, ‘The Impacts and Non-Impacts on Unions of Enterprise Bargaining’ (2012) 22(3) Labour and Industry 237, 246–7; Chris Briggs, ‘Australian Exceptionalism: The Role of Trade Unions in the Emergence of Enterprise Bargaining’ (2001) 43(1) Journal of Industrial Relations 27, 37–9; Peter Gahan and Andreas Pekarek, ‘The Rise and Rise of Enterprise Bargaining in Australia, 1991–2011’ (2012) 22(3) Labour and Industry 195; Breen Creighton, ‘Getting to the Bargaining Table: Coercive, Facilitated and Pre-Commitment Bargaining’ in Shae McCrystal, Breen Creighton and Anthony Forsyth (eds), Collective Bargaining under the Fair Work Act (Federation Press, 2018) 25; Rosalind Read, ‘The Role of Trade Unions and Individual Bargaining Representatives: Who Pays for the Work of Bargaining’ in Shae McCrystal, Breen Creighton and Anthony Forsyth (eds), Collective Bargaining under the Fair Work Act (Federation Press, 2018) 69; Umeya Chaudhuri and Troy Sarina, ‘Employer-Controlled Agreement-Making: Thwarting Collective Bargaining under the Fair Work Act’ in Shae McCrystal, Breen Creighton and Anthony Forsyth (eds), Collective Bargaining under The Fair Work Act (Federation Press, 2018) 138.

[11] See, eg, Richard Hyman, Strikes (Fontana, 1972) 34–6; Michael P Jackson, Strikes: Industrial Conflict in Britain, USA and Australia (Wheatsheaf Books, 1987) ch 6; Josh Healy, ‘Peace at Last? Recent Trends in Australia’s Industrial Action’ (2002) 28(2) Australian Bulletin of Labour 80.

[12] See, eg, Stephen J Frenkel, ‘Patterns of Industrial Action: Analysis and Conclusions’ in Stephen J Frenkel (ed), Industrial Action: Patterns of Labour Conflict (George Allen & Unwin, 1980) 132; David Peetz, ‘Industrial Conflict with Awards, Choices and Fairness’ in Breen Creighton and Anthony Forsyth (eds), Rediscovering Collective Bargaining: Australia’s Fair Work Act in International Perspective (Routledge, 2012) 159.

[13] For example, Australia, Canada, New Zealand and the UK all mandate pre-strike ballots: Breen Creighton and Shae McCrystal, ‘Strike Ballots and the Law in Comparative Perspective’ (2016) 29(2) Australian Journal of Labour Law 121.

[14] See, eg, Tonia Novitz, ‘UK Regulation of Strike Ballots and Notices — Moving Beyond “Democracy”’ (2016) 29(2) Australian Journal of Labour Law 226.

[15] See, eg, Alan Bogg, ‘Beyond Neo-Liberalism: The Trade Union Act 2016 and the Authoritarian State’ (2016) 45(3) Industrial Law Journal 299; Michael Ford and Tonia Novitz, ‘Legislating for Control: The Trade Union Act 2016’ (2016) 45(3) Industrial Law Journal 277; Ruth Dukes and Nicola Kountouris, ‘Pre-Strike Ballots, Picketing and Protest: Banning Industrial Action by the Back Door?’ (2016) 45(3) Industrial Law Journal 337; Keith Ewing, John Hendy and Carolyn Jones (eds), A Manifesto for Labour Law: Towards a Comprehensive Revision of Workers’ Rights (Institute of Employment Rights, 2016); Shae McCrystal and Tonia Novitz, ‘“Democratic” Pre-Conditions for Strike Action: A Comparative Study of Australian and UK Labour Legislation’ (2012) 28(2) International Journal of Comparative Labour Law and Industrial Relations 115.

[16] Trade Union and Labour Relations (Consolidation) Act 1992 (UK) ss 226–34.

[17] FW Act (n 1) ss 444, 449.

[18] Jane Elgar and Bob Simpson, Industrial Action Ballots and the Law (Institute of Employment Rights, 1996) 16. See also Roderick Martin et al, ‘The Decollectivisation of Trade Unions? Ballots and Collective Bargaining in the 1980s’ (1991) 22(3) Industrial Relations Journal 197, 207.

[19] Elgar and Simpson, Industrial Action Ballots and the Law (n 18) 20; Roger Undy et al, Managing the Unions: The Impact of Legislation on Trade Unions’ Behaviour (Clarendon Press, 1996) 215–16.

[20] Undy et al (n 19) 216.

[21] Martin et al (n 18) 202.

[22] Undy et al (n 19) 230; Martin et al (n 18) 202; Jane Elgar and Bob Simpson, ‘The Impact of the Law on Industrial Disputes Revisited: A Perspective on Developments over the Last Two Decades’ (2017) 46(1) Industrial Law Journal 6, 10 (‘Impact of the Law on Industrial Disputes Revisited’).

[23] Elgar and Simpson, Industrial Action Ballots and the Law (n 18) 20; Jane Elgar and Bob Simpson, ‘The Impact of the Law on Industrial Disputes in the 1980s’ in David Metcalf and Simon Milner (eds), New Perspectives on Industrial Disputes (Routledge, 1993) 70, 105.

[24] Undy et al (n 19) 229.

[25] Elgar and Simpson, ‘Impact of the Law on Industrial Relations Revisited’ (n 22) 8, 21.

[26] Ibid 22.

[27] See, eg, Breen Creighton et al, ‘Protected Industrial Action Ballots: An Empirical View’ (2018) 60(1) Journal of Industrial Relations 53, which empirically describes the pre-strike ballot process under the FW Act (n 1).

[28] Ron McCallum, Michael Moore and John Edwards, Towards More Productive and Equitable Workplaces: An Evaluation of the Fair Work Legislation (Report, 2012) 180.

[29] Productivity Commission, Workplace Relations Framework (Report No 76, 30 November 2015) 869–79.

[30] Ibid 870–1.

[31] Graeme Orr and Suppiah Murugesan, ‘Mandatory Secret Ballots before Employee Industrial Action’ (2007) 20(3) Australian Journal of Labour Law 272, 273. See also Shae McCrystal, ‘Smothering the Right to Strike: Work Choices and Industrial Action’ (2006) 19(2) Australian Journal of Labour Law 198, 204.

[32] Orr and Murugesan (n 31) 272.

[33] Ibid 294.

[34] Reith (n 8) 1.

[35] Work Choices (n 4) pt 9 div 4.

[36] See, eg, Senate Employment, Workplace Relations, Small Business and Education Legislation Committee, Parliament of Australia, Consideration of the Provisions of the Workplace Relations Amendment (More Jobs, Better Pay) Bill 1999 (Final Report, November 1999)

254–61.

[37] The history is explored further in Creighton, Denvir and McCrystal, ‘Strike Ballots and the Law in Australia’ (n 5) 155–9; Shae McCrystal, ‘A New Consensus: The Coalition, the ALP and the Regulation of Industrial Action’ in Anthony Forsyth and Andrew Stewart (eds), Fair Work: The New Workplace Laws and the Work Choices Legacy (Federation Press, 2009)

141, 141–3.

[38] As to the definition of ‘industrial action’ in s 19(1) and the ambiguities created by the definition, see Breen Creighton, Catrina Denvir and Shae McCrystal, ‘Defining Industrial Action’ (2017) 45(3) Federal Law Review 384.

[39] As to agreement making under the FW Act (n 1), see Stewart et al (n 3) ch 14.

[40] FW Act (n 1) s 415. As to the potential liabilities and the concept of ‘protected industrial action’, see generally Stewart et al (n 3) chs 26, 27.

[41] FW Act (n 1) s 437.

[42] Ibid s 438.

[43] Ibid s 443(1)(a).

[44] Ibid ss 173(2), 437(2A).

[45] Ibid s 443(1)(b); see also at s 413(3). As to the meaning of ‘genuinely try to reach an agreement’, see Stewart et al (n 3) 982–5.

[46] Ibid ss 444, 449. The AEC is an independent statutory body which has responsibility, inter alia, for the conduct of federal elections, by-elections and referenda: Commonwealth Electoral Act 1918 (Cth) s 7.

[47] FW Act (n 1) ss 450, 451.

[48] Ibid s 499(2).

[49] Ibid s 451(2).

[50] Ibid s 450(2).

[51] Ibid s 459(1).

[52] See especially ibid s 414.

[53] Ibid s 459(1)(d). For a discussion of the ‘30-day rule’, see Catrina Denvir and Shae McCrystal, ‘Researching Labour Law “in Practice”: Challenges in Assessing the Impact of Protected Industrial Action Ballot Procedures on Enterprise Bargaining Processes’ in John Howe, Anna Chapman and Ingrid Landau (eds), The Evolving Project of Labour Law: Foundations, Development and Future Research Directions (Federation Press, 2017) 161, 173–5.

[54] Reith (n 8) 1.

[55] PABO applicants were union officials who were either: (a) directly involved in the negotiations with an employer for an enterprise agreement; or (b) union legal officers who took responsibility for submission of the PABO application and the legal process governing the granting of a ballot order (including appearing at a FWC hearing where necessary). PABO application respondents were employer representatives (typically human resources managers or small business owners) involved in negotiations with union officials for a single-enterprise agreement.

[56] Anslem Strauss and Juliet Corbin, Basics of Qualitative Research: Grounded Theory Procedures and Techniques (Sage Publications, 1990) 23.

[57] Juliet Corbin and Anslem Strauss, Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory (Sage Publications, 3rd ed, 2008) 52.

[58] For an explanation of a grounded theory approach to data analysis, see generally Kathy Charmaz, Constructing Grounded Theory: A Practical Guide through Qualitative Analysis (Sage Publications, 2006).

[59] Some negotiations gave rise to more than one ballot application, for example, where more than one union was involved, or a union balloted more than once. However, the primary cause of the large disparity was separate employers bargaining together under a single interest authorisation. A single interest authorisation allows certain related employers to negotiate together for the purposes of creating a single-enterprise agreement. However, the employees of these enterprises may pursue industrial action separately. In the study period, 740 PABO applications were received in respect of five enterprise agreement negotiations involving single interest employers.

[60] This figure was reached based on an inference of withdrawal where no PABO order could be found by the investigators and/or a withdrawal order was located. Whilst sustained efforts were made to locate all orders, it is possible that in some of these cases an order was made which was not made public or locatable through public searches before the PABO application was withdrawn.

[61] 51 PABO applications subsequently were withdrawn before a ballot was conducted. A further three applications were withdrawn after the ballot had been conducted. Because those three ballots were conducted, they are included in the following analysis.

[62] Paul Weiler, Reconcilable Differences: New Directions in Canadian Labour Law (Carswell, 1980) 49.

[63] Interview with union (Interview 49).

[64] Interview with union (Interview 73).

[65] Interview with union (Interview 22); Interview with union (Interview 28); Interview

with union (Interview 49); Interview with union (Interview 56); Interview with union (Interview 67); Interview with union (Interview 70).

[66] Interview with union (Interview 22).

[67] Interview with employer (Interview 21).

[68] Interview with employer (Interview 19). Note that the employer added, ‘we really don’t mind if we lose this contract, we’re not making any money out of it’, and that the employees have ‘more to lose’ because without the contract, the employer would not be able to pay them.

[69] Interview with union (Interview 56).

[70] Interview with union (Interview 8).

[71] Interview with union (Interivew 66).

[72] Interview with union (Interview 48).

[73] Interview with union (Interview 43).

[74] The broader factors that dictate the effectiveness of industrial action in the course of bargaining negotiations lie beyond the scope of this paper, but include: the degree of cooperativeness in the bargaining relationship, the degree of experience of participants, economic and reputational costs, cost of delays and incentives to make concessions.

See Peter Cramton and Joseph Tracy, ‘Unions, Bargaining and Strikes’ in John T Addison and Claus Schnabel (eds), International Handbook of Trade Unions (Edward Elgar, 2003) 86; Edward Montgomery and Mary Ellen Benedict, ‘The Impact of Bargainer Experience on Teacher Strikes’ (1989) 42(3) Industrial and Labor Relations Review 380; Joseph S Tracey, ‘An Empirical Test of an Asymmetric Information Model of Strikes’ (1987) 5(2) Journal of Labour Economics 149.

[75] Interview with union (Interview 28).

[76] Interview with union (Interview 63).

[77] Interview with employer (Interview 34). Similar comments were expressed in interview with union (Interview 16).

[78] Interview with union (Interview 47).

[79] Interview with union (Interview 73).

[80] Interview with employer (Interview 31).

[81] Interview with union (Interview 57).

[82] Interview with union (Interview 12).

[83] Richard E Walton and Robert B McKersie, A Behavioral Theory of Labor Negotiations (McGraw-Hill, 1965) 5–6.

[84] Interview with union (Interview 73).

[85] Interview with union (Interview 62).

[86] Interview with union (Interview 18).

[87] Interview with union (Interview 8).

[88] Interview with union (Interview 64).

[89] Interview with employer (Interview 52); Interview with employer (Interview 33).

[90] Interview with union (Interview 51).

[91] Interview with union (Interview 55).

[92] Interview with union (Interview 47).

[93] Interview with union (Interview 70).

[94] Breen Creighton et al, ‘Pre-Strike Ballots and Collective Bargaining: The Impact of Quorum and Ballot Mode Requirements on Access to Lawful Industrial Action’ (2018) Industrial Law Journal (advance) 28–9 (‘Pre-Strike Ballots and Collective Bargaining’).

[95] Interview with union (Interview 47).

[96] Interview with union (Interview 56).

[97] Interview with union (Interview 47).

[98] Interview with union (Interview 24).

[99] FW Act (n 1) s 464.

[100] Ibid s 465.

[101] Of 1204 pre-strike ballots in the study, the AEC conducted 99.2% (n = 1193) whilst 0.8%

(n = 10) involved another protected action ballot agent.

[102] Only the choice of either a postal or attendance vote is offered by the AEC: see Australian Electoral Commission, ‘Protected Action Ballots’ (Web Page, 23 September 2016) <https:// www.aec.gov.au/Elections/pab/>, archived at <https://perma.cc/BH6P-VREA>.

[103] FW Act (n 1) s 451.

[104] In our sample, attendance ballots were associated with a 91.3% chance of meeting quorum, compared to an 80.2% chance of meeting quorum for postal ballots: see Creighton et al,

‘Pre-Strike Ballots and Collective Bargaining’(n 94) 19.

[105] Interview with AEC.

[106] Interview with union (Interview 28).

[107] Interview with union (Interview 58); Interview with union (Interview 47).

[108] Interview with union (Interview 28).

[109] Interview with union (Interview 55).

[110] Interview with union (Interview 62).

[111] Reith (n 8) 1.

[112] See, eg, Australian Postal Corporation v Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union of Australia [2009] FWAFB 599; (2009) 189 IR 262,

267–8 [16]–[17] (Acton SDP, Hamilton DP and Blair C).

[113] See, eg, John Holland Pty Ltd v Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union [2010] FWAFB 526; (2010) 194 IR 239, 244–6 [11]–[23] (Giudice J, Watson SDP and Blair C).

[114] See FW Act (n 1) ss 414(2), 443(5). See also Stewart et al (n 3) 976–7 [27.12].

[115] Interview with Australian Council of Trade Unions (Interview 2).

[116] Interview with union (Interview 41).

[117] Interview with union (Interview 69).

[118] Interview with employer (Interview 52).

[119] Interview with employer (Interview 33).

[120] Interview with union (Interview 42).

[121] Ibid.

[122] FW Act (n 1) s 441.

[123] Interview with employer (Interview 52).

[124] Interview with employer (Interview 25).

[125] Interview with employer (Interview 36).

[126] Assumed non-contest is the conclusion drawn by the research team when an order for a ballot was made without a decision also being made available. The FWC does not always publish a decision when it makes an order.

[127] Interview with union (Interview 51).

[128] Interview with employer (Interview 29).

[129] Interview with Australian Council of Trade Unions (Interview 2).

[130] Creighton et al, ‘Pre-Strike Ballots and Collective Bargaining’ (n 94).

[131] Interview with AEC.

[132] Note that to attain their preferred choice of ballot mode, it is open to a ballot applicant to elect to use an independent ballot agent other than the AEC. However, the cost and inconvenience of doing so militates against this in many instances: Creighton et al, ‘Pre-Strike Ballots and Collective Bargaining’ (n 94) 25.

[133] Interview with union (Interview 28).

[134] Ibid.

[135] Ibid 294.

[136] Reith (n 8) 1.

[137] See McCrystal and Novitz (n 15) 144–5; Creighton et al, ‘Strike Ballots and the Law in Australia’ (n 5) 167–70.

[138] Jim Stanford, The Australia Institute, Historical Data on the Decline in Australian Industrial Disputes (Briefing Note, 30 January 2018).

[139] FW Act (n 1) s 3.