INDUSTRIAL CONFLICT WITH AWARDS, CHOICES AND FAIRNESS by David Peetz, Professor, Department of Employment Relations& Human Resources, Griffith University
“To properly understand contemporary patterns of industrial conflict it is necessary to place them in their historical and national context. Attachment C to this submission, published as a chapter in a 2012 book, does this by analysing industrial conflict at the macro level under the Fair Work Act and its predecessor in the context of long term trends.
It considers patterns in the incidence of disputes, their causes, duration and method of settlement, distinguishes between contingent and unconditional strikes and considers the implications for how changes to bargaining regimes have affected bargaining behaviour.
It shows how, in recent years disputes have become less common, more concentrated in the times during which new collective agreements are negotiated, and longer. The role for negotiation in ending strikes has not increased.
The easing of restrictions on union actions under the FW Act has so far had little impact on strike activity. Short, unconditional strikes were never just about grievances and protests; they were always an important tactic in bargaining campaigns. They have now increased in duration, and importance, because of procedural requirements around disputation.
Despite nearly two decades of enterprise bargaining, there are still a significant number of short disputes and disputes on non-enterprise bargaining matters, though the Fair Work Act has been accompanied by a fall in the level and significance of unprotected action. Unprotected strikes remain shorter than protected strikes, as unions and members need to minimise the risks associated with them.
The patterns of conflict reflect the institutional history of Australian industrial relations and the way that parties, habituated by former systems, have adapted the rules of successive legislative regimes.”
Peetz, D, ‘Industrial Conflict with Awards, Choices and Fairness’, Anthony Forsyth & Breen Creighton (eds), Rediscovering Collective Bargaining: Australia’s Fair Work Act in International Perspective, New York: Routledge, 2012, 159-181.
This is Attachment C in Professor David Peetz’s submission to the Productivity Commission. His submission covers many issues, here reproduced only his interesting research on strikes.
See Peetz in submissions see http://www.pc.gov.au/inquiries/current/workplace-relations/submissions
“When all else fails, under the logic of an arbitral system, you go to the independent umpire. When all else fails, under the logic of a collective bargaining system, you strike. The shift in Australia from an arbitration-based model to a bargaining-based model would be expected to have had some major effects on strikes: their number, causes and duration and how they were resolved.
Likewise, the re-collectivisation of employment relations through the Fair Work Act 2009 (FW Act), after the Coalition’s policy of individualisation, could be expected further to affect strike patterns. This chapter examines changing patterns of industrial conflict in Australia under a number of different policy regimes, including the early period of the FW Act. It first, briefly summarises some of the economic literature on strikes, then posits an alternative model which attempts to integrate economic and institutional explanations, relates the model to the historical characteristics of phases in Australian industrial relations policy, discusses data sources and methods, analyses aggregate trends in strike density and considers patterns regarding their causes, duration and termination.
Theorising industrial conflict patterns
A series of economic models have been used to try to explain strike activity. In Hicks’ (1966) early model of demands and offers, the majority of strikes arise from inadequate appreciation by each side of the other’s situation. Ashenfelter and Johnson’s (1969: 36-7) model introduced a separate role for union leadership as in effect a mediator between members and the employer. Later Kennan (1980) and Reder and Neumann (1980) separately and simultaneously proposed the ‘joint cost theory’, in which strike duration was inversely related to the sum of its cost to the firm and union. Theories of strikes were based on unilateral asymmetric information: the firm knows more about its profits than does the union, which uses strikes to test the firm’s response and thereby infer information about profits – or the firm signals its (un) willingness to pay by how long it is willing to endure a strike (Hayes 1984; Tracy 1986,1987; Booth and Cressy 1987; Cramton and Tracy 2003). While subject to various critiques (Paldam and Pedersen 1982; Shalev 1980; Ehrenberg and Smith 1988:471), each of these models offers something of interest to the theory of strikes.
Yet none of the above models appear to offer an adequate explanation of strike activity in the Australian context. Most focus heavily on wage related disputes, ignoring the possibility that there may be very different forces at work in the determination of non-wage disputes.
Moreover, none of the models seriously considers the possibility of unconditional, fixed duration strikes – that is, strikes without any negotiation between the parties or any wage increase (or other form of settlement) being agreed between or conceded by the parties. Each model essentially assumes that the strike continues until a new agreement or contract is reached between the parties.
These models’ relevance to the Australian situation is further complicated by the historical role played by arbitral tribunals in dispute settlement in this country. Although there is a body of literature concerning the use and effects of arbitration (eg Farber and Katz 1979; Hirsch and Donn 1982), it has neither been integrated into the above strike models nor applied to Australia, even in Australian studies that have made use of the above models (eg Bentley and Hughes 1970; Phipps 1977; Paldam and Pedersen 1982).
This chapter addresses some of these deficiencies by use of a simple model of industrial conflict that incorporates institutional influences on behaviour. It posits that strikes occur when, for both parties, the expected costs of the strike are lower than the expected benefits, and will continue until, for one of the parties, the marginal expected costs of continuing exceed the marginal expected benefits.
A thorough overview of all the factors involved is beyond the scope of this chapter, though factors affecting the expected costs and benefits to workers and the union would likely include:
the expected wages foregone while labour is withdrawn; expected penalties associated with the strike (the likelihood of possible penalties multiplied by their size); expected gains (eg better wages and conditions) from resolving the dispute (influenced inter alia by perceptions of an ‘intensity’ – that is, the firm’s intent, depth of feeling and militancy in its approach to resolution of the strike); expected (dis)utility from the experience of participating in a strike (sense of solidarity, retaliatory harassment, additional (or foregone) leisure etc); the expected impact on employment; and the impact on future bargaining power (Hicks 1966).
An important benefit from a strike may be to indicate the union’s own intensity: to impart a message about the costs it is willing to impose on the employer. The expected benefits will also be dependent on the mobilising ability of the union. If employees lack the institutional mechanisms to mobilise effective industrial action the costs of taking strike action will be too high (eg Kelly 1998).
For firms, the costs and benefits of a dispute will be influenced by its impact upon profits, inventories and goodwill; the proportion of the workforce who strike; the ability to replace striking workers; the expected financial cost of conceding compared to that of not conceding; and the firm’s expectations as to the union’s intensity. The firm and the union have imperfect knowledge of each other’s positions, particularly their intensity, and the greater the uncertainty (or the misalignment of perceptions), the greater the chance of a dispute occurring. Through the behaviour of their opponent during a dispute (and, for the union, their own members) they have inferred information about the likely outcome of the strike, which may cause them to revise the expected outcome.
This chapter is about policy, and as some of the factors described above are not directly affected by policy, they are not examined further here. However, some of them are clearly influenced by policy.
For example, the expected level of penalties for taking industrial action affects costs facing unions and hence their members and their likelihood of taking industrial action. So too will the scope of activities which are likely to incur penalties, and the procedural costs parties must incur to avoid penalties.
The expected cost to unions of taking strike action is influenced by policies that affect their ability to mobilise, the insecurity of employees or the propensity of employers to be belligerent in their dealings with unions.
That said, the behaviour of parties in disputes reflects ingrained patterns of behaviour – ‘cultures’ – that develop and are reinforced over years, indeed decades.
Major policy changes are frequently aimed at changing the culture of industrial relations. For example, the ‘two tier’ wage system of 1986-87 (under which wage increases were available at both the local level, to facilitate bargaining over productivity, and the national level) was aimed in part at promoting a union culture that was more supportive of workplace efficiency.
The period of the WR Act, including the intended demonstration effect of confronting key unions (such as the Maritime Union of Australia in the waterfront dismissals of 1998), and the Work Choices legislation of 2005 (with its prohibitions on union-friendly clauses in agreements between consenting parties, as discussed by Cooper and Ellem in this volume), were implicitly aimed at changing employer culture, away from accommodation with unions and in favour of union avoidance.
But culture is not so easily changed, and even under new policy regimes dispute behaviour may retain elements of cultures developed at earlier times.
A union benefits from imparting information about its ‘intensity’ provided it means the firm revises upwards the minimum outcome it expects to concede. At the same time, the union would like to reduce uncertainty about the duration and hence cost of strikes.
Accordingly, the union has an incentive to undertake short, fixed-duration strikes that impart information about its intensity and minimise the level of and uncertainty about its own costs. Consequently, some strikes, on both wage and non- wage matters, would be short and end without immediate resolution of the issues concerned. These can be called unconditional strikes.
In contrast, when contingent strikes commence there is uncertainty about likely strike duration. The regulatory and cultural environments may influence the balance between the two.
Under the traditional award system, the federal industrial tribunal had powers of compulsory conciliation and arbitration. An award made in settlement of an industrial dispute had a fixed term, but continued in force after its expiry until it was subsequently varied or replaced. Any party had the capacity to reopen the dispute, the settlement of which had led to the award in the first place. Consequently, many believed there was no real commitment to implementing the provisions of an award or agreement once it was made. In theory, penalties for breaches of awards (including by taking industrial action) applied at any time, so there was no distinction between when industrial action could be taken and when it could not.
And penalties were rarely applied (especially after the jailing of union leader Clarrie O’Shea in 1969: Deery and Plowman 1980).
‘Since all strikes are outlawed’, wrote Niland (1978: 52), ‘it is not possible to effectively prohibit those types of strikes which are seen as especially detrimental’.
In effect, there was no significant distinction between interest disputes and rights disputes in Australia. There was little incentive for unions to refrain from industrial action for any but a short time after an award had been created in settlement of a previous dispute.
In the mid 1980s, as the economy was subjected to greater competitive pressures, employers in some very public disputes (such as at the Robe River iron ore mine in Western Australia, the Mudginberri meatworks in Northern Territory, the Dollar Sweets confectionery factory in Melbourne and the State-owned South East Queensland Electricity Board) made full use of the multiplicity of forms of legal recourse available (McCarthy 1985; Creighton 1987).
These could be used to halt industrial action or to recover any losses suffered during the course of the bargaining process.
In this context, the ILO Committee of Experts observed that existing law ‘appears to deny workers the right to take industrial action’ and called for remedial action (ILO Conference Committee on the Application of Conventions and Recommendations 1989).
Moreover, as enterprise bargaining took hold from 1991, there were contradictions in seeking to promote direct bargaining between the parties while employers had an array of common law and statutory weapons that could be used against a union taking industrial action over a direct bargaining claim.
The Industrial Relations Reform Act 1993 consequently established that parties engaged in an interest dispute negotiating a single-employer agreement now had a qualified right to take industrial action. The 1993 Act also clarified that penalties would apply for parties taking industrial action in a rights dispute in breach of an agreement.
The right to strike while negotiating an agreement was retained but amended in the Coalition’s Workplace Relations Act 1996 (WR Act) and again under the Work Choices legislation that took effect from 27 March 2006.
Work Choices also shifted the balance of power in favour of employers. It strongly privileged individual Australian Workplace Agreements (AWAs) that could undercut award conditions.
It improved the ability of employers to temporarily replace striking permanent workers with casual or labour hire staff or contractors by prohibiting restrictions on their use in agreements as well as in awards.
It substantially increased the fines for unprotected industrial action and the requirements that must be followed for action to be protected while restricting union access to workplaces (McCallum 2006; Cooper and Ellem 2008).
It reduced the confidence and hence bargaining power of workers in small and medium firms by abolishing protections against unfair dismissal – or in all firms where dismissals could, in part, be attributed to ‘operational reasons’. This left many workers reportedly fearful of even challenging managerial decisions, let alone going further and organising industrial action against their employer (Elton et al 2007; Workplace Rights Advocate 2007), even if at law an employer could not dismiss striking workers.
As explained by McCrystal elsewhere in this volume, many of the features of Work Choices regarding sanctions for undertaking industrial action remained largely intact under the FW Act, though secret ballot procedures before taking industrial action were marginally streamlined. However, several features of the 2009 Act represent a significant break from Work Choices and could affect parties’ behaviour in disputation.
These include: the potential for Fair Work Australia (FWA) to make majority support determinations, issue scope orders, and make bargaining orders requiring parties to bargain in good faith; the removal of employers’ ability to use
AWAs to marginalise unions and de-collectivise employment relations; the removal of employers’ capacity to use individual agreements (in combination with the insecurity arising from the lack of unfair dismissal protections in smaller organisations) to reduce conditions to levels below those established in awards; and improved general protections, including for employees exercising their industrial rights (Sappideen, O’Grady, Riley and Warburton 2011:457-72; French, Harpur and Muurlink 2012 (forthcoming):47-51).
These legislative and policy developments raise a number of questions that can be addressed using data on industrial disputes, covering the period up to and including the commencement of the FW Act.
Does the introduction of a capacity lawfully to take industrial action increase strike levels? Or by clarifying when action is and is not permissible, does it reduce strikes in those circumstances where it is no longer permissible? The potential for greater penalties under the WR Act, increased under Work Choices, and the impediments created to union mobilising, might be expected to lead to substantial reductions in strike activity under that regime. The same might be said for the reduced overall power of individuals due to unfair dismissal changes, even if they could contest sackings attributable to participating in protected industrial action. Did that happen? Was that reversed with limited liberalisation under the FW Act? In what ways do the impacts of policy changes differ, between different types of disputes? How important were unconditional strikes as a signalling device under the traditional award system, with its absence of distinction between contingent and unconditional strikes? After the introduction of the capacity lawfully to take industrial action, were unprotected strikes shorter than protected strikes, because employees undertaking unprotected action need to minimise the risk arising from the action? After the shift to enterprise negotiation, was there a major increase in the proportion of disputes ended through negotiation, consistent with policy makers’ objectives? Or did the persistence of existing industrial relations culture mean unconditional strikes still had a significant role? Did the establishment of a right to strike regime lead to the virtual disappearance of unprotected strikes, or did the parties adapt to such a regime by continuing to use patterns of behaviour that had served them well in the past?
Data and concepts
The data for this chapter are derived from those published by the Australian Bureau of Statistics (ABS) and relate to the direct withdrawal of labour. That is, they cover strikes and stop work meetings, but make no distinction between the two. They do not encompass ‘quasi-strikes’: bans, limitations on work or boycotts that might not involve the employees losing wages (Dabscheck and Niland 1981: 41). The data used here come from several electronic and (discontinued) hard copy annual, quarterly and monthly ABS sources (Cat Nos 6322.0, 6321.0, 6321.0.55.01, 6101.0, 6310.0, 6323.0, and 6325.0). Because of the way in which the data are collected (the ABS relies on a range of intelligence sources to be alerted to industrial disputes, and then sends questionnaires to employers it knows are affected, but conducts no regular census or random survey of employers to identify whether strikes have occurred), some disputes are missed.
The data are presented below in different periods, corresponding to policy regimes in Australian industrial relations. They are:
• the traditional award system, which lasted up until the end of 1982;
• the centralised Accord period, from 1983 to 1990, during which an incomes policy involved the Labor government and Australian Council of Trade Unions (ACTU) negotiating over wage increases payable at the national level (covering four ‘Accord’ agreements known as Accords I to IV);
• the decentralised or ‘enterprise bargaining’ period of the ACTU-Labor Accord (the ‘EB Accord’), from 1991 to 1996 (as centralised wage determination was abandoned under processes set out vaguely in Accord V and explicitly in Accords VI and VII);
• the early and mid phases of the Coalition government’s WR Act, from 1997 to March quarter 2006 – the split between ‘early’ and ‘mid’ phases is driven by data availability: in 2003 the ABS changed its method of defining and publishing disputes data, fundamentally altering its definitions of ‘cause of dispute’ and, less radically, ‘method of settlement’ (renamed as ‘reason work resumed’) and began to distinguish between disputes over enterprise bargaining (‘EB disputes’) and disputes over other matters (‘non-EB disputes’);
• the ‘Work Choices’ (late) period of the WR Act, from June quarter 2006 to March quarter 2008;
• the transition period between Work Choices and the FW Act (June quarter 2008 to June quarter 2009), during which several of the most important elements of Work Choices (eg the capacity to make new AWAs) were no longer operative, but most of the remaining provisions of the WR Act remained in place;
• the FW Act period, commencing in September quarter 2009. The period encompassed by this chapter (the first eighteen months of the FW Act) is referred to as the ‘early FW Act’ period.
Data on the method of settlement of strikes, that is the reason work resumed, concern the ending of the strike, not the settlement of the underlying dispute. Several measures of industrial conflict are referred to for purposes of this chapter:
strike frequency (the number of strikes); strike duration; strike size (the number of workers involved); strike volume (the number of working days lost as a result of strikes); and strike density (strike volume divided by the number of workers, typically measured as working days lost per thousand employees (WDL/1000 employees)). The ABS does not publish cross-tabulations of these various indicators of industrial action, but it is possible to use correlation or regression analysis to explore relationships between such measures. Strike duration may be unweighted (all strikes are weighted equally, as per the data that appear in the officially published ABS tables) or weighted (according to strikes’ size), that is total strike volume divided by total strike size (a measure commonly employed in statistical studies of strike patterns).
The data cover different time periods according to the availability and relevance of data. Due to the changes in definitions, adjustments were made by the author to splice data before and after 2003 to create some continuous series. Unconditional strikes were essentially encompassed by what the ABS termed those ‘resumed without negotiation’ and, after 2003, ‘pre-determined return to work’. Data for the early FW Act cover the eighteen months to December quarter 2010.
Also, since 2006 (end especially since 2009), the ABS has become more concerned about not publishing data on strike size and volume in some categories (eg disputes due to a certain cause in a particular quarter, such as disputes over health and safety in June quarter 2010) where only small numbers of strikes have occurred, because of the fear of releasing data about a specific dispute which are treated as confidential. The categories with missing data vary from quarter to quarter, and where this occurs data have been interpolated.
Up to March 1994, most strikes in Australia were technically unlawful and there was no formal right to strike or lockout. Australian law had followed English law which regarded strikes as a breach of contract (Ewing 1989). Moreover, the introduction of the conciliation and arbitration system at the turn of the 20th century had been aimed at making strikes redundant, and strikes had been explicitly illegal until 1930 (Deery and Plowman 1980). In practice, for most of the 20th century the absence of this formal right appears to have had little effect in reducing the level of industrial disputes; indeed, industrial conflict peaked in the 1917-1920 period while strikes were still explicitly illegal.
Under the traditional award system that operated until 1982, Australia had a relatively high rate of industrial disputation when compared to several other industrialised nations (Dabscheck and Niland 1981:69). From 1950 to 1982, an average of 415 working days were lost per thousand employees due to industrial action each year. This three-decade aggregate disguised a long term growth in strike density, from an average of 265 WDL/1000 employees in 1950-1966 to 575 WDL/1000 thousand employees per year over the period from 1967 to 1982, peaking at 1273 WDL/1000 employees in 1974.
The prices and incomes Accord between the ACTU and Labor Government led to a significant drop in strike volume, considerably greater than what could be explained by economic circumstances or by international trends (Beggs & Chapman 1987a, 1987b, 1988; Morris and Wilson 1994).
As shown in Table 1, during the first, centralised Accord period, from 1983 to 1990, strike volumes more than halved to just
229 WDL/1000 employees, as wages were largely taken off the bargaining table at the industry or workplace level and replaced by nationally determined increases.
The second phase of the Accord saw the shift to enterprise bargaining. Notably, there was a substantial drop in the density of industrial conflict with the shift from the centralised Accord to the EB Accord. The downward impact on strike volumes, of clarifying the periods when strikes were permissible and impermissible and establishing clear periods when penalties would apply, outweighed any upward impact from removing penalties for striking during the negotiation of new agreements.
Attachment A: Charts on wages and industrial conflict
￼Source: ABS Cat No 6345.0, Wage Price Index, Australia Table 1. Total Hourly Rates of Pay Excluding Bonuses: Sector, Original, Seasonally Adjusted and Trend, December 2014.
The early and mid phases of the WR Act saw a further substantial drop in strike density, with further significant falls again during the Work Choices era. This is consistent with the idea that reducing the costs to employers of withstanding strike action and increasing the costs of industrial action to employees reduces the volume of strikes.
Legislation encouraged casual, labour hire and contract employment. It made it more difficult for employees to mobilise in the first place and increased the likelihood of sanctions for taking unlawful action (McCrystal 2006).
By the time of Work Choices, the procedural hoops that had to be jumped through were so numerous and formidable that some saw the law as ‘virtually extinguishing the right to strike’ (White 2005) while employers often had ready access to alternative labour.
There is also the possibility that the continuing falls in strike density through this period reflected the continuing effects of a cultural shift against readily taking industrial action, a shift that had commenced in the 1980s with the Accord. It is difficult to measure the relative importance of this factor.
Over the first 18 months of the FW Act, average strike density was 28 per cent higher than during the two years of Work Choices.
This is not surprising, as the FW Act reduced some of the costs associated with strike action compared to Work Choices, particularly the costs associated with mobilising (eg by making ‘good faith’ bargaining orders available).
Still, strike density under the early FW Act was only a quarter of the levels experienced under the early and mid phases of the WR Act. Compared to the previous period of enterprise bargaining under a Labor government (the EB Accord), strike density under the early FW Act was almost 90 per cent lower. The fall in strike density is much greater than the decline in union density (and by implication, union power), as illustrated by the lower row of Table 1, and has occurred despite lower unemployment, which is often linked to higher levels of disputation (Beggs & Chapman 1987a, 1988; Bentley & Hughes 1970; Paldam & Pedersen 1982).
There have also been major changes in the industry composition of disputation, with mining accounting for much less than in the past, education, health and community services for more, but no evidence of a sustained fall in the construction industry’s
share, despite the extensive attempts at ‘reform’ in that industry (Forsyth, Gostencnik, Ross and Sharard 2007; Creighton and Stewart 2010:838-80).
In sum, the level of industrial conflict under the early FW Act is more like that under Work Choices than like any system that preceded it, and is much lower than under the EB Accord or the period preceding the move to enterprise bargaining.
Table 2 shows the importance of unconditional strikes in Australia. Under the traditional award system, unconditional strikes (those where work was resumed without negotiation) accounted for 47 per cent of strike volume. Unconditional strikes have not always been a feature of the Australian system. Up until the 1930s they consistently accounted for less than one fifth of strike volume, but as union strategy changed they grew in significance through the 1930s and 1940s until they accounted for roughly half of strike volume from the 1950s onwards (Perry 2005).
[INSERT TABLE 2 HERE]
Although no cross-tabulations of ‘type of strike’ against ‘cause of strike’ are published, ordinary least squares (OLS) regressions can be used to indicate the dispute issues that are most important in explaining variations in dispute terminations. Using quarterly data over the period 1970-1982, the frequency of unconditional strikes (and of strikes of one day in length or less) was most strongly explained by variations in disputes over wages issues, and vice versa. This was contrary to the conventional but untested wisdom that short strikes, terminated without negotiation, were grievance-related, ‘usually…protests against decisions and policies regarded as being detrimental to labour’ (Oxnam 1971:55; see also Niland 1978:49; Bentley and Hughes 1970:152).
Rather, unconditional strikes appeared part of normal industrial campaigning, largely explaining why unconditional, fixed duration strikes, like wage- related strikes, tended to be larger (717 workers/strike for unconditional strikes, compared to 337 workers/strike for contingent strikes).
This result suggested an historical pattern of wage bargaining very different to US-inspired strike models, in which strikes continue until a wage agreement is achieved. Wage bargaining frequently, though by no means always, involved industrial campaigns (often multi- employer) in which strikes of fixed rather than indefinite duration were unions’ most efficient weapon.
Under the traditional award system between 1963 and 1982, around a third of strikes were ended through the arbitration systems, and a little over a sixth through direct negotiation. Again, it was not always thus.
Over the first three decades of the twentieth century, strikes that ended after negotiation accounted for approaching two thirds of strike volume, but they declined to just one fifth by the 1950s. Despite the establishment of arbitration systems in the 1890s and 1900s, strikes that ended through legislative mechanisms (conciliation and arbitration) constituted only around one sixth of strike volume till the 1930s, but rose to nearly one half by the 1940s, before dropping to a third in the 1950s and 1960s (Perry 2005).
As the parties acculturated to how the arbitration system could be made to operate, union practices changed. They placed less reliance on protracted strikes where the withdrawal of labour continued until one side or the other capitulated, or at least an agreement could be reached, after a period of negotiations (consistent with many of the US-inspired strike models). They increased their emphasis on methods (contrary to those models) in which the strike was used as a signalling device to indicate the intensity of union positions, and to activate the processes of conciliation and arbitration.
The move to enterprise bargaining has seen the expected decline in the use of arbitration to end disputes. Arbitration’s share of strike volume fell to 23 per cent under the EB Accord, 19 per cent in the early and mid WR Acts, then only 8 per cent under Work Choices (under which the arbitration tribunals were largely neutered). Despite the increase in powers given to the tribunal under the FW Act compared to Work Choices, arbitration’s share of dispute volume is unchanged, at 8 percent, under the early FW Act.
However, the other outcome envisaged in this area – a greater reliance on negotiation to end strikes – has not materialised. After allowing for the likely impact of changing definitions in 2003, it seems that no period since the shift to enterprise bargaining has had a significantly higher usage of negotiation to end disputes than occurred in the traditional award period. Indeed, the use of negotiation to end disputes fell to a low point of 9 per cent of strike volume under Work Choices, and just 10 per cent under the FW Act.
The slack instead appears to have been taken up by unconditional disputes. Strikes where work resumed without negotiation, including those with an identifiable pre-determined return to work, accounted for a majority of strike volume through each of the enterprise bargaining periods, and have reached over 80 per cent of strike volume under the early FW Act.
It appears that the methods of handling conflict that evolved over a century of arbitration have not easily been swept away, and have instead been adapted to suit the new environment. Unconditional strikes are still commonly used as a major signalling device to establish the ‘intensity’ and power of employees.
This persistence of elements of the earlier culture is illustrated in Table 3, which shows the average size of disputes by method of resolution over the four periods since 2003. In each period, the largest disputes have been unconditional strikes. On average, over the whole period since 2003, unconditional disputes have had an average size over three times that of contingent strikes. This remains the case under the FW Act.
[INSERT TABLE 3 HERE]
The Coalition government of 1996-2007 encouraged the use of external mediation. Strikes that ended through mediation rose briefly to account for 9 per cent of strike volume in the Work Choices era. But they fell to non-significance after the demise of Work Choices, accounting for approximately 5 per cent of strike frequency and
around 1 per cent of strike volume under the FW Act. Since the demise of Work Choices, where mediation was used at all, it was for settlement of very small disputes.
Overall, the pattern of dispute settlement under the early FW Act is rather more like that under Work Choices than it was under the EB Accord period and is far removed from the patterns that existed before the shift to enterprise bargaining.
There were major shifts in strike duration through the first half of the twentieth century as the parties gradually adapted to using the arbitration system. In the 1910s and 1920s, average weighted strike duration was typically around 15 days.
Unions then adapted their tactics to the arbitration system and moved away from ‘knock-em- down-drag-em-out’ strikes that persisted until a settlement was reached, to a more strategic engagement with the arbitration system, including using strikes primarily as signalling devices. Average strike duration plummeted to just 2 days by the mid 1950s and was below that level through the 1960s.
Table 4 shows the distribution of strike volume by duration of strike over the various periods since 1963. There was a fall in the importance of longer disputes (of 5 or more days unweighted duration) from 49 per cent of strike volume under the traditional award system to 15 per cent under the EB Accord. The contribution of longer strikes has fluctuated since then, but increased to 41 per cent under the early FW Act. Very short disputes peaked under Work Choices. This partly reflects several ‘national days of action’ protesting against Work Choices, though most participants took time off work or participated outside work hours and so did not count as strikers. The Work Choices figures include stop work meetings, which received more publicity, and hence were more likely to become known to the ABS, because of a number of incidents where workers were, as required by the statute, docked four hours’ pay for unauthorised meetings, no matter how short their duration (eg ABC 2006; Koutsoukis 2006), so stop work meetings then sometimes ‘lasted’ for four hours. Overall, however, it is likely that this made only a marginal contribution to the figures.
Matching the move back towards longer disputes, the share of WDL due to disputes lasting a day or less, which grew to 63 per cent under Work Choices, fell to 13 per cent of WDL under the early FW Act, the lowest figure for any period since the 1950s. Nevertheless, 44 per cent of recorded disputes lasted one day or less under the early FW Act, while only 16 per cent lasted 5 or more days.
[INSERT TABLE 4 HERE]
Despite the dominance of enterprise bargaining for almost two decades, Australia retains a pattern of strike duration that is more like its own past than it is like other countries with bargaining-based systems.
In Britain, for example, in 2007, 65 per cent of working days were lost in disputes that lasted 5 days or over (Hale 2007), compared with previously mentioned figures of 15 per cent under Work Choices and 40 per cent under the FW Act. Indeed, long strikes were more common under the traditional award system than they are now. Recall that over the 1963-1982 period in Australia, 49 per cent of strike volume consisted of strikes lasting five days or more. This is well below the recent British figure, but still in excess of recent experience in Australia.
The comparison with Canada is even more stark. Between 1966 and 2001 an average of 53 per cent of strikes lasted 10 days or more (Briskin 2005). Under the early FW Act, the comparable figure was only 8 per cent.
Even under the FW Act, the strike is primarily used to signal intensity, alerting employers (and, very rarely, tribunals) to the significance of an issue and the willingness of the union to impose economic costs on the employer if its demands are not met (or the employer’s demands not withdrawn). Only occasionally does a union use a strike directly to inflict sufficient economic damage on an employer to force its capitulation. That is, the strike is strategically used to create a clearer indication of the expected costs to the employer if they fail to negotiate, rather than to impose actual costs that force them to concede.
[INSERT TABLE 5 HERE]
As shown in Table 5, in recent years unconditional disputes have typically had a shorter average duration than contingent disputes – on average, over the pre-FW Act periods, by 2.5 days to 1.1 days. However, the difference was not as great as in the traditional award period, when the mean weighted duration of contingent strikes was 5.1 days, compared to 1.5 days for unconditional strikes.
Interestingly, under the FW Act, average duration of unconditional strikes has increased, to double previous levels and 50 per cent above the level under the traditional award system. In each quarter of 2010, the average duration of unconditional strikes exceeded that of contingent strikes. This happened only twice out of 13 quarters before Work Choices, once during Work Choices and then in two of the five transitional quarters.
Unions are adjusting their strike behaviour to the requirements of the FW Act, which in turn are a modification of those introduced under Work Choices. Under the FW Act, they must specify, well in advance, the nature and timing of proposed industrial action (Creighton and Stewart 2010:820-4). With the level of planning and procedural hurdles that must be cleared, it makes sense for unions to make these unconditional strikes as potent a signal as possible. To act as an effective signalling device, unconditional strikes must be longer than before, because they are harder to deploy than before.
Cause of dispute
Throughout the greater part of the 20th century, the most common cause of disputes (in terms of strike volume) was wages. The principal exception (prior to the 1980s) was in the late 1930s and early 1940s, when managerial policy and working conditions were more common causes. In the late 1960s and early 1970s, the relative precedence of wages was accentuated as the wages system decentralised within an award framework, and the pressure leading up to the wages explosion of 1974 grew (Plowman 1981:23-4). As shown in Table 6, during the traditional award system period of 1966 to 1982, disputes over wages accounted for 56 percent of strike volume.
This changed markedly during the centralised Accord period, when wages disputes dropped to only 24 per cent of strike volume. In terms of absolute numbers, working days lost through wages disputes dropped by four-fifths. The fall in wages disputes accounted for 96 per cent of the total fall in WDL. They have remained a low proportion of total dispute volume – under one fifth – ever since (highlighting the weakness of strike models based solely around wages).
By contrast, working days lost through disputes over managerial policy rose by more than half under the centralised Accord, with the implication that some of the conflict that was suppressed by the wages rules of the Accord was transferred into disputes over managerial policy.
The EB Accord experienced a further drop in all types of industrial dispute apart from ‘other disputes’ (this last result reflecting the protests against the Victorian Government’s industrial relations legislation in 1992). This reinforces the notion that there was a change in strike culture associated with the introduction of enterprise bargaining and formal recognition of the right to strike. Interpreting the transition between these periods is also complicated by the fact that enterprise bargaining disputes that encompassed disputes over managerial policy as well as wages were normally classified as ‘managerial policy’, so the drop in the relative share of wages disputes between those periods is probably overstated.
[INSERT TABLE 6 HERE]
At the time, no separate data were collected on EB disputes but in 2003 this was remedied and Table 7 contains data on causes of disputes from that time.
Several things are noteworthy about Table 7.
First, under the early FW Act, around 24 per cent of working days lost were in disputes that were not related to enterprise bargaining.
There are two ways of looking at that number. Some may see it as remarkable that, given the sanctions that surround unprotected action, at least 24 per cent of working days lost were due to such disputes. Some behaviours that developed under arbitration
regimes were still effective under the FW Act (and, for that matter, Work Choices) despite being unlawful. Old cultures may change, but they do not necessarily disappear.
On the other hand, this number represents a significant drop in the incidence of unprotected disputes. Under Work Choices, 58 per cent of working days lost were in non-EB disputes. Under the FW Act, that proportion has more than halved. The FW Act is the first period in which a large majority of WDL are occurring only through the mechanism of disputes over EB. This may reflect several things, including greater enforcement of non-EB disputes by more heavily resourced inspection agencies, the slightly more generous availability of legitimate opportunities for industrial action under the FW Act, and access to beneficial elements of protected bargaining such as access to good faith bargaining processes.
That the majority of WDL during the Work Choices period were in non-EB disputes is especially noteworthy, as this regime brought about a much higher likelihood of sanctions against unprotected action (and hence higher expected cost) than had existed under the WR Act (McCrystal 2006). During the WR Act, 49 per cent of WDL were in non-EB disputes, so Work Choices, rather than abolishing unprotected action, saw a relative increase in its incidence (compared to that of protected action). Of course, this was in the context of a major reduction in total industrial action under Work Choices, but the fall in WDL in EB-related disputes (79 per cent) was greater than the fall in WDL in non-EB disputes (70 per cent).
[INSERT TABLE 7 HERE]
One factor in this is probably that some unions considered that it was so hard to engage legally in industrial action, they may as well do it unlawfully anyway.
(Note, though, that strikes in support of EB claims that were deemed unprotected because they failed to meet technical requirements concerning notice or ballots, would have been defined by the ABS as EB-related disputes. So the level of non-EB strikes understates the level of unprotected industrial action. As well, unprotected strikes are less likely to come to the attention of the ABS than protected EB claims, leading to further understatement of the level of unprotected action.) In addition, a number of unions sought to finalise enterprise agreements before Work Choices took effect, bringing forward some EB-related disputation from the Work Choices period to the preceding period.
Table 8 shows average duration per employee by cause of dispute. Disputes over enterprise bargaining tended to be longer than disputes over non-EB matters. This is not surprising, as non-EB disputes, being typically unlawful, would need to be short to avoid inviting employers or authorities to take action in the tribunal and/or courts to terminate the action and (perhaps) penalise the union. Perhaps more surprising is that the difference in average duration between the two is not greater.
[INSERT TABLE 8 HERE]
Table 9 shows the average size (number of workers involved) of disputes since 2003. Disputes over matters relating to enterprise bargaining typically involved more workers than disputes over non-EB matters. This is especially the case for disputes over remuneration. Non-EB disputes over remuneration ranged from half to less than a quarter the size of EB disputes over similar issues, probably because larger disputes are more visible to authorities and the media and hence more demanding of legitimacy. Yet by the time of the FW Act, disputes over non-EB matters had a slightly higher average size than disputes over non-EB matters. This did not reflect substantial growth in the average size of non-EB disputes. Rather, average EB dispute size under the FW Act fell noticeably.
[INSERT TABLE 9 HERE]
This is in the context of a long term decline in average size of disputes related to EB – though one not as great as might be expected in the context of the shift in the locus of bargaining. In the period from 1970-1982, disputes over wages were quite large, averaging 736 workers per strike. Some of these were multi-employer or even industry level strikes. Perhaps surprisingly, in the mid WR Act period, the average size of disputes over remuneration was still 599 employees per strike, including 789 in EB disputes over remuneration.
A substantial fall in the average size of wage disputes might have been expected as bargaining shifted from larger to smaller employers, but perhaps an offsetting factor was that workers in smaller enterprises may have been less likely to be strongly organised (there is a well-established positive relationship between size of workplace or organisation and union density: Millward and Stevens 1986; Visser 1991) and may have suffered a sufficiently large drop in bargaining power to make them unwilling to strike over wages. By the time of the FW Act, the average size of strikes over remuneration had fallen to 146 employees per dispute (150 in EB disputes over remuneration). This is unlikely to reflect a further shift in bargaining to lower levels. Indeed, the average size of enterprise agreements (the number of employees they cover) increased slightly in the private sector, and increased substantially in the public sector, between the Work Choices period and the early FW Act. The drop in average size of EB disputes is probably due to some improvement in power for workers in smaller workplaces and organisations.
However, as with many matters, caution is required because these data concern only the first six quarters of the FW Act.
Australia has its own unique pattern of industrial conflict that is distinct from the North American and British patterns and which is not readily explained by US- inspired strike models. It reflects the institutional history of Australian industrial relations and the way that parties habituated by former systems have adapted the rules of successive legislative regimes.
The data suggest a number of preliminary conclusions about industrial conflict before and under the early FW Act, bearing in mind they are based on the experience of the first eighteen months of the Act’s operation. Several of these conclusions are consistent with the model discussed earlier, which focused on the institutional (and economic) factors and strategic issues influencing the expected costs and benefits of industrial action.
The introduction of a qualified right to strike clarified when action is permissible and when it is not. This led to a reduction in strikes in those circumstances where it was no longer permissible. This trend has continued under the FW Act. The impediments to effective union organisation and collective bargaining that were enshrined in the WR Act and then Work Choices resulted in substantial reductions in strike activity. The subsequent easing of those restrictions under the FW Act has so far had little impact on strike activity.
While the volume of strike activity under the early FW Act is more like that under Work Choices than any earlier period, the patterns of strike behaviour have changed. The volume of unconditional strikes is proportionately greater and they are smaller but last for substantially longer. Disputes are much more focused on enterprise bargaining.
Unprotected strikes under the FW Act and earlier bargaining regimes remain shorter than protected strikes, as unions and members need to minimise the risks associated with them. Perhaps one of the most notable things is that despite nearly two decades of enterprise bargaining, there is still a substantial number of short, unlawful disputes on non-EB matters. If this is an indication of lawlessness, however, then it is a form of lawlessness that was lower under the FW Act than under Work Choices or any other regime.
Under the traditional award system, unconditional strikes evolved into an important signalling device of unions’ intensity, and became a frequent alternative to striking indefinitely until one party conceded. Even after the shift to enterprise bargaining, unconditional strikes still have a significant role. The introduction of burdensome procedural requirements under Work Choices, largely retained under the FW Act, favours unconditional strikes, as every aspect of industrial action, including timing, has to be planned well in advance. In consequence, unconditional strikes have become the dominant pattern under the FW Act.
The procedural requirements emphasise to unions their need to provide a strong signal to employers when action is taken. As a result, the average duration of those unconditional strikes that do take place has increased under the FW Act. Unions have deployed longer, more potent unconditional strikes. That said, unconditional strikes still only averaging a couple of days in duration.
Australian unions typically avoid becoming caught up in the much longer disputes, lasting two weeks or more, that have been more typical of the North American and, to a lesser extent, British systems. Short, unconditional strikes were never just about grievances and protests (cf Oxnam 1971), they were always an important tactic in bargaining campaigns. They have now increased in duration, and importance, because of the procedural requirements surrounding disputation.
One thing that did not happen was an increased role for negotiation in ending strikes, either under Work Choices or the FW Act. This is contrary to the implied expectations of policy makers. Enterprise-level bargaining might make parties focus more on enterprise-level negotiations, but once a strike has commenced negotiations play no greater role in termination them than they have in the past. Strikes simply do not usually last long enough for negotiation to become critical.
Changes in strike patterns take a long time fully to unfold due to the resilience of culture and tactical behaviours that develop under previous regimes. Over the long term, the patterns under the FW Act are consistent with the idea that Australia is gradually moving towards disputes patterns that are closer to those in other developed English-speaking countries. But the emphasis is on ‘gradually’ and full convergence seems unlikely.
Disputes under the FW Act retain many of the characteristics that the Australian system developed over the preceding hundred years. Unions have responded to an enterprise-based bargaining system by adapting and modifying existing traditions and behaviour, in ways that suggest Australian dispute patterns might never resemble those in the North American enterprise-based systems.
ABC (2006). Workers lose pay for helping widow. ABC News Online, Australian Broadcasting Corporation: 11 April. < http://www.abc.net.au/news/newsitems/200604/s1613278.htm>.
Ashenfelter, O. & Johnson, G.E. (1969) `Bargaining Theory, Trade Unions and Industrial Strike Activity’, American Economic Review, 59, March, 35-49.
Beggs, J.J. & Chapman, B.J. (1987a) `Australian Strike Activity in an International Context, 1964-85′, Journal of Industrial Relations, 29(2), 137-49.
—— & —— (1987b) `An Empirical Analysis of Australian Strike Activity: Estimating the Industrial Relations Effect of the First Three Years of the Prices and Incomes Accord’, Economic Record, March, 46-60.
—— & —— (1988) `Australian Strike Activity in an International Context, 1964-85: Rejoinder’, Journal of Industrial Relations, 30(2), June, 318-20.
Booth, A. & Cressy, R. (1987) `Strikes with Asymmetric Information: Theory and Evidence’, Centre for Economic Policy Research Discussion Paper No. 178, Australian National University, Canberra, August.
Briskin, L. (2005) ‘The Work Stoppages Data From Human Resources and Skills Development Canada: A Research Note’, Just Labour, 5, 80-89.
Cooper, R. and B. Ellem (2008). ‘The Neoliberal State, Trade Unions and Collective Bargaining in Australia.’ British Journal of Industrial Relations 46(3): 532-554.
Cramton, P. and J. Tracy (2003). Unions, Bargaining and Strikes. International Handbook of Trade Unions. J. T. Addison and C. Schnabel. Cheltenham, UK, Edward Elgar.
Creighton, B. (1987). ‘Trade Unions, the Law and the New Right.’ The New Right’s Australian Fantasy. K. Coghill. Melbourne, Penguin: 74-92.
Creighton, B. and A. Stewart (2010). Labour Law. Sydney, Federation Press.
Dabscheck, B. and J. Niland (1981). Industrial Relations in Australia. Sydney, George Allen & Unwin.
Deery, S. and D. Plowman (1980). Australian Industrial Relations. Sydney, McGraw- Hill.
Ehrenberg, R.G. & Smith, R.S. (1988) Modern Labour Economics: Theory and Public Policy, 3rd edition, Scott Foresman, Glenview (Ill.).
Elton, J., et al (2007). Women and WorkChoices: Impacts on the Low Pay Sector. Centre for Work + Life, University of South Australia, Adelaide.
Ewing K (1989) ‘The Right to Strike in Australia’, Australian Journal of Labour Law, 2(1), March, 18-39.
Farber, H.S. & Katz, H.C. (1979) `Interest Arbitration, Outcomes and the Incentive to Bargain’, Industrial and Labor Relations Review, 33(1), October, 55-63.
Forsyth, A., V. Gostencnik, I. Ross and T. Sharard (2007). Workplace relations in the building and construction industry. Sydney, LexisNexis Butterworths.
French, B., P. Harpur and O. Muurlink (2012 (forthcoming)). A Pandora’s Box of General Protections under the Fair Work Act 2009. The Fair Work Act: Revision or Restitution. K. Abbott, B. Hearn-Mackinnon, L. Morris and K. Saville. Melbourne Heidelberg Press.
Hale, D. (2007) Labour disputes in 2007, Economic & Labour Market Review, Vol 2 No 6 June 2008
Hayes, B. (1984) `Unions and Strikes with Asymmetric Information’, Journal of Labour Economics, 2(1), 57-83.
Hicks, J.R. (1966) The Theory of Wages, 2nd edition, New York: St. Martin’s Press. Hirsch, B. T. & Donn, C.B. (1982) `Arbitration and the Incentive to Bargain: The
Role of Expectations and Costs’, Journal of Labour Research, 3(1), Winter, 55-68.
ILO Conference Committee on the Application of Conventions and Recommendations (1989). Direct request relating to Convention No 87 from the Committee of Experts on the Application of Conventions and Recommendations to the Australian Government,. Geneva, International Labour Office.
Kelly JE (1998) Rethinking Industrial Relations: Mobilization, collectivism, and long waves. London, New York: Routledge.
Kennan, J. (1985) `The Duration of Contract Strikes in US Manufacturing’, Journal of Econometrics, 28, 5-28.
Koutsoukis, J. (2006). “Wages docked over two minutes.” Age. 26 November.
McCallum, R. (2006). “Plunder Down Under: Tramsplanting the Anglo-American Labor Law Model to Australia.” Comparative Labor Law and Policy Journal 26: 381-400.
McCarthy, P. (1985). “Power Without Glory: The Queensland Electricity Dispute.” Journal of Industrial Relations 27(3): 364-382.
McCrystal, S. (2006). “Shifting the Balance of Power in Collective Bargaining: Australian Law, Industrial Action and Work Choices ” Economic and Labour Relations Review 16(2): 9.
Millward, N. and M. Stevens (1986). British Workplace Industrial Relations, 1980- 1984: The DE/ESRC/PSI/ACAS surveys. Aldershot, England, Gower.
Morris, A. & Wilson, K. (1993) An Economic Analysis of Australian Strike Activity 1959 to 1991. Paper to Australian Labour Market Research Workshop, Perth, February.
Niland, J.R. (1978) Collective Bargaining and Compulsory Arbitration in Australia, New South Wales University Press, Sydney.
Oxnam, D W (1971) `The Incidence of Strikes in Australia’, in J E Isaac & G W Ford (eds) Australian Labour Relations Readings: Second Edition, Sun, Melbourne, 19- 58.
Paldam, M. & Pedersen, P.J. (1982) `The Macroeconomic Strike Model: A Study of Seventeen Countries 1948-1975′, Industrial and Labour Relations Review, 35(1), July, 504-521.
Perry, L J (2005) A Long-Term Perspective on Industrial Disputes in Australia: 1913– 2003, Economic Papers Vol. 24 No. 3 September 2005 pp. 263−279
Phipps, A. (1977) `Strike Activity and Inflation in Australia’, Economic Record, 53(143), September, 297-319.
Plowman, D. (1981). Wage Indexation. Sydney, George Allen & Unwin. Sappideen, C., P. O’Grady, J. Riley and G. Warburton (2011). Macken’s Law of
Employment. Sydney, Lawbook Co.
Shalev, M. (1980) `Trade Unionism and Economic Analysis: The Case of Industrial
Conflict’, Journal of Labour Research, Spring, 1(1), 133-73.
Smith, B. (2010). “Fair and Equal in the World of Work: Two significant Federal Developments in Australian Discrimination Law.” Australian Journal of Labour Law(23): 3.
Tracy, J.S. (1986) `A Investigation into the Determinants of U.S. Strike Activity’, American Economic Review, 76, June, 423-36.
—— (1987) `An Empirical Test of an Asymmetric Information Model of Strikes’, Journal of Labour Economics, 5(2), 149-73.
Visser, E. J. (1991). Trends in trade union membership. Employment Outlook. Paris, Organisation for Economic Cooperation and Development: 97-134.
White, C. (2005). Inside the tent: The right to strike in Australia. Sydney, Evatt Foundation.
Workplace Rights Advocate. (2007) Report of the Inquiry into the Impact of the Federal Government’s Work Choices Legislation on Workers and Employers in the Victorian Retail and Hospitality Industries. Melbourne: Office of the Workplace Rights Advocate, 2007.
Table 1: Working days lost per thousand employees and per thousand union members, by period, 1967-2010
Source: Reserve Bank of Australia, Historical Tables 4.21; ABS Cat Nos 6321.0.55.001; 6310.0; 6323.0; 6325.0.
￼￼￼￼￼￼Traditional award system ￼￼￼Centralised Accord ￼￼￼￼￼EB Accord ￼WRA Early & Mid ￼￼￼￼￼Work Choices Transition
￼￼￼￼￼FW Act ￼￼￼￼￼￼￼￼1967- 1982
￼￼￼￼￼Working days lost per thousand employees
￼￼￼Working days lost per thousand union members
Table 2: Reason work resumed, share of all working days lost, by period, 1963- 2010
￼￼￼￼￼￼Traditional award system ￼￼￼Centralised Accord
EB Accord ￼WRA Early & Mid ￼￼￼￼￼Work Choices Transition ￼￼￼￼￼FW Act ￼￼￼Period
￼￼￼- Pre-determined return to work
￼￼￼- Resumption without negotiation
￼￼￼￼￼Conciliation and arbitration
￼￼￼￼- State legislation
￼￼￼￼￼- Federal (plus joint Federal-State legislation, pre-2003)
￼￼￼# These cells cover the period from 2003-2006 only, and hence do not add to their category subtotals.
* includes some interpolated data for at least one quarter
Ø Less than 0.5 per cent
Note: The change in definitions of method work resumed in 2003 means that there are some discrepancies in data between the pre-2003 and post-2003 estimates. Data were double-coded in 2003, and correspondence analysis suggests that, relative to the post-2003 definitions, the pre-2003 definitions led to the following understatements (-) or overstatements (+) of WDL: negotiation -2.7 percentage points; conciliation and arbitration +0.1 points; state
legislation +2.1 points; federal legislation -2.0 points; unconditional strikes (resumption without negotiation) +1.0 points; other +1.6 points.
Table 3: Average number of workers involved per dispute per quarter, by
method of resolution, by period, 2003-2010
(Mid) Work Choices 2003- 2006(2)- 2006(1) 2008(1)
737 571 533 2193
Transition FW Act 2008(2)- 2009(3)- 2009(2) 2010(3) 1158 361
1532 380 229 138
184 101 160 120 203 157 17 50
815 267 Note: excludes quarters where data suppressed from publication for a particular category
Pre-determined return to work
Resumption without negotiation
intervention of a third party
Other methods 82
170 94 119 109 247 203 192 348
Table 4: Unweighted duration of dispute, share of total working days lost, by period, 1963-2010
￼￼￼￼￼￼Traditional award system
￼WRA Early & Mid
￼￼￼￼Up to and including 1 day
￼￼￼￼Over 1 day and up to & including 2 days
￼￼￼￼Over 2 and less than 5 days
￼￼￼￼5 and less than 10 days
￼￼￼￼10 days and over
￼￼￼Total all WDL
￼￼￼* includes some interpolated data for at least one quarter Source: ABS Cat No 6321.0.55.001 and 6321.0
Table 5: Average weighted duration of disputes per employee, by method of
resolution, by period, 2003-2010
Pre-determined return to work
Resumption without negotiation
(Mid) Work Choices Transition 2003- 2006(2)- 2008(2)- 2006(1) 2008(1) 2009(2) 1.2 0.9 1.2
1.7 1.1 1.1 0.9 0.8 1.4
2.6 2.3 1.8
FW Act 2009(3)- 2010(4) 2.2
2.9 3.7 1.3 1.3
intervention of a third party
Other methods 12.2
1.5 1.1 1.2
2.7 3.4 2.2 1.6 1.1 1.3 2.6 1.6 1.4 2.4 2.0 1.0
￼Note: excludes quarters where data are suppressed from publication for a particular category. As a result, the average durations for particular categories may appear inconsistent with average duration for ‘total reasons’.
Table 6: Cause of dispute (issue), share of total working days lost, by period, 1966-2010
￼￼￼￼￼￼Traditional award system
￼WRA Early & Mid
￼￼￼￼￼Managerial policy, hours and leave
￼￼￼￼￼Physical working conditions
￼￼￼* includes some interpolated data for at least one quarter Ø less than 0.5 per cent.
Note: Definitional changes in cause of dispute in 2003 may affect some comparisons, though the effects are not likely to be large.
Table 7: Cause of dispute (enterprise bargaining versus non-EB), share of total
working days lost, by period, 2003-2010
Remuneration Employment conditions Other
Remuneration Employment conditions Health and safety
Job security Managerial policy Union issues
Work Choices Transition 2003- 2006(2)-2008(1) 2008(2)-
12% 65% Ø*
2006(1) 2009(2) 51% 42% 56%* 34% 16% 16% 14% 25% 40%* 3% 2% Ø
17% 39% Ø* 100.0% 100.0% 100.0%
* includes some interpolated data for at least one quarter for a particular category Ø less than 0.5 per cent.
44%* 24%* 1%* Ø* 28%* 11%* 3% 4%* 1%* 1%* 9% 1%* 1%* 5%*
Table 8: Average weighted duration of disputes per employee, by cause of dispute, by period, 2003-2010
￼Enterprise Bargaining Remuneration Employment conditions Other
Non-enterprise Bargaining Remuneration Employment conditions Health and safety
Job security Managerial policy Union issues
(Mid) Work Choices Transition 2003- 2006(2)- 2008(2)- 2006(1) 2008(1) 2009(2) 2.4 1.5 1.0
2.2 2.0 1.1 3.2 1.2 1.0 2.2 3.5 0.8* 1.1 0.9 1.5 2.2 1.2* 1.3
1.2 1.0* 1.6*
1.3 1.1 1.1
2.2 1.3 1.2 1.5 1.4 2.1 0.9 0.8 0.9 0.7 0.8 0.6* 1.5 1.1 1.2
FW Act 2009(3)- 2010(4) 2.6
1.3* 4.1* 1.0 3.0* 0.6 0.7* 0.4* 2.2
￼￼* These cells should be interpreted with great caution as they are based on observations from less than
Note: All cells except the bottom row excludes quarters where data are suppressed from publication for a particular category. As a result, the average durations for particular categories may appear inconsistent with average duration for all disputes.
Table 9: Average number of workers involved per dispute, by cause of
dispute, by period, 2003-2010
(Mid) Work Choices 2003- 2006(2)- 2006(1) 2008(1)
Transition FW Act 2008(2)- 2009(3)- 2009(2) 2010(4) 1339 259
Enterprise Bargaining 599
Remuneration 789 Employment conditions 432 Other 235 Non-enterprise Bargaining 303 Remuneration 157 Employment conditions 122 Health and safety 185 Job security 222 Managerial policy 143 Union issues 263 Other 2675 All disputes 397
798 150 1954 306
60* 120* 119 518 148
191 75* 4311* 617* 153 204 171 100* 245 129 225 850* 125* 320* 815 267
* These cells should be interpreted with great caution as they are based on observations from less than ten strikes.
Note: All cells except those in the bottom row exclude quarters where data are suppressed from publication for a particular category.