by Nick Wailes & Russell D. Lansbury

I. The types of flexibility introduced:
. . C. Pay determination procedures and systems

2. Decentralization of pay determination

Since the late 1980s, there has been a trend towards decentralization of wage determination in Australia, which has been driven by the policies agreed to by the federal Labour government and the Australian Council of Trade Unions. The agreements reached between the ALP and the ACTU, known as the Accords, have set the framework within which this decentralization has taken placed, reflecting the desire to manage and coordinate the process of decentralization of wage determination and bargaining (see Lansbury, 1985). The constitutional limitations placed on the federal industrial relations power has meant that decentralization has relied on the incorporation of decentralized wage determination principles into the decisions made by the arbitral body in National Wage Cases. In the process of decentralization of wage determination, NWCs have been transformed from a mechanism of centralized wage determination to a vehicle for determining the principles that should apply to decentralized wage determination (Dabsheck, 1994: 157). Recent changes to industrial relations legislation introduced by the current federal Liberal and National Coalition government have continued, and attempted to accelerate, this process of decentralization of wage determination. This section traces the decentralization of wage determination that has taken place in Australia through the vehicle of the Accords and the implementation (and alteration) of these agreements in National Wage Cases. It also draws attention to legislative efforts to enshrine principles of decentralized wage determination and to overcome the limitations placed on decentralization by the continued role of the AIRC. For a summary of the various Accords and recent NWCs, see Appendix 1.

The first Accord, negotiated before the 1983 federal election, allowed for the reestablishment of wage indexation in Australia for the first time since 1981. In the period September 1983 to December 1985, 96 per cent of wage increases in Australia resulted from the decisions of the AIRC in NWCs, and in the same period there were no substantial movements in over award pay (Dabsheck, 1994:155). While the centralization of wage determination, around the National Wage Cases, was initially associated with good economic performance, a series of exogenous shocks weakened Australia's economic performance during the mid 1980s. In the context of economic uncertainty, centralized wage indexation was abandoned in December 1986. In its place, attempts were made to tie wage increases to improvements in productivity and performance of industries and firms, and over time this has meant that wage determination has become increasingly decentralized.

The first stage in the movement to "managed decentralism" (Rimmer and MacDonald, 1989) in wage determination was contained in the renegotiations of the Accord between the ACTU and the ALP in the context of the economic downturn in the mid 1980s. Accord Mark III resulted in an agreement that wage increases should be divided into two tiers. The first tier at the national level was to provide a minimal increase of all wage levels covered by federal awards. The second source of wage increases were to be made available at the industry and enterprise level where agreement had been reached on improving efficiency and productivity between employers and workers.

Accord Mark III was translated into specific wage principles by the AIRC in National Wage Cases in 1987, 1988 and 1989. The 1987 NWC granted two separate wage increases to be paid to all workers covered by federal awards. It also permitted wage increases of up to 4 per cent to be granted in individual awards where employers and unions had addressed issues of restrictive work and management practices, multi-skilling and broad banding of work classifications, reduction of demarcation barriers and changes in award classification. The 1988 NWC developed this approach further in the "structural efficiency principle". This allowed wage increases of 3 per cent to be incorporated in individual awards and agreements where unions and employers had reviewed awards in relation to skill-related career paths, award relativities, flexibility, minimum rates and discriminatory aspects of awards. This resulted in a mixture of industry-by-industry and employer-by-employer negotiations. The structural efficiency principle of wage determination was reinforced in the 1989 NWC which made available two wage increases: the first associated with Commission scrutiny of proposals for award restructuring and the second related to the progress made in introducing changes (Davis and Lansbury, 1997: 21-22).

Evidence presented by Rimmer and Zappala (1988: 586) suggests that there were significant increases in labour market flexibility associated with the process of managed decentralization. However, wage flexibility "was almost untouched by second tier bargaining" (ibid). Faced with evidence of this nature, continued economic uncertainty and mounting criticisms from employer groups and the political opposition, the ACTU and the federal Labour government reached agreement that "enterprise bargaining" would become the main source of wage increases in Accord Mark IV which was concluded in February 1990. The National Wage Case related to the principles in Accord Mark IV did not take place until late 1990. In its decision, released in April 1991, the IRC rejected the wage determination principles contained in Accord Mark IV. The reasons it gave for this rejection pointed to the unevenness of award restructuring introduced though the structural efficiency principle and its belief that further changes were necessary before enterprise bargaining could be considered. In particular, the Commission was concerned that not enough attention had been given "to the criteria for future wage increases other than those that [arise] from enterprise bargaining" (National Wage Case, April 1991: 38-40 quoted in Dabsheck, 1994: 159). The October 1991 National Wage Case decision by the AIRC eventually gave official approval to the principle of wage increases being negotiated primarily at the enterprise level, but still raised doubts over the practical consequences of the change. While the October 1991 NWC removed limitations previously placed on the level of wage increases, it preserved the structural efficiency principle as the framework through which enterprise agreements could be made.

During the course of 1992, the Labour Government, led by Paul Keating, introduced legislative changes to speed up the adoption of enterprise bargaining by limiting the ability of the AIRC to alter agreements reached at an enterprise level. It did so by allowing enterprise agreements to be registered subject to a "no disadvantage" test as opposed to the more stringent "public interest" test, through which the Commission had promoted the structural efficiency principle. The effect of this legislation was to create two routes through which wage deals in enterprise agreements could be registered.

In the lead up to the 1993 federal election, the ACTU and the federal Labour government concluded Accord Mark VII. The agreement focussed on the need to increase employment and unions agreed to make wage claims in line with this objective. Accord Mark VII reasserted the view that wage increases should be negotiated at the workplace level in enterprise agreements, and "directly linked to improving the productive performance of enterprises ...; continuation of change, in particular work organization and job design, and the process of reform should be integral to agreements" (ACTU/ALP, 1993: 5 quoted in Green, 1994: 100-101). Productivity-based pay increases were to be underpinned by access to "safety net increases" through arbitration, in cases where enterprise agreements could not be reached. The agreement set out the level and timing of safety net increases. The agreement therefore guaranteed the continued existence of an arbitrated award system underpinning workplace bargaining and industry level framework agreements (Short and Buchanan, 1995: 119).

The application for a National Wage Case, designed to implement the content of Accord Mark VII, was delayed until late 1993 primarily to encourage the growth of enterprise bargaining by limiting access to wage increases through other means. Faced with what it regarded as potential contradictions between the wage principles in Accord Mark VII, and the requirement to be able to arbitrate disputes on their merits, the AIRC issued its October 1993 NWC decision under the title "Review of Wage Fixing Principles". The Commission's primary concern was to prevent wage increases gained at the enterprise from flowing through to the award structure. In doing so it preserved and expanded the enterprise bargaining principles handed down in the October 1991 NWC and required that wage increases be linked to implementation of efficiency measures and increases in productivity. It also outlined a new arbitrated "safety net adjustment" principle which was designed to ensure that the movement to a decentralized system did not undermine minimum conditions provided by awards (ibid: 112-114). Under this principle, the NWC introduced a safety net wage increase of eight dollars a week for employees who had been able to benefit from wage increases through enterprise bargaining.

Consistent with the principles outlined in Accord Mark VII, and attempting to reconcile these with the approach outlined by the AIRC in the October 1993 NWC, the 1993 Reform Act created an industrial relations system with distinct award and bargaining streams. The bargaining stream provided for two types of enterprise agreements: certified agreements and enterprise flexibility agreements. Certified agreements required union involvement and were subject to a no disadvantage test in relation to the relevant award. Enterprise flexibility agreements did not require union involvement but were subject to a public interest test which was more onerous than the nondisadvantage test, and ensured a role for the AIRC in overseeing these agreements. The award stream was intended to provide safety net increases for those workers who could not be covered by an enterprise agreement. Therefore, the Reform Act sought to overcome the confusion that had developed over certification of enterprise agreements and to resolve the issues of concern that had been raised by the IRC in the April 1991 NWC and reiterated in the October 1993 NWC over the relationship between the award system and enterprise bargaining, and particularly the nature of wage increases from both streams.

In September 1994, the AIRC made three safety net increases of eight dollars available to workers who had not been able to achieve wage increases through enterprise bargaining⁶. This reflected an attempt by the Commission to ensure that the award safety net was at a level that was "secure, relevant and consistent but which did not discourage bargaining at the... enterprise level" (Short and Buchanan, 1995: 122). However, Short and Buchanan (1995) argue that the small size of the increases and their phasing were designed to prevent safety net increases from undermining workplace bargaining. To this extent the safety net increases were consistent with agreements over wage determination outlined in Accord Mark VII. Between October 1994 and March 1995, the second safety net increase was granted at the enterprise level in 65 cases. Between March and August 1995, 376 federal Awards were varied to include the second increase (Pragnell et al., 1996: 99).

Given these changes in legislation, and the extent to which wage increases beyond a minimal level became tied to enterprise agreements, it is important to assess the growth and coverage of enterprise agreements. By mid November 1993, 1322 agreements had been ratified by the Commission, covering about 38 per cent of all employees covered by federal awards (approximately 789,000 employees). About half of the agreements were registered under the provisions of the 1992 legislation. The other half were ratified under the enterprise bargaining principles contained in the October 1991 NWC. Most of these agreements provided for wage increases of between 4 and 6 per cent (Green, 1994: 109). By November 1994, there were approximately 2'700 federally registered enterprise agreements which covered 56 per cent of employees under federal jurisdiction. The average annual wage increase associated with these agreements was 4.1 per cent. The growth of union-based enterprise agreements was not matched by a dramatic rise of nonunion agreements during 1994. By November 1994, 20 such agreements had been finalized covering 3,000 employees (although one of these agreements covered 2,500 workers) (Short and Buchanan, 1995: 124). By September 1995, 5, 310 enterprise agreements had been registered at the federal level, covering 58 per cent of employees covered by federal awards. Pragnell et al. (1996) argue that the majority of new agreements registered during 1995 were from small businesses or represented second generation agreements, because coverage increased by only 1 per cent. The average wage increase associated with agreements registered in 1995 was 4.8 per cent. By November 1995, 113 non-union enterprise flexibility agreements had been approved by the Commission and 206 had been received. Many of these agreements continued to be associated with small enterprises (Pragnell et al., 1996: 100-101). As these figures suggest, enterprise bargaining grew significantly in the early 1990s in Australia and became the main source of wage increases for many workers.

The figures in table 1.6 show that as bargaining has become decentralized, the traditional pattern of wage increases flowing through the award structure has been altered significantly. The existence of differential wage increases can be regarded as an indicator of increased substantive wage flexibility in the Australian economy.

Despite the growth of enterprise bargaining in the early 1990s, significant sections of the workforce have their wages determined by the award system, and have not received wage increases through enterprise productivity bargaining. Concern over the increasing gap between wage increases granted by safety net adjustments and those gained by workers represented in enterprise bargaining, depicted in table 1.6, raised the issue of how to ensure equity in wage outcomes. Furthermore, considerable debate has ensued about the variation of wage increases gained by different groups of workers under enterprise bargaining. In particular, a number of commentators have pointed to the adverse consequences of decentralized wage determination on the gender wage gap (see Hammond and Harbridge, 1995).

In June 1995, Accord Mark VIII was released. It was designed to address some of the equity concerns raised by the differential effects of the growth of enterprise bargaining on wage levels. It was also aimed to provide overall guidelines for the negotiation of wage increases through enterprise bargaining. It reaffirmed the 5 per cent unemployment target contained in Accord Mark VII, guaranteed low inflation and committed unions to increasing productivity and seeking responsible wage outcomes. In part this was a response to criticisms from employer groups and the political opposition that decentralized wage determination could result in a wages explosion. However, in March 1996, the ALP lost the federal election to the Liberal and National Party Coalition. The change in government heralded an end to the Accord process and also brought to an end the process of "coordinated flexibility" that had characterized policy over wage determination in Australia since the early 1990s.

In reaction to the growing concerns about the adverse impact of decentralized bargaining on equity outcomes, the ACTU launched its Living Wage Claim to the AIRC in early 1996. The ACTU claim aimed to raise the minimum hourly rate of pay to all award employees to $AUD 12 an hour through three $AUD 20 a week wage increases for all employees who had not gained wage increases through enterprise bargaining. It also sought to reestablish the relationship between rates in the Metal Trades Award with other award rates to ensure a fair relativity structure. The primary aim of the claim was to stem the real wage decline experienced by low-paid workers, who have relied on awards, and have not been able to gain increases through enterprise bargaining (Buchanan and Watson, 1997). A joint submission by the federal government and all the state governments (except the NSW government, which is the only remaining Labour government in Australia) to the AIRC argued against the ACTU's claim. They argued that the award system should be regarded as a minimum safety net only and that the determination of actual wage rates should be solely based on enterprise bargaining. Deviation from this position, the government argued, would be at odds with the objectives of the WROA Act. The government submission also argued that if the Living Wage Claim were accepted, the result would be a blow out of wage levels, increases in inflation and a tightening of monetary policy. The negative economic consequences associated with these developments would impact disproportionately on the low paid. To prevent these negative economic consequences, the government submission argued that any safety net adjustment should be minimal and limited in its application to the low paid (Buchanan et al., 1997: 10-15).

In a split decision by the full bench, the AIRC did not grant the ACTU's claim. Instead the AIRC rewarded a $AUD 10 safety net adjustment to be made available to workers who have not received increases through enterprise bargaining. The majority decision placed heavy influence on the potential inflationary impacts of the Living Wage Claim, the position advanced by the federal government submission to the case. At the same time the Commission reiterated concerns about the growing differentials between workers in the award system and those in the enterprise bargaining stream (CCH, 1997).

The changes introduced by the WROA Act which are designed to allow pay determination to be solely based on enterprise-based agreements represent a continuation of the changes introduced by the previous Labour government in an effort to increase labour market flexibility, rather than a fundamental reform (Ronfeldt, 1997). The new features of the WROA Act, which differ from the previous system, include: the introduction of an individual employment contracting scheme outside the jurisdiction of the arbitration body; changes which undermine the monopoly rights and "interest representation" status of unions in collective bargaining; and the specification of 20 allowable matters which are to form the content of awards.

The individual employment contracting provisions allow for the introduction of Australian Workplace Agreements. These agreements are regarded as private employment contracts and are to be approved by a specialist body, the Office of the Employment Advocate, which is separate from the Arbitration Commission. Once approved, AWAs displace awards. Therefore, wage increases from the award stream will not flow through to AWAs and any wage increase will be based on agreement between an individual employer and individual employee. The experience of workers in New Zealand under the Employment Contracts Act 1991 suggest that workers covered by AWAs will find it difficult to conclude wage increases through individual bargaining (Maloney, 1994). However, considerable doubt remains over the extent to which AWAs will be adopted by Australian employers. The provisions for the establishment of AWAs are complicated and will require the specification of a comprehensive range of issues (Ronfeldt, 1997; ACIRRT, 1997, No. 12: 29). Despite the costs that could be associated with concluding individual bargains, there are some reasons to suggest that employers will be prepared to bear these costs. In particular, the character of the debate about enterprise bargaining in Australia during the 1990s has meant that many Australian employers see individual bargaining as necessary for better economic performance (Wailes, 1997).

The second major change associated with the WROA Act relates to representation in collective bargaining. While previous legislation allowed for the certification of nonunion collective enterprise agreements, i.e. enterprise flexibility agreements, unions were able to object to these agreements and the test imposed on these agreements by the AIRC was more stringent than that for union-based agreements. The change in legislation implies that nonunion enterprise agreements will be easier to conclude. The WROA Act also makes it more difficult for unions to act as effective representatives in enterprise bargaining. Unions' status has been altered from "interest representation" to "agency representation" by changes in the Act which require unions to demonstrate that a "valid majority" of workers want the union to bargain on their behalf prior to negotiation and that a "valid majority" accept the outcome of the bargaining (Ronfeldt, 1997). At the present stage, it is unclear what will be regarded as a valid majority but it is clear that these changes cast unions in a fundamentally different role and could potentially undermine the ability of unions to coordinate wage bargaining within and across sectors in the economy.

The third major change associated with the WROA Act is the limitation of awards, and consequently the arbitration power of the AIRC, to 20 allowable matters. The Act includes a timetable for the "stripping back" of awards in line with these allowable matters and implies that a great many more matters will have to be negotiated in enterprise agreements or will become unregulated. Having said this, the Commission continues to have to power to consider other issues which are required to make awards workable. The Commission has the power to adopt a broad definition of such powers. Nevertheless, the reduction of award regulation to 20 allowable matters completes the process begun by Labour in the 1993 Reform Act of "clipping the wings of the Commission" in an effort to increase labour market flexibility (Dabsheck, 1995). As such, it may further undermine the role of award regulation in setting conditions of work and pay rates in particular.

Therefore, during the course of the 1980s and into the 1990s, there have been significant changes in the process of wage determination for workers covered by federal awards in Australia. The process of wage determination has changed from highly centralized to decentralized. The process by which decentralization was introduced has been characterized by adjustment of the agreements over wage determination between unions and the government, contained in the Accords, by the AIRC through NWCs to ensure "coordinated flexibility". The 1993 Reform Act attempted to speed up the pace of decentralization by legislatively limiting the role of the AIRC in overseeing enterprise agreements. This approach has been continued by the current federal government, which is less concerned with coordinating flexibility than in more "genuine" decentralized wage determination. However, as the Living Wage Case demonstrates, the AIRC still has a considerable role in determining the relationship between the safety net award stream of wage determination and the decentralized enterprise system of wage determination.