Bulletin – October 2024 Australian Economy Developments in Wages Growth Across Pay-setting Methods

Introduction

Over the past few years, the Wage Price Index (WPI) has grown at its fastest rate in more than a decade, although appears to have passed its peak for the current cycle. This strength has been driven by a combination of the tight labour market and high inflation outcomes such that, despite the strong growth, real WPI has declined. Assessing the outlook for wages growth is important for assessing the inflation outlook, as labour costs are a major factor in firms’ pricing decisions. Further, wages are the largest source of household income, meaning wages growth has a significant impact on household consumption.

The wage system in Australia is made up of three distinct wage-setting methods: awards, enterprise bargaining agreements (EBAs) and individual arrangements. Wage dynamics can differ across these pay-setting methods and these differences can be important to consider when assessing the outlook for wages growth. For this reason, the Australian Bureau of Statistics (ABS) publishes estimates of contributions to wages growth by pay-setting method (ABS 2024a). One of the RBA’s methods of forecasting WPI growth is to combine forecasts for wages growth in each pay-setting method into an aggregate forecast. Across each method, WPI growth has picked up over the past two years but appears to have peaked (Graph 1). Growth is expected to continue to slow as the labour market eases but the rate of easing is expected to differ across the methods.

This article outlines recent developments in wages growth and describes the RBA’s approach to forecasting wages growth by pay-setting method. As part of its forecasting process, the RBA considers the decisions of the Fair Work Commission (the Commission), which sets minimum wages in modern awards. For this reason, this article also discusses the factors the Commission considers when making its annual wage review (AWR) determinations.

Individual arrangements

Around 40 per cent of employees have wages and conditions set on an individual basis. Given just over 70 per cent of employees on individual arrangements work full-time, individual arrangements make up the largest share of the wage bill of the three pay-setting methods (nearly 50 per cent) (ABS 2024b).

Wages set by individual arrangements tend to be more responsive to the economic cycle than wages set by other pay-setting methods (Bishop and Cassidy 2019). Given this, the RBA’s forecasts for wages growth in individual arrangements is informed by a ‘wages Phillips curve’ model. This model predicts wages growth based on its negative relationship with spare capacity in the labour market.^[1] The cyclical sensitivity of individual arrangements, along with their large share of the wage bill, means they account for much of the high-frequency cyclical variation in WPI growth.

Given that wages growth in individual arrangements is most responsive to changes in demand, the tightness in the labour market following the COVID-19 pandemic led to wages growth in individual arrangements picking up earlier than other pay-setting methods (Graph 1). With the labour market becoming less tight since late 2022, it appears that wages growth in individual arrangements has peaked. The Phillips curve suggests that wages growth in individual arrangements will continue to decline, consistent with conditions in the labour market expected to ease further over the forecast period.

The standard Phillips curve specification uses the unemployment gap as the measure of labour market spare capacity. The unemployment gap is the difference between the unemployment rate and an estimate of the non-accelerating inflation rate of unemployment (NAIRU).^[2] The RBA is currently expanding its suite of labour market indicators. At times, these measures can tell varying stories about the state of the labour market, and therefore may be useful in providing different insights on the outlook of wages growth. These include the hours-based underutilisation gap, and the quits rate and employer-to-employer transition rate measures developed using Longitudinal Labour Force Survey data from the ABS. The range of estimates produced from the Phillips curve using these alternative measures of slack is very wide, highlighting the difficulty in forecasting wages growth (Graph 2).

Enterprise bargaining agreements

EBAs are collective agreements negotiated at the enterprise level between an employer and a group of employees. Around 35 per cent of Australian employees are covered by an EBA (ABS 2024b). Changes to wages are pre-determined for the life of the agreement, which is an average length of three years. During the life of an agreement, employees cannot lawfully engage in industrial action in pursuit of further claims. EBAs therefore tend to be affected by labour market conditions with a lag.

Private sector enterprise bargaining agreements

Information on wage outcomes in private sector EBAs is available in the Workplace Agreement Database (WAD) maintained by the Australian Government Department of Employment and Workplace Relations (DEWR). The WAD provides information on the average annualised wage increase (AAWI) of federally registered EBAs, which includes all private sector EBAs that are in effect. The RBA has a model that uses the AAWIs to predict private EBA wages growth, which involves three steps:

1. Forecast AAWIs in new agreements – This is done with a Phillips curve model whose explanatory variables are a lag of AAWIs in new private sector EBAs, the unemployment gap and inflation expectations.
2. Forecast AAWIs in the stock of all agreements – This is done using the forecast from step 1 as an input.
3. Forecast private EBA WPI growth – This is done using the forecast from step 2 as an input.

There are some limitations to the existing model. First, the AAWI measures the average increase over the life of an agreement, and therefore does not account for the precise size and timing of wage increases. The underlying microdata of WAD contains detailed information on all agreements in effect including wage increases, the length and expiry of agreements, and number of employees covered. The microdata suggest that wage increases are not always uniform over the duration of an EBA, and agreements often have ‘front-loaded’ wage increases – that is, the first pay rise tends to be larger than subsequent increases over the agreement. There are several reasons for this, including a perception that inflation will decline over the life of an agreement or compensation for delays in negotiations.

Over the past two years, there has been a higher degree of front-loading in agreements, consistent with other periods of high inflation (Graph 3). As the AAWI measures the average wage increase over the agreement, front-loaded agreements may lead to AAWIs that underestimate wages growth in the near term and overestimate growth in later years. These factors are included in the private sector EBA forecast using judgement, although the RBA is continuing to evolve its forecasting framework for private sector EBAs.

A second limitation of the RBA model is that the AAWI series only includes agreements that provide for quantifiable wage increases over the life of the agreement. This means a large proportion of agreements are not being captured in the current forecasting framework. Agreements may be determined as ‘non-quantifiable’ for a number of reasons, including that the agreement contains wage changes that are:

• not consistent between groups of employees
• linked to performance
• linked to the Commission’s AWR or the Consumer Price Index (CPI).

Around 20 per cent of private sector agreements are non-quantifiable, and these agreements cover 40 per cent of private sector employees on an EBA. Given the proportion of agreements linked to the CPI or the AWR is known, the RBA can make assumptions about wage increases in these agreements based on CPI and award wage forecasts. For agreements with increases that are not consistent between groups or linked to performance, the RBA is currently looking into the usefulness of machine learning techniques in extracting the wage increase from these non-quantifiable agreements.

Wages growth in private sector EBAs appears to have peaked, broadly consistent with other EBA wages growth indicators (Graph 4). However, given the stickiness of EBA wages growth, growth in private sector EBAs is forecast to ease more gradually across the forecast horizon compared with individual arrangements.

Public sector enterprise bargaining agreements

Wages in the public sector tend to behave differently to those in the private sector, and tend to be driven by other factors such as government wage policy settings and budget balances. Public sector wages are predominantly made up of EBAs, with around 80 per cent of public sector employees being covered by an EBA.

To model wages growth in the public sector, the RBA monitors a representative sample of public sector EBAs based on published information.^[3] Public sector wages growth is forecast for each state or territory by aggregating scheduled wage increases in each major EBA relevant to that state or territory, with each EBA weighted by the number of employees. Public sector wages growth for Australia is given by aggregating the wages growth forecasts from all states and territories.

Prior to and during the pandemic, many state governments imposed annual caps on the maximum allowable wage increases for public sector workers. These caps were introduced to reduce state and territory budget deficits and ensure public sector employees were receiving wage increases consistent with those in the private sector. Since 2022, state and territory governments have raised or abolished the wage caps put in place before and during the pandemic. State and territory essential workers have also received large pay increases under recently negotiated EBAs. This has led to an increase in public sector wages growth over the past year (Graph 5). Although public sector WPI growth appears to be past its peak, it is expected to remain robust over the period ahead.

Awards

Awards are legally enforceable determinations that set out minimum terms and conditions of employment in addition to any legislated minimum terms. In its AWR, an Expert Panel of the Commission reviews the modern award minimum wages and the National Minimum Wage Order (NMW Order), and determines if they should be adjusted.

There are currently 121 modern awards that set minimum wages and conditions for a wide range of industries and occupations (FWC 2023). Around 20 per cent of all employees are paid at the applicable minimum wage rate in awards and are directly affected by the AWR (ABS 2024b). The characteristics of this cohort of employees are significantly different from those of the workforce as a whole:

• they predominately work part-time hours and are female
• almost half are casual employees
• compared with the general workforce, they are disproportionately low paid and employed by small businesses.

The NMW Order applies only to employees who are not covered by a modern award or enterprise agreement. An award- or agreement-free employee cannot be paid less than the applicable rate specified in the NMW Order.^[4] The practical application and effect of the NMW Order is very limited, with less than one per cent of all employees estimated to be paid the NMW.

The AWR process is set out in the Fair Work Act 2009 (Cth) (Fair Work Act). The Commission must conduct and complete the AWR in each financial year. Any orders and determinations made that change award minimum wages must come into operation on 1 July in the new financial year, unless there are exceptional circumstances.^[5] For example, over the pandemic period, the Commission adopted a staggered approach to implementing the increases to modern award minimum wages.

The Commission has also said that its decision-making process in an AWR should be as transparent as possible and disclose the factors most relevant in a particular year.^[6] Although the Expert Panels may be differently constituted year-to-year, they tend to adopt a consistent interpretation of the legislative framework. As the Commission has said, ‘[j]ustice requires consistent decision-making unless a difference can be articulated and applied’.^[7]

A reasonable opportunity must also be provided to all persons and bodies to make written submissions. Submissions are typically provided by governments, unions and employer associations, and academics.

The statutory framework and approach

In the AWR, the Commission must ensure the maintenance of a safety net of fair and relevant minimum wages.^[8] The relevant statutory objectives are broadly expressed and do not necessarily exhaust the matters that the Commission may consider to be relevant.

Economic, labour market and business considerations

The Commission must consider the likely impact of its determinations on employment growth, inflation and the sustainability, performance and competitiveness of the national economy.^[9] It has interpreted this as meaning that it must take into account ‘the effect of its decision on national economic prosperity’ and in doing so give ‘particular emphasis’ to employment growth and inflation.^[10]

The Commission considers both actual and forecasts of economic indicators, with actual indicators the primary consideration as they are viewed by the Commission as more reliable. The Commission considers the Living Cost Index for employee households alongside the CPI for changes in living standards and purchasing power, noting that price increases in non-discretionary items are more likely to adversely affect the household budgets of the low paid.

The Commission pays particular attention to trend data and routinely looks to developments over the medium and longer term, as well as to changes over the past year. Consistent with this, the Commission has also noted that short-term changes in productivity should be interpreted with caution and productivity growth is best measured over the business cycle.^[11] The main measure of productivity examined by the Commission is Gross Domestic Product (GDP) per hour worked.

Relative living standards and the needs of the low paid

The Commission must consider the relative living standards and the needs of the low paid.^[12] ‘Relative living standards’ is a comparative concept and requires a comparison of the living standards of award-reliant workers with other groups. The comparison of living standards is at the household level using equivalised household disposable income.

The ‘low paid’ have been defined as those employees whose ordinary-time earnings are below two-thirds of median adult ordinary-time earnings of all full-time employees. There are two measures of this benchmark:

• $1,066.67 per week (as at August 2023) from the ABS Characteristics of Employment data.
• $1,131.33 per week (as at May 2023) from the ABS Employee Earnings and Hours (EEH) survey data.

The ‘needs of the low paid’ requires an examination of the extent to which low-paid workers can purchase the essentials for a decent standard of living and to engage in community life. The Commission has accepted that if low-paid workers live in poverty then their needs are not being met. In measuring poverty, the Commission relies on ‘poverty lines based on a threshold of 60 per cent of median equivalised household disposable income’ and has stated that ‘those in full-time employment can reasonably expect to earn wages above a harsher measure of poverty’.^[13]

The Commission also considers legislated superannuation guarantee increases and changes in the tax/transfer system, noting that the latter can provide a more targeted approach than increases in minimum wages.

Gender equality

In 2022, amendments were made to the Fair Work Act requiring the Commission to consider the promotion of gender equality when performing its functions and exercising its powers.^[14] This applies to the modern award and minimum wage objectives, which now require the Commission to consider the need to achieve gender equality in the workplace by:

• ensuring equal remuneration for work of equal or comparable value
• eliminating gender-based undervaluation of work
• providing workplace conditions that facilitate women’s full economic participation.

In the AWR 2022–23, the Commission noted that there were significant issues concerning the potential undervaluation of work in modern award minimum wage rates applying to female-dominated industries and occupations. However, the scope of the AWR prevented these gender equality issues from being sufficiently addressed.^[15] Since then, the Commission commenced proceedings to consider variations to five identified priority awards on work value grounds to remedy potential gender undervaluation (see Box A for details).

Job security

The job security consideration primarily refers to whether the AWR outcome might affect the capacity of employers to continue to offer, or maintain permanent employment, in the future.^[16]

Collective bargaining

The Commission must consider ‘the need to encourage collective bargaining’, which requires attention to be given to whether the exercise of modern award powers may affect the extent to which enterprise bargaining is occurring.^[17] The Commission has consistently observed that a complex mix of factors may contribute to employee and employer decision-making on whether to bargain, and has expressed the view that increases in award reliance do not support the contention that minimum wage increases act as a disincentive to bargaining.^[18]

The outlook for annual wage reviews

The Commission has repeatedly noted that it will not adopt a mechanistic approach to award determinations, such as, for example, real wage maintenance as a ‘decision rule’. However, over the years it has outlined a range of principles that are used to guide its decision. Rather than using a model to forecast wages growth in awards, the RBA uses the principles outlined by the Commission and other information to forecast future AWR decisions across the forecast horizon.

The Commission has said that the AWR is not an ‘adjudication between competing proposals’, but a ‘statutory task’ that requires it to make its own ‘assessment of what constitutes a safety net of fair minimum wages having regard to the prescribed considerations’.^[19]

The Commission has noted that awarding an increase that is less than increases in prices and living costs would amount to a cut in real wages, and such an outcome would mean that many award-reliant employees, particularly low-paid employees, would be less able to meet their needs.^[20] On average, the increases in award wages since 2010 have been higher than year-ended headline inflation (Graph 6). However, in recent years the annual award wage increases have been closer to headline inflation.

In the two most recent AWR decisions, the Commission concluded that the immediate economic circumstances mitigated against awarding an increase above headline inflation. This included inflation being above the target range, insufficient evidence of productivity growth having returned to its pre-pandemic average rate, Stage 3 tax cuts and other Federal Budget cost-of-living measures, and the legislated superannuation guarantee increases.^[21]

Further, recently the Expert Panel has accepted that ‘[i]n the medium to long term, it is desirable that modern award minimum wages maintain their real value and increase in line with the trend rate of national productivity growth’.^[22] The simplest measure of labour productivity, GDP per hour, grows in part because of changes in the composition of the workforce. For example, the decline in the number of labourers and increases in the number of software engineers over time contributes to GDP per hour growth. The quality-adjusted labour productivity measure currently being developed by the RBA may be relevant in future AWR proceedings (Bruno, Hambur and Wang 2024).

The effects of annual wage review decisions on wages growth

Direct effects

The direct effect of the AWR on wages growth is limited to the award-reliant workforce. Given these employees tend to work part-time hours and are disproportionately low paid, award-reliant employees account for around 10 per cent of the wage bill (ABS 2024b).

Direct effects of the AWR also occur through EBAs directly linked to the AWR. Over 300,000 employees are on federally registered EBAs linked to the AWR (DEWR 2024). Nearly all EBAs in the hospitality and retail industries are linked to the AWR, as are many EBAs in the health care and social assistance industry. Further, the base pay rate in an enterprise agreement must be at least equivalent to the pay set out in the relevant award. As a result, employees on EBAs paid a rate close to the award may need to receive a pay increase, even if the EBA is not explicitly linked to the AWR, to ensure their wage remains at least in line with the award rate.

Indirect effects

Outside of direct effects, the AWR decision can also influence the wages for non-award reliant employees. This is because the decision can influence wage expectations, which can lead to different wage outcomes than would have occurred under a weaker or stronger AWR decision. For example, if an AWR decision is higher than expected, non-award employees may receive a higher wage change to maintain the same differential as award and non-award jobs.

Measuring the indirect effects of the AWR decision – also referred to as ‘spillovers’ – is difficult as the prevalence and size of spillovers is not directly observable. The RBA has several methods for capturing the proportion of jobs indirectly affected by the award decision using underlying WPI microdata. For example, the RBA looks at the share of ‘award-influenced’ jobs by using information provided by firms in the WPI survey about why a particular job’s wage changed in the quarter (Graph 7). The RBA also receives qualitative information from firms through its liaison program that is useful for estimating spillovers. For such purposes, the RBA assumes the award decision will spillover to around 10 per cent of individual arrangements and 15 per cent of EBAs (including those directly linked). However, these estimates are sensitive to assumptions.

Forecasting wages growth and broader labour costs

The forecast for each pay-setting method is aggregated together based on the WPI weights to provide a profile for total WPI. For technical reasons, the weights in the WPI for awards is a little less than its share of the wage bill, and the reverse is true for individual arrangements.^[23]

This forecasting framework complements the Reserve Bank’s suite of other wages models, including the Phillips curve model for the private sector in aggregate. Given the framework is only a few years old, the RBA has not been able to thoroughly evaluate the accuracy of forecasts derived from this disaggregated method. However, it has the benefit of providing a framework that accounts for the different dynamics across pay-setting methods. The framework is also useful for scenario analysis, including modelling the impact of award wage increases and proposed changes in government wage policies.

When compiling the wages growth profile, the RBA considers all models and makes further judgements to account for elements of wages growth the models do not sufficiently capture. This includes, for example, the legislated superannuation guarantee increases, which analysis suggests should detract from growth in base wages,^[24] and the extent to which real wages are expected to ‘catch up’ to their pre-pandemic level across the forecast horizon. Information from timely wages indicators, liaison and business and household surveys are also incorporated into the near-term forecast (one to four quarters ahead).

As the WPI measures the changes in wage rates for a given quantity and quality of labour, it is a narrow measure of labour costs. The RBA also forecasts measures of labour costs that are wider in scope than the WPI and are more relevant for assessing living standards and inflationary pressures. This includes average earnings from the National Accounts (AENA), which is designed to measure the average earnings per hour and incorporates non-wage costs, such as superannuation and redundancy payments, along with pay increases resulting from changes in the composition of the workforce. Forecasts for AENA and labour productivity are then used to create a profile for unit labour costs (ULCs), which measures the labour costs per unit of output produced. This is the most relevant concept for assessing inflationary pressures from labour costs and feeds into the RBA’s mark-up model used to forecast trimmed mean inflation (see Cassidy et al 2019).

A key judgement in the RBA’s forecasts is the degree to which wages ‘catch up’ to the substantial increase in consumer prices since 2021. Real wages, as measured by the WPI, have declined by around 5 per cent since 2021 and remain around their 2023 trough. In the August 2024 Statement on Monetary Policy, it was assessed that the level of real WPI will only modestly pick up over the forecast horizon, with the pace of nominal wages growth declining more slowly than inflation (Graph 8). However, real AENA is above its pre-pandemic level and is expected to increase at a faster rate than real WPI over the forecast horizon. This forecast for stronger growth is mostly due to the legislated superannuation guarantee increases. The August Statement on Monetary Policy includes RBA’s latest forecast for wages growth and broader labour costs (RBA 2024).^[25]

Conclusion

This disaggregated framework for forecasting wages growth is useful in informing the RBA’s outlook for wages growth and assessing the effects of the Fair Work Commission’s award wage decisions. The RBA will continue to use and improve this framework alongside its other methods for forecasting wages growth.

Endnotes

Martin McCarthy and Lydia Wang contributed while working in Economic Analysis Department, Iain Ross is a member of the Reserve Bank Board and Madison Terrell is from Economic Analysis Department. [*]

Bishop and Greenland (2021) estimate the strength of this negative relationship using labour markets data on 291 regions. [1]

Ballantyne, Sharma and Taylor (2024) describe how the RBA assesses full employment, including the role of the NAIRU. [2]

This includes the WAD database, which covers almost 40 per cent of all public sector employees in the Victorian, Commonwealth, Australian Capital Territory, Northern Territory and Tasmanian governments. Information on agreements for other state governments is retrieved from state industrial relations commissions (or equivalent). [3]

See Fair Work Act 2009 (Cth) s 294(2) (Fair Work Act). [4]

Fair Work Act s 287. [5]

Annual Wage Review 2019–20 [2020] FWCFB 3500 at [7]. [6]

AWR 2019–20, n 6 at [9]. [7]

The Commission is also required to take into account considerations specified in the Fair Work Act, including the ‘minimum wages objective’ (s 284(1)), the ‘modern awards objective’ (s 134(1)), and the Act’s object (s 3). [8]

Fair Work Act ss 134(1)(h), 284(1)(a). [9]

Annual Wage Review 2018 –19 [2019] FWCFB 3500 at [12]. [10]

Annual Wage Review 2012 –13 [2013] FWCFB 4000 at [167]–[169]; Annual Wage Review 2022 –23 [2023] FWCFB 3500 at [87]. [11]

Fair Work Act ss 134(1)(a), 284(1)(c). [12]

Annual Wage Review 2021 –22 [2022] FWCFB 3500 at [71]. [13]

Fair Work Act ss 134(1)(ab), 284(1)(aa). [14]

AWR 2022 –23, n 11 at [11]. [15]

See Fair Work Act s 134(1)(aa). [16]

Fair Work Act s 134(1)(b); AWR 2022 –23, n 11 at [148]. [17]

Annual Wage Review 2014-15 [2015] FWCFC 2500 at [461]. [18]

AWR 2022 –23, n 11 at [7]. [19]

Annual Wage Review 2010–21 [2021] FWCFB 3500 at [170]. [20]

AWR 2022 –23, n 11 at [179]; Annual Wage Review 2023 –24 [2024] FWCFB 3500 at [156]–[157]. [21]

AWR 2022 –23, n 11 at [179]; AWR 2023 –24, n 21 at [154]. [22]

In the WPI release, the ABS uses the jobs in its sample to estimate wages growth for each ‘elementary aggregate’, which is a combination of state/territory, industry and sector (public/private). The ABS then weights each elementary aggregate using the shares of the wages bill in the EEH survey. This method does not explicitly weight each pay-setting method. However, the implicit weight of each pay-setting method in the aggregate WPI can be deduced. These implicit weights are about right for EBAs, a bit low for awards, and a bit high for individual arrangements. To complement the WPI release, the ABS publishes an analytical series that shows contributions to aggregate wages growth by pay-setting method, where each method is weighted according to EEH data (see ABS 2024b). This analytical series typically shows similar wages growth to the published WPI, except in September quarter 2023 where the AWR decision caused much higher wages growth for workers on awards. [23]

See the comments from Treasury in Senate Economics Legislation Committee (2023). [24]

This excludes the June quarter 2024 WPI and National Accounts releases, which were published after the August Statement on Monetary Policy. [25]

References

ABS (Australian Bureau of Statistics) (2024a), ‘Method Used to Calculate Analytical Series "Contribution to Wages Growth by Method of Setting Pay"’, Wage Price Index, Australia Methodology, June.

ABS (2024b), Employee Earnings and Hours, Australia, January.

Ballantyne B, A Sharma and T Taylor (2024), ‘Assessing Full Employment in Australia’, RBA Bulletin, April.

Bishop J and N Cassidy (2019), ‘Wages Growth by Pay-setting Method’, RBA Bulletin, June.

Bishop J and E Greenland (2021), ‘Is the Phillips Curve Still a Curve? Evidence from the Regions’, RBA Research Discussion Paper No 2021-09.

Bruno A, J Hambur and L Wang (2024), ‘Measuring Labour Quality in (Closer to) Real Time Using Emerging Microdata Sources’, Joint ABS–RBA Conference on Human Capital, Sydney, 11–12 June.

Cassidy N, E Rankin, M Read and C Seibold (2019), ‘Explaining Low Inflation Using Models’, RBA Bulletin, June.

DEWR (Department of Employment and Workplace Relations) (2024), Workplace Agreement Database, June.

FWC (Fair Work Commission) (2023), Enterprise Agreements Benchbook, August.

RBA (Reserve Bank of Australia) (2024), Statement on Monetary Policy, August.

Senate Economics Legislation Committee (2023), Estimates, Canberra, 30 May.