Analytical measures of labour costs



In the National accounts, remuneration for the labour provided by employees involved in economic production is described as Compensation of Employees (COE). COE includes salaries earned by employees, bonuses, in kind payments such as free or discounted goods, superannuation, workers compensation and termination payments.

COE is combined with other labour market indicators such as employment headcount and hours worked to produce a range of analytical statistics to help users understand changes in the labour market.

This article explains how to interpret average COE per employee, COE per hour worked and unit labour costs (ULC) during the COVID-19 pandemic and provides further insight into COE growth as one measure of the labour cost.

Average COE per employee

Average COE per employee is defined as the total value of COE divided by the total number of employees.

\(Average\ COE\ per\ employee = \large{COE \over Number\ of\ employees}\)

Average COE is published for the total economy and non-farm sector.

Despite volatile quarterly outcomes, the 10-year average of growth in average COE per employee is generally comparable to growth in the Wage Price Index (WPI), especially prior to the COVID-19 pandemic. This relationship has diverged in the past 2 years.

Table 1. Time period average quarterly growth
Average COE per employee0.98%0.66%0.70%


Note: time periods for average quarterly growth correspond to 2000-2009, 2010-2019 and 2020-current

Growth in the WPI measures changes in the price of labour, unaffected by compositional shifts in the labour force, hours worked or employee characteristics.

In contrast, compositional changes, such as shifts in the proportion of higher paid or full-time jobs or in hours worked by employees, have a large impact on average COE per employee.

Pandemic impacts on average COE per employee

Job losses through COVID-19 lockdowns were more pronounced in part time employment and concentrated in industries that on average pay lower wages, for example Accommodation and Food Services. As a result, average COE per employee increased during this period with lower paid employees dropping out of the work force.

Due to COVID-19 restrictions on businesses, some employees worked reduced or zero hours. Employees qualifying for JobKeeper continued to be classified as employed as they retained a formal job attachment and received a wage or salary. See more here: Classifying people in the Labour Force Survey during the COVID-19 period

In the first iteration of JobKeeper that ran from March to September 2020, eligible employees were paid a minimum of $1,500 per fortnight regardless of the hours they worked. For employees earning less than the minimum JobKeeper amount in the March quarter 2020, the extra payment resulted in an increase to average COE per employee in June quarter 2020. 

Average COE per employee increased 3.7% in the June quarter 2020 due to these impacts. This increase was also observed in estimates of average weekly earnings over 2020, as published in Average Weekly Earnings, Australia, May 2020.

In recent quarters, strong demand for labour drove the unemployment rate to levels not seen since 1974. This has also been reflected in a pick up in WPI growth, which rebounded from historic lows to above pre-COVID-19 rates. Strong growth in average COE per employee has also been seen over this time period.

COE per hour

COE per hour is defined as total COE divided by the hours worked by employees.

\(COE \ per \ hour = \large{COE \over Hours \ worked \ by \ employees}\)

Estimates of hours worked used in the calculation of COE per hour are based on the ABS Labour Accounts, supplemented with data from the ABS Labour Force Survey.

COE per hour provided additional insights into changes in COE over the COVID-19 period. Hours worked in the economy were reduced by COVID-19 restrictions but some employment was artificially maintained due to the requirements of the JobKeeper and other business support programs such as the NSW Government’s JobSaver payment.

The ABS will begin publishing COE per hour worked in the September quarter 2022 publication to provide users additional insights into COE movements.

Pandemic impacts on COE per hour

The number of hours worked showed unprecedented volatility due to the economic and health impacts of the pandemic. This was predominately driven by a reduction in demand for services due to movement restrictions which reduced employees’ hours worked. Increased COVID-19 sick leave, particularly in the March quarter 2022 during the peak of the Omicron wave, also reduced the number of hours worked.

Like average COE per employee, COE per hour is affected by compositional change in the labour market. Job losses and restrictions in customer facing industries resulted in part-time employees’ hours dropping significantly. Working from home arrangements enabled other industries to be more resilient to COVID-19 restrictions. Employees who were able to work from home were typically in higher paid industries. Some essential workers were required to work overtime to meet demand. All these factors contributed to higher COE per hour.

JobKeeper payments also contributed to increased COE per hour worked. Eligible employees that had reduced or zero hours still received $1,500 per fortnight during the June and September quarters 2020. Regardless of their pre-COVID-19 wage, employees who earned $1,500 per fortnight for reduced hours contributed to the strong COE per hour during this period.

Comparatively, the decline in hours worked due to the Global Financial Crisis was prominent in higher paid industries such as Financial and Insurance Services, resulting in a fall in COE. By contrast, COVID-19 had a disproportionate impact on lower paid industries and thus increased COE relative to hours worked.

In the past two years, volatile hours worked drove large quarterly movements in COE per hour. The rolling average for growth in hours worked by employees has stabilised near pre-COVID-19 levels, while growth in COE has continued to increase.

Unit labour costs

Nominal unit labour costs (ULC) are defined as average labour costs (ALC) divided by average labour productivity (ALP).

\(Nominal\ ULC = \large{Average \ labour \ costs \ (ALC) \over Average \ labour \ productivity \ (ALP)}\)

Average labour costs (ALC) are defined as labour costs divided by hours worked by employees.

\(ALC = \large{Labour \ costs \ (LC) \over Hours \ worked \ by \ employees}\)

Average labour productivity (ALP) is defined as volume GDP divided by total hours worked by employees and self-employed.

\(ALP = \large{Volume \ GDP \over Total \ hours \ worked \ by \ employees \ and \ self \ employed}\)

ULC measures the average cost of labour per unit of output produced in the economy. This highlights the main difference between ULC and other average labour cost indicators (COE per hour or average COE per employee), as ULC allows for the effect of productivity improvements on labour costs. ULC remains unchanged if an increase in the average cost of labour is accompanied by the same proportionate increase in average labour productivity.

ULC reflects the cost of labour from a firm’s perspective. The labour cost component of ULC is calculated as COE plus payroll taxes minus employment subsidies. This provides additional insight into how government support programs, such as JobKeeper, influenced change in ULC during the COVID-19 pandemic.

Nominal ULC includes general price changes in the economy, due to the use of nominal labour costs.

Real ULC (RULC) removes price impacts by deflating average labour cost with the GDP deflator. RULC offer a more direct indication of labour cost pressures from a firm’s perspective by accounting for changes in labour costs relative to the prices received for goods and services produced (GDP deflator).  

Real unit labour costs (RULC) are defined as:

\(RULC = \large{{{ALC }/{GDP \ deflator}} \over ALP}\)

Pandemic impacts on ULC

Record employment subsides through JobKeeper, combined with reduced employment and the introduction of payroll tax relief, caused labour costs for firms to decline to historic lows by June 2020. This outweighed the significant reduction in hours worked from the nationwide COVID-19 lockdown and drove dramatic falls in both nominal and real ULC during this period.

Labour costs quickly rebounded as economic activity recovered and eligibility requirements and rates of payment for JobKeeper tightened, leading into the ultimate cessation of the support program on 28 March 2021. Elevated growth in labour costs in recent quarters reflected the increasingly tight labour market, with high job vacancies and ongoing labour shortages. This resulted in higher non-wage payments such as retention and sign-on bonuses contributing to labour costs. Quarterly growth in labour costs per hour averaged 0.9% since the beginning of the pandemic, compared to the ten-year pre-pandemic average of 0.7%.

Average labour productivity has been volatile over the course of the pandemic, reflecting disruptions to hours worked through the various COVID-19 strains and associated compositional change between industries. Labour productivity increased 3.5% since December 2019, translating to an average quarterly growth of 0.35%. This is similar to the 0.30% average quarterly growth during the most recent productivity growth cycle (2009-10 to 2017-18).

Nominal ULC rose 5.1% since the beginning of the pandemic, driven by an increase in average labour costs and partly offset by a smaller rise in labour productivity.

In contrast, real ULC fell 8.3% over this period. The difference between nominal and real ULC during the COVID period was due to a 14.6% rise in the GDP implicit price deflator. The largest driver of the GDP deflator increase was the terms of trade, with the prices of exports received by Australian firms increasing 54.9%.

The decline in real ULC reflects labour costs rising less than prices received by Australian firms. Falling real ULC need not imply falling real wages from an employee perspective, especially if increases in prices were largely concentrated in exports which are not purchased by domestic consumers.

Back to top of the page