6302.0 - Average Weekly Earnings, Australia, Nov 2014 Quality Declaration 
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 26/02/2015   
   Page tools: Print Print Page Print all pages in this productPrint All

This document was added or updated on 02/07/2015.

A GUIDE TO UNDERSTANDING AVERAGE WEEKLY EARNINGS STATISTICS


INTRODUCTION

The Survey of Average Weekly Earnings (cat. no. 6302.0) is designed to produce estimates of the level of average weekly earnings of employees in Australia, at a point in time. It is one of a suite of ABS statistics providing information about earnings in Australia.

Average Weekly Earnings (AWE) statistics are referenced in a range of Commonwealth, state and territory legislation to adjust a variety of government payments, including the age pension. They are also used to analyse average earnings to inform wage claim submissions, monitor wage equity and develop taxation and social policies. In addition, AWE data are used in the compilation of the Australian National Accounts.

The aim of this article is to provide a guide for users of AWE by explaining:

  • How the data are collected.
  • What statistics are produced.
  • Factors to consider when using AWE data.


HOW AWE DATA ARE COLLECTED

AWE statistics are collected via a sample survey, which means data are collected from a selection of businesses instead of all businesses in Australia. The individual businesses included in the sample change over time.

The following data items are collected every six months from over 5,000 employing businesses:
  • Number of full time adult employees and number of other employees.
  • Taxable gross weekly earnings (including overtime but excluding salary sacrificed earnings) for full time adults and other employees.
  • Weekly overtime earnings, for full time adults.
  • Weekly salary sacrificed earnings, for full time adults and other employees.

For the purposes of AWE, employees refer to all wage and salary earners who received pay for any part of the reference period. Earnings are the pre-tax (gross) regular wages and salaries received by employees for work done or time worked (including paid leave) before any deductions (taxes, life insurance premiums, union dues, etc.) are made. Irregular and infrequent payments, such as annual bonuses, are excluded.

These sample data are weighted to provide estimates for the whole population of in scope businesses.

Data for AWE are collected at the business level, rather than the job or employee level. That is, the above data items are collected as aggregates, separated only into male and female employees.


WHAT AWE STATISTICS ARE PRODUCED?

The data items, listed above, obtained from businesses are used to calculate mean, or average, earnings for the following series:
  • Full time adult average weekly ordinary time earnings (excludes overtime).
  • Full time adult average weekly total earnings (includes overtime).
  • All employees average weekly total earnings (all employees including part-timers and juniors; and earnings for all hours worked, including overtime).

The main AWE statistical series are available both inclusive and exclusive of salary sacrificed earnings.

The following example shows how All employees average weekly total earnings (includes overtime) is calculated:

Image: How all employees average weekly total earnings (includes overtime) is calculated

Each of the above series are available for males, females, and persons.

Estimates are available by state/territory, industry division, and sector. State by sector estimates are available for persons only. Seasonally adjusted and trend estimates are also produced.

AWE statistics closely follow the International Labour Organization's concept of 'Statistics of average earnings'. The data are collected in respect of a typical week. For more information on what is included or excluded from AWE definitions of earnings, refer to the Explanatory Notes of Average Weekly Earnings, Australia, and Labour Statistics: Concepts, Sources and Methods (cat. no. 6102.0.55.001).

Estimates of the average level of earnings provide a level benchmark that can be used in comparisons. For example, individuals can compare their own earnings to the average level of earnings in their industry or state/territory.

It is important to note that AWE estimates do not relate to average award rates, nor to the earnings of the 'average person'.


FACTORS TO CONSIDER WHEN USING AWE DATA

AWE is not a survey of individual workers and how their earnings change over time. It looks at overall earnings in the economy relative to the number of employees. Therefore, changes in the composition of the labour market can have significant impacts on average earnings estimates that are unrelated to changes in individuals' rates of pay. The key factors to consider when using AWE data are highlighted in the following three examples.

1. Comparing average earning estimates between different groups of employees

The level of average earnings may differ between groups of employees across different states and territories, industries or sex due to a number of factors. Because of this, the information collected and produced from AWE does not enable the impact of a single factor on average wages across these groups, such as sex, to be easily isolated. This issue is explored further below.

Comparisons of average earnings: males and females
AWE statistics show a difference in average earnings between males and females. This is at least in part due to compositional factors that impact on earnings and are not adjusted for in the AWE series, such as:
  • employee characteristics such as age, experience and training,
  • hours worked,
  • employment conditions, and
  • occupation.

For example, data from Employee Earnings and Hours, Australia (EEH; cat. no. 6306.0) reported that in May 2014, male employees were predominately full-time (76.6% of male employees). In contrast, more female employees were employed part-time (56.2%) than full-time (43.7%). As such, it would be expected that, on average, men would be paid for a higher number of hours of work than women. If, for comparison purposes, the population is restricted to only include full-time non-managerial employees paid at the adult rate, differences still remain in the number of hours paid for men compared to women. EEH data reported that the average weekly total hours paid for full-time non-managerial employees paid at the adult rate was 40.7 hours for males and 38.3 hours for females.

The Labour Force Survey provides a wide range of data relevant for examining the composition of the workforce. For example, data from Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003) show large differences in the proportion of males and females employed by industry. In May 2014, most employees (85.9%) in the highly-paid Mining industry were male. In the lower paid industries of Accommodation and food services and Retail trade, the majority of employees were female (both 56.4%).

Because of these compositional factors, AWE statistics cannot answer whether males and females receive 'equal pay for equal work'. Also as it does not collect the relevant information, AWE is not suitable for determining the causes of differences in average earnings between males and females.

Other ABS sources of earnings statistics are better placed to address some of these types of questions. For example, Employee Earnings and Hours, Australia and Employee Earnings, Benefits and Trade Union membership, Australia (cat. no. 6310.0) collect and publish information on characteristics of employees. Gender Indicators (cat. no. 4125.0) may also be of use for researchers investigating gender differences. For more information on other measures of earnings, see Understanding earnings in Australia using ABS statistics.

Comparison of average earnings: states and territories
As with comparing average earnings between males and females, the composition of the labour force in different states and territories will influence the estimates. For example, data from November 2014 indicate that the Average weekly ordinary time earnings (AWOTE, original) for full time adults is higher for employees in Western Australia (WA; $1,673.10) than for those in New South Wales (NSW; $1,492.30).

One factor likely contributing to the observed differences in the level of average weekly earnings between WA and NSW is the impact of the recent resources construction boom in WA. The resources boom increased the proportion of WA employees working in Mining (which is the industry with the highest average weekly ordinary time earnings for full-time adults), contributing to an increase in the average weekly earnings for WA. Also skills shortages and competition for labour have previously led to higher earnings in WA, particularly in skilled occupations in industries such as Construction and Mining.

As with comparing males and females, differences in compositional factors such as hours worked, occupations and labour market conditions may affect average earnings between states.

The higher level of average weekly earnings in WA does not suggest that, all other things being equal, an employee in NSW earns less for performing the same job.

2. Sampling error

Sampling error is the difference between the estimates produced by taking a sample of employing businesses, and the values that would have been produced if the information had been obtained from a census of all employing businesses.

The sampling error associated with any estimate can be estimated from the sample results. One measure of sampling error is given by the standard error. The standard error indicates the degree to which an estimate may vary from the value which would have been obtained from a census of all businesses (the 'true value'). There are about two chances in three that a sample estimate differs from the true value by less than one standard error, and about 19 chances in 20 that the difference will be less than two standard errors.

To enable interpretation of AWE estimates, the ABS publishes standard errors for original series data. These are available in the Time Series Spreadsheets under the Downloads tab of the AWE release. Any analysis of AWE data should be undertaken with the standard errors in mind. For more general information on sampling error, refer to the Statistical Language - Measures of Error section of the ABS website.

An example of the use of the standard error is as follows. If the estimated Full time adult average weekly total earnings (Australia; Original; Persons), November 2014 were $1,542.40, with a standard error of $8.30, then:
  • there are about two chances in three (67%) that the true value is between $1,534.10 and $1,550.70 ($1,542.40 ± $8.30);
  • there are about 19 chances in 20 (95%) that the true value is between $1,525.80 and $1,559.00 ($1,542.40 ± 2 x $8.30).

3. Movements in AWE estimates

With the exception of annual movements at the broadest level (Australia and sector), the ABS does not currently release estimates of movements for AWE for the reasons described below. However, many users of AWE data calculate estimates of movements at the state or industry level from the published average earnings estimates. While, over the longer term, AWE gives a reliable measure of changes in average earnings, caution needs to be exercised in interpreting AWE movements from period to period.

First, AWE estimates can be affected by sampling (ie, changes in which businesses are selected in the AWE sample over time). Short term changes in the estimates of average earnings in many cases are quite small compared to the size of the standard errors of those estimates. In these cases, the changes in earnings may be due to sampling variability and may not be statistically significant.

Second, a frequent misconception is that changes in average earnings estimates largely reflect changes in individuals' salaries or hourly rates of pay. While changes in wage rates do impact AWE estimates, they cannot be isolated from other factors and often may not be the primary driver of changes in estimates. Other factors impacting AWE estimates from one time period to the next include real world events - such as changes in hours worked, upskilling of jobs over time in the labour force, or changes in occupations.

Users interested in changes in wage rates over time, independent of changes in quality and quantity of labour, should refer to the Wage Price Index, Australia (cat. no. 6345.0). For more information on the differences between AWE and the Wage Price Index, refer to the feature article Average Weekly Earnings and Wage Price Index - What do they measure? published in the May 2014 issue of AWE.


CONCLUSION

AWE is one of a suite of ABS statistics that can inform discussions of earnings in Australia. The survey is designed to provide an estimate of average wages and salaries at a point in time. The survey is not designed to represent the earnings of the 'average person'. Changes in earnings over time may be caused by a large variety of factors and may not reflect changes in pay experienced by a typical individual employee. The survey is not designed to accurately measure short-term movements in wages growth, or for investigating the causes of differences in earnings between groups. Other ABS releases, such as those listed below, provide statistics on earnings but differ in a number of ways from AWE. The decision on which is the most appropriate source of earnings data should be determined by the purpose and type of analysis to be undertaken.


FURTHER INFORMATION

Employee Earnings and Hours, Australia (cat. no. 6306.0) - Collects detailed information on employee earnings and characteristics of individuals from businesses. In addition to the mean, the median, quartiles and deciles can be calculated, giving an indication of the distribution of earnings across the population. Earnings data can be explored by characteristics of employees, such as age, occupation or the number of hours paid for.

Employee Earnings, Benefits and Trade Union membership, Australia (cat. no. 6310.0) - Collects detailed information on employee earnings and characteristics of individuals from households. As for Employee Earnings and Hours, the median, quartiles and deciles of earnings are available, giving an indication of the distribution of earnings across the population. Earnings data can be explored by characteristics of employees, such as age, occupation or country of birth.

Wage Price Index, Australia (cat. no. 6345.0) - Measures changes in the price of wages and salaries in the Australian labour market, independent of changes in the composition of the labour force, hours worked and employee characteristics.

Average Weekly Earnings and Wage Price Index - What do they measure? - A feature article published in the May 2014 issue of AWE. It outlines the purpose and key uses of AWE and the Wage Price Index, and explains how the two surveys can respond differently to economic events using hypothetical examples.

Understanding earnings in Australia using ABS statistics - A feature article providing an overview of measures of earnings produced by the ABS, their key outputs, benefits and limitations.