Statement on Monetary Policy – May 2016 Box B: Wage Developments by Industry
As discussed in the ‘Price and Wage Developments’ chapter, wage growth in Australia has been very low, and lower than implied by its historical relationship with the unemployment rate. Wage growth is well below its decade average in all industries, and dispersion across industries is around its lowest level since the late 1990s when the wage price index (WPI) began (Graph B1 and Graph B2).
While wage growth is low in every industry, it is currently lowest in industries that are more exposed to the end of the mining investment boom, such as mining, construction and administrative & support services (which include labour hire companies that provide a range of workers – such as construction labourers, truck drivers and administrative assistants – to mining and mining-related firms). In addition, wage growth has been relatively weak in professional, scientific & technical services, and rental, hiring & real estate services, which also include firms that support the mining industry.
While these industries have experienced wage growth below the national average of late, this follows a period of above-average wage growth for many of them during the mining investment boom. The overall level of earnings in the mining, construction and professional, scientific & technical service industries (based on average weekly earnings data) appears to have risen somewhat relative to the national average since the mid 2000s (Graph B3).[1] In contrast, the level of earnings in the rental, hiring & real estate industry has fallen relative to the national average. Wage relativities have been more stable for most other industries.
Most of the industries that have experienced relatively low wage growth recently also experience more volatile wage growth (from quarter to quarter) than other industries over time (Graph B4 and Graph B5). It is unclear to what extent this volatility reflects: the nature of wage setting in these industries; the possibility that these industries are more cyclically sensitive than others; and/or the possibility that they have just been affected by the largest changes in labour demand and supply associated with the rise and fall of mining investment.
Wage growth has generally been higher in industries where employment growth has been stronger of late, with a few notable exceptions (Graph B6). For example, while wage growth has been weak in administration & support services, employment in the industry rose quite strongly in 2015. Liaison suggests that the wages in labour hire companies, which fall into this category, have declined as there has been a shift in demand for their workers from mining-related businesses that paid relatively high wages to other firms that pay lower wages. In addition, labour hire firms may have also experienced an increase in the availability of labour, as similar workers have left mining and mining-related firms. Professional, scientific & technical services firms have also had weak wage growth and strong employment growth in the past year. This may also reflect a change in the composition of the clients they work for and an increase in labour availability as similar workers have left the mining industry. In contrast, manufacturing wage growth remains high relative to other industries at the same time as employment in manufacturing has declined.
Growth in average earnings per hour from the national accounts, a broader measure of labour income, tends to be higher in industries that have stronger growth in labour productivity over the long run.[2] This would be expected if wages remain a constant share of an industry's total income. Indeed, the wage share of total factor income has been relatively stable in the non-mining sector, despite below-average wage growth (Graph B7). However, there has been an increase in the wage share of income in the mining industry because the decline in commodity prices over recent years has weighed more heavily on profits than wages, just as the earlier increase in commodity prices accrued more to profits than to wages.
Wage-setting methods vary significantly across industries. Enterprise agreements between employers and groups of employees are used to set the pay and conditions for a little more than 40 per cent of employees. Individual agreements between employers and employees cover almost another 40 per cent of employees, while around 20 per cent of employees have their pay determined directly by awards. Most awards are determined by the Fair Work Commission and also indirectly affect a significant proportion of employees covered by enterprise agreements or individual contracts where they set minimum standards for an occupation or industry.
Industries that have the highest share of workers whose pay is directly determined by awards include accommodation & food services (43 per cent), administrative & support services (37 per cent) and retail trade (29 per cent) (Graph B8). Industries with a significant public sector presence, such as education & training, public administration & safety, and health care & social assistance, tend to have a large share of enterprise agreements. In most other industries, at least half of all employees have their pay and conditions determined by individual agreements.
The relationship between wage-setting methods and wage outcomes is unclear. Awards are more prevalent in industries with lower wages as they provide minimum standards. Changes in wage growth and labour market outcomes by industry may reflect differences in wage flexibility or bargaining power, but these are difficult to distinguish from a wide range of other determinants of wages, including variation in industry performance, the balance of demand and supply for different skills, and productivity.
Footnotes
The average weekly earnings series is designed to estimate the level of wages, rather than the change over time. It is affected by compositional change in employment unlike the WPI, which holds the quality and quantity of labour constant. See Australian Bureau of Statistics (2014), ‘Feature Article: Average Weekly Earnings and Wage Price Index – What Do They Measure?’, Average Weekly Earnings, May. [1]
Average earnings per hour from the national accounts is affected by compositional change and a broader range of labour income payments than the WPI, such as payments related to redundancies, allowances and fringe benefits. The relationship between earnings and productivity growth by industry is not always clear from one year to the next because productivity growth is volatile and productivity improvements are more difficult to measure for some industries than others. Productivity is more challenging to measure for services than for goods. It is also more difficult to measure productivity for services where public sector provision is a significant share of output than for services where prices are market determined. [2]