The public sector’s role in propping up growth and jobs

Strong growth in public demand has accounted for all of Australia’s weak GDP growth over the past year as activity in the rest of the economy has slowed sharply. Above-average growth in a broad, but volatile measure of public-sector employment has accounted for about two-thirds of strong total jobs growth over the past year, while private-sector hiring has slowed to a still-solid pace (private-sector employment is much weaker in other countries and has contracted in the UK and NZ). Health and social assistance – which is already Australia’s largest employer – continues to drive public-sector employment, with very crude estimates showing NDIS-related hiring adding to total employment from 2018 to 2022 when the scheme was ramping up. 

The national accounts show that growth in the economy is currently being propped up by public-sector demand, which accounts for all of the 1% increase in Australia's real GDP over the past year (note that public demand comprises public consumption and investment). 

The last time this happened was in the year before COVID hit and before that in the global financial crisis.

This unusual turn of events reflects the combination of ongoing strong growth in public demand and activity in the rest of the economy sharply tapering off after a very strong recovery from the short-lived COVID-driven recession of 2020.

Most of the strength in public demand reflects public consumption, which covers the government wages bill, as well as goods and services that are provided either free of charge or sold at below-market prices.

The latter includes direct government subsidies that are tied to private-sector spending on specific goods and services – such as “cost-of-living relief” payments for household electricity bills – where the ABS reallocates the subsidised expenditure from private- to public-sector demand.(1)

This standard accounting treatment means that private demand has likely not been quite as weak as recently reported and that growth in public demand has been flattered by Commonwealth and state governments both making more use of cost-of-living payments.(2) 

The fact that public demand has propped up economic growth raises the question of whether the same is true for the labour market, where employment has risen by a strong 2.7% over the past year.

Using the broadest definition of the public sector – which is the ABS “non-market sector” that comprises the public administration, health and social assistance, and education industries – public-sector employment currently stands at about 4½mn, or 31% of total employment, up from 26% a decade ago.(3)

This is the same as the equivalent share in the USA and places it in line with the median share of most other advanced economies, although there is a marked variation across countries. 

A few economies – viz, the euro area, Canada, and New Zealand – are at the bottom end of the range, clustered around 25%, while the Nordics are at 33% and over.   

On this basis, above-average growth of about 6% over the past year has seen this broad measure of the public sector account for 1.8pp of the 2.7% increase in total employment over the same period. 

With the broad private-sector measures of employment, defined as the “market sector” by the ABS, up 1.2% from a year ago, this means that the private sector has added 0.8pp to total growth.

Smoothing the data because the industry split of employment is volatile, even when reported as annual growth rates, shows that both public- and private-sector employment display mini-cycles, but public-sector employment hardly ever falls, with the private-sector accounting for job losses in recessions and also contracting during the global financial crisis.(4)  

All this is similar to the experience of other countries, where estimated broad public-sector employment is generally not correlated with the business cycle and currently accounts for most, and sometimes all, growth in total employment.

Correspondingly, growth in broad private-sector employment is much weaker than in Australia, broadly stalling for payroll jobs in the US in Q2 and recently contracting in the UK and New Zealand. 

In terms of what has driven the recent strength in broad public-sector employment in Australia, about half of all public-sector workers are employed in the health and social assistance sector, which is also the largest single employer in Australia, accounting for 16% of total employment (education is fifth at 8% and public administration is sixth at 7%).

Health and social assistance has grown in importance over the past sixty years, although the share of total employment has picked up at a faster rate from 2009 onwards.

The National Disability Insurance Scheme (NDIS), which took some years to ramp up after starting in 2013, likely accounted for much of this acceleration.

On our highly experimental time series estimates, which are based on scheme payments and separate industry data on wages and profits, most of the strength in health and social assistance employment from 2018 to 2022 was likely due to NDIS.

This suggests that over that period, NDIS-related employment potentially added about ½pp to average annual growth in total employment of about 2¼%, but that there has been only a minimal contribution over the past couple of years.

That said, the latest census suggests that these extremely imprecise estimates could overstate the earlier contribution to growth in total employment based on a very fine split of jobs that shows a likely NDIS impact, but also strong hiring in hospitals and aged care between 2016 and 2021.   

Finally, in terms of the outlook, the broad measures of public- and private-sector job vacancies both peaked at multi-decade highs as a share of the labour force a couple of years ago. 

Private-sector vacancies have since fallen faster and are approaching their pre-pandemic levels, while declining public-sector vacancies are still at a historically high level.     

Note:
(1) In contrast, welfare payments and indirect government subsidies, where the money can be used for any purpose, are separately reported as transfers of income and are not counted in GDP.
(2) The ABS will publish data on direct subsidies to households in 2023-24 later this month.
(3) Narrower measures of public-sector employment are available, but the broader measure was used to account for the blurring of some jobs between the public and private sectors (e.g., a healthcare professional might consult to both private and public hospitals) and the fact that the public sector significantly influences the pricing decisions in the health and education sectors via subsidies and regulation.
(4) The volatility of sectoral split of employment partly reflects sampling variability. For example, survey respondents are sometimes inconsistent when reporting which industries employ other members of their household from one survey to the next.