University job losses in the first year of COVID-19

Long overdue data on university staff in 2021 was released yesterday, giving us the most detailed information yet about job losses since COVID-19 hit the higher education sector.

Aggregate losses

The Department of Education’s staff statistics are mostly based on a 31 March census date. For staff with permanent or fixed term contracts I assume few job losses before 31 March 2020. The full travel ban on incoming international students was less than two weeks old, although some countries – including, importantly, China – had earlier travel restrictions. But retrenchments take time to process so I doubt the impact at 31 March exceeded some new hires abandoned at the last minute. These weren’t enough to prevent a 3.7 per cent headcount increase between 2019 and 2020.

In the next twelve months to 31 March 2021 total permanent or fixed term contract staff fell by 9,050, or 6.9 per cent of the 31 March 2020 total. This is only the third decline in staff numbers since 1989, and by far the largest. Difficult as 2020 was for everyone involved, total staff numbers at 31 March 2021 (121,364) were roughly what they had been on 31 March 2018 (121,718). The higher education sector is still a big employer by its own recent historical standards.

The full-time equivalent fall for permanent or fixed term contract staff was 7,985, or 6.8 per cent. This number, however, needs a caveat. For these staff the FTE is an extrapolation based on work arrangements as at 31 March. Normally this would understate actual FTE, as hours worked by additional staff hired after 31 March will not be counted until the following year. But in 2020 the 31 March estimate would have overstated FTE, by not taking into account net staff reductions during the rest of the year.

For casuals DESE reports actual FTE (ie not an extrapolation) with a lag, so that 2021 actuals will be reported in the next staff data release. Before then DESE publishes university estimates of casual FTEs. With no notice periods or retrenchment payouts required, universities could start reducing casual numbers before 31 March 2020. As the chart below shows, casual estimates were trending down at 31 March 2020 compared to 2019 – a contrast to the increase in permanent and fixed term staff.

The actuals show an even bigger decline, with losses of 4,258 FTE or 17.5 per cent in 2020 compared to 2019 .

The estimated casual FTE figures for 2021 are essentially the same as the actuals for 2020. It’s worth noting, however, that in the two years prior to COVID as well as 2020 estimated casual FTEs exceeded the actuals reported a year later.

Who lost their job? – academics versus non-academics

Nearly three-quarters of all university permanent or fixed term contract job losses on a headcount basis were to the non-academic workforce. Normally non-academics make up around 57 per cent of university permanent or fixed term staff.

It was a similar story with casuals in 2020 compared to 2019. Non-academic staff are normally around 35 per cent of casual FTE, but experienced 47 per cent of lost hours.

Academics who lost their job – by type of academic appointment

One important policy trend of the last 30 years has been separating funding for teaching and research, so that they are distributed on different criteria (mainly some version of enrolments for students and some version of performance for research) and grow at different rates (teaching faster in most years). Job-ready Graduates was the last step in this process, funding student places on a teaching and scholarship basis only, removing some implied research funding, especially in science, health and engineering.

With weak funding support for the 40/40/20 teaching-research employment model it has been under strain for a long time. The logic of the funding system is that specialised funding should lead to specialised staff, and this is indeed what has happened as the chart below shows. The high fees paid by international students, however, to some extent reunited teaching and research funding, and helped sustain teaching-research academic jobs at higher rates than would otherwise have been possible.

But with international student revenues and profits in decline, the largest hit to academic employment was in teaching-research roles (-1,837) compared to teaching only (-321) and research only (-254; I reported this as -176 on Twitter yesterday, but that was from a time series that included non-academic research only staff).

A caveat on the research-only number is that from the federal Budget of October 2020 universities knew that they were getting an extra $1 billion in federal research funding for 2021. This was a once-off COVID-19 supplement, and so without it in 2022 research only staff may decline significantly.

Job losses by career stage

One fear as COVID-19 struck was that this would affect early career staff the most. They are often employed on fixed-term contracts which make them relatively cheap to retrench. Contractual vulnerability was a factor in who lost jobs, with a 10 per cent decline in the number of staff on fixed term contracts compared to 5 per cent for those with permanent contracts.

Age data shows that relative to their 2020 numbers the 20-somethings were hit hard. It looks like some older staff were sent off into retirement, with mid-career staff at the lowest risk of job loss.

Among academic staff, level B (lecturer) academics had the highest job loss (-1,009) and percentage decline (-5.9 per cent) of all academic levels. This is consistent with the relatively high rates of job loss for younger staff.

Job loss by university

Job losses varied significantly by university. At the upper end of the range UNSW, Monash, RMIT and UTS all lost more than 500 staff. Charles Darwin and Southern Queensland went against the trend and added small numbers of staff.

How does these job losses compare with other figures?

The 9,000 headcount losses reported in DESE’s data, while large, are lower than other estimates. A year ago Universities Australia put 2020 job losses at 17,300. The ABS Participation, job search and mobility survey estimated tertiary education retrenchments at 16,800 for February 2020 to February 2021. While the much-cited 40,000 tertiary education job loss estimate released by the Australia Institute, with 35,000 from public universities, has methodological issues (I explain why below) getting good numbers on job losses is very difficult.

The biggest problem with DESE’s data is that casuals are reported on a full-time equivalent basis only. Other sources, although with their own methodological problems, put casual contract numbers at around 100,000 pre-COVID. Many casuals work at more than one university, so 100,000 casual contracts over-states the number of people employed as casuals. Assuming a strong relationship between casual FTE and the number of casual positions, a 17.5 per cent FTE loss of casual staff would translate into about 17,500 lost casual contracts, taking total job losses to the 26,000-27,000 range. The headcount number of persons who lost or missed out on work, however, would be lower than this estimate.

DESE’s headcount figures compare totals at two different dates, but don’t report the flows. In a normal year more people commence jobs in higher education than get retrenched, resign or retire. Clearly 2020 wasn’t a normal year, but there were still academic job advertisements during 2020. These new staff to some extent offset job losses, so that we can be confident that the number of people ceasing permanent or fixed term higher education employment exceeded 9,000 in 2020, without knowing by how many or the mix of voluntary and involuntary departures.

A complexity created by these flows is that more people hold a job in the higher education sector at some point in the year than are employed on a given day, the census date approach used by DESE and most other statistical collections. ATO payroll data by industry – unfortunately all tertiary education, including vocational education which is about 30 per cent of the total based on other sources – highlights the fluctuations.

The Australia Institute’s analysis was based on tertiary education industry employment data produced by the ABS, comparing May 2020 and May 2021. But the May 2020 baseline figure (in the chart below) is not credible. For these numbers to be right we have to believe that tertiary education employment was growing strongly between February and August 2020. The ATO numbers show that it was falling. DESE’s numbers don’t give us timelines, but casual FTE in decline by 31 March 2020 compared to 2019 suggest that jobs were lost in the February to August 2020 period. The well below usual number of academic job ads in this time also counts against growth in employment. (There is evidence that retrenchments of staff peaked later in the year).

The ABS numbers are from their labour force survey of around 50,000 people, with an eighth of the sample rotated each month. My theory is that one or more of these rotations into the survey over-sampled people working in tertiary education, distorting the results. As they were rotated out we start to get more plausible numbers for the November 2020 and February 2021 quarters.

Later in 2021

I can’t prove it, but some universities may have cut harder than in hindsight they needed to during 2020. Online education revenue turned out to be much stronger than expected. I’ve argued before that the timing of the government’s $1.5 billion of COVID assistance might have led to premature job loss. It was not announced until October 2020 and the money did not flow until 2021. Retrenchments that might have been avoided with an earlier announcement had possibly already been made. Job-ready Graduates introduced additional uncertainty, with some funding levels not clarified until 2021.

Once things settled at least some universities realised that they could start hiring again. While I have reservations about the ABS labour force numbers, they are consistent with the ATO trends and academic job ads in showing a jobs recovery in the second half of 2021.

Where to now?

The borders are now open to international students and their numbers are slowly trending up. That’s good news, but a lack of flights still limits arrivals. In 2022 only small amounts of government COVID assistance will be available compared to 2021. While most universities probably already have their staff numbers at a level they can sustain, the sector is not out of trouble yet.

2 thoughts on “University job losses in the first year of COVID-19

  1. Thanks Andrew, really interesting analysis. One aspect that, while difficult, would be great to look at is what approach different University’s took with the ‘extra $1 billion in federal research funding for 2021’. Some at least have held this centrally to support the overall bottom line – which makes sense in terms of institutions financial viability but may have done little to prevent job losses.


  2. The very broad conditions on the RSP and the fungibility of money would make it very hard to determine how the money is used. But even allowing for the general priority of research over teaching, I found the low decline in RO a bit surprising. Or maybe is just that many RO staff have externally funded grants or contracts that were not affected by COVID.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s