In a previous blog post, I argued that stagnating or declining government revenues encourage universities to seek additional international student fee income. By 2018, international student fees provided 26 per cent of all university revenue, up from 10 per cent in 2000.
However, I doubted that aggregate public funding levels fully explained university dependence on international students, whose numbers grow when public spending is increasing as well as decreasing.
But in thinking about how government policy affects university decision making it is not just revenue that matters. The cost of the services universities deliver for their public money is also crucial to understanding university behaviour.
A recent article in The Conversation suggested that government student-linked revenue did not cover the full cost of growth in student numbers. Another Conversation piece this morning also suggested that universities have become reliant on international student fee revenue to cover the cost of teaching, as well as research and other activities.
However, a chart in my first post shows that since the mid-2000s average per student funding for Commonwealth supported students grew by more than inflation and then stabilised in real terms, although with a small recent decline.
But one point made in response to my original post was that wages usually grow by more than general inflation. This means that my CPI indexation of revenue does not fully adjust for the changing purchasing capacity of grants, given the bundle of goods and services universities actually buy. In 2018, 56 per cent of university expenditure was on wages.
Although the wage point is correct, what it means is that academics at a given level (lecturer, senior lecturer, etc) get paid more in real terms now than they did in the past. But teaching costs have been held down by employing more junior academics to do the teaching work.
The chart below assumes that academics employed on a teaching and research basis spend 40 per cent of their time teaching. While that is a common expectation, I am not aware of recent published research on actual time use, and so the detail of the trends are sensitive to varying that assumption. Teaching and research staff have higher ranks than teaching-only staff, the vast majority of whom are casual staff employed at ‘Level A – Below lecturer’, usually a tutor.
Interestingly, the proportion of teaching full-time equivalents at the top ranks has increased slightly, by 2 percentage points since 2001. But there has been a 6 percentage point increase in Level A full-time equivalents at the expense of lecturers and senior lecturers. This is likely to have held teaching wage costs down.
Another way of saving on teaching costs is to get each teaching staff member to teach more students. Student-staff ratios have always been a bit hard to interpret – they count teaching and research staff as one, which overstates hours available for teaching; in what sense is a student with 12-15 weekly contact hours for six months of the year full-time from a teaching perspective? But with these interpretive qualifications, student-staff ratios have also increased, from 14 or 15 to one in the early 1990s to about 21 to one now.
At a field of education by university level, the recent Deloitte Access Economics study of university costs found, unsurprisingly, that there are economies of scale. The universities that have more students enrolled in a field report lower average costs. Counting both domestic and international students, full-time equivalent enrolments increased by nearly 90 per cent in Australian universities between 2001 and 2018, generating additional economies of scale.
In one of the more surprising statistics I have seen this year, more than half the floor space in Australian universities is in buildings that are 40 years old or more. Given that on-campus student numbers have more than tripled over that time period, and staff numbers have more than doubled, it suggests that universities use their physical infrastructure much more efficiently than in the past.
Consistent with this analysis, university cost data shows that although a few disciplines are loss-making at Commonwealth supported rates, most are not. However, in each field there are at least some universities with costs above the Commonwealth supported student funding rate, as the chart below shows. Most universities are in this situation for management and commerce courses, which is also a very popular field for international students.
Explaining management and commerce has a chicken and egg aspect; do these faculties have lots of international students because the Commonwealth funding rate is too low, or are their average costs high because it costs more to recruit, educate and support so many international students? Probably a bit of both, but management and commerce median costs are fairly similar to fields with similar teaching methods such as psychology and ‘society and culture other’, which includes much of the humanities and social sciences. Overall, I am inclined to think under-funding of Commonwealth supported management and commerce students has contributed to the need for international student fees.
Losses by particular universities in other fields may also have contributed to their decision to pursue international student discretionary revenue. However, the funding rates have their origins in a system to calculate overall block grants. It was always expected that there would some mismatches at the discipline level, but block grants would work so long as profits and losses balanced across the institution.
Despite the exceptions, overall Commonwealth supported students are profitable. That these students are not typically loss-making is also the only conclusion compatible with a massive increase in domestic bachelor degree student numbers under demand driven funding while university surpluses were high by recent historical standards (p.59). Universities don’t generally need international student profits to deliver Commonwealth supported places.
Again, I think we are left with only part of the story as to why international student revenue has grown so much. The next post looks at how changes to research grant policy have affected university finances.
2 thoughts on “Why did universities become reliant on international students? Part 2: The cost of educating Commonwealth supported students”
Do the FTE numbers you base this analysis on take into account the extent of university teaching delivered by casual academic staff? It’s my understanding that full-time Teaching-only appointments at Level A are a small proportion of appointments, and a partial response at best to increasing casualisation and the lack of permanent positions on offer. The real secret to keeping resourcing costs down in relation to teaching is to allow higher and higher numbers of sessional academics to deliver your courses at the frontline. No job security, paid for 26 weeks of the year maximum, furloughed during each semester break and over the summer. Also if you think that 12-15 face to face hours doesn’t constitute a full time teaching load, multiply it by four to account for all the painstaking preparation that precedes delivery, and the thankless marking that follows and takes many, many more hours than is budgeted for, then see how that fits into a standard 40:40:20 model for those staff lucky enough to have full-time research and teaching position.
Yes, casual Level A vastly outnumber those with on-going or fixed term contracts (https://twitter.com/andrewjnorton/status/1266901895211216896/photo/1). The chart in the post includes casual FTE as reported to the Department, which is paid hours rather than actual hours worked as you note. While I agree that there are significant numbers of non-classroom hours involved in teaching, it is not like the school student:staff ratios, where staff have to teach or supervise students all the time, and the academic year is longer.