My recent blog post on the cost of the teaching-research academic employment model prompted various Twitter comments on its analytical assumptions.

Friday’s post and the comments made on them also link back to other recent posts that try to understand how universities finance themselves. The posts have consistently acknowledged data issues, and that precise dollar figures cannot be attached to most of the conclusions. At best, we can get to a credible range.

The most important criticism is that my analysis assumes that the two main sources of university expenditure data, the Deloitte Access Economics study for teaching, and the ABS for research, can adequately distinguish between and separately cost scholarship and research.

According to the ABS, which uses international definitions, research is ‘creative and systematic work undertaken in order to increase the stock of knowledge – including knowledge of humankind, culture and society – and to devise new applications of available knowledge’. Scholarship, by contrast, I take as activity leading to or maintaining in-depth understanding of existing knowledge.

It is hard to have research without scholarship. How can academics claim to have increased knowledge if they are unaware of the current state of their topic or field? The literature reviews that appear in many ‘research’ articles in academic journals are scholarship.

The ‘devise new applications of available knowledge’ part of ‘research’ creates even more subtle distinctions. It will often be difficult to say where scholarship ends and research starts.

Similarly, it is hard to teach without scholarship in the subject being taught (though I have heard plenty of horror stories of people teaching subjects outside their existing expertise).

Scholarship is the foundation of both teaching and research.

An implication of this is that the teaching and research employment model may not be in as much trouble as I said, because some of the time I assumed was unfunded ‘research’ is ‘scholarship’ already included in Deloitte’s cost model.

A further issue is that, for lack of a better data source, I fell back on the assumption that teaching-research academics spend 40 per cent of their time on research. But if teaching-research academics spend less of their time on research than 40 per cent there is, correspondingly, less of a research funding deficit to make up. More of their salary costs are already covered in the Deloitte analysis of teaching and scholarship expenditure.

There is also the problem that, despite their classification, ‘research only’ staff do sometimes teach. If that teaching work is captured in the Deloitte teaching cost statistics then again we have the danger of double counting, if the ABS does not do the same thing.

In the Twitter thread copied in below, cost expert Michelle Brooke confirms that only some universities have scholarship separately identified in their cost model, along with further complexities about research students and administration.

Neither the Deloitte nor ABS reports directly say how they divide up academic time, although Deloitte highlights both the sometimes subtle conceptual distinctions between different activities and the practical difficulties in measuring that time.

However, it looks like, in recent years at least, the ABS numbers assume 40 per cent of teaching-research staff is spent on research. In the chart below I used the Department of Education’s staff statistics and allocated 100 per cent of research only staff time and 40 per cent of teaching and research staff time to research. That gets a very close match with the ABS academic ‘person years of effort’ on research numbers.

The chart above also raises questions about why the ABS apparently increased their research time estimate from 2012, and whether some growth in research expenditure is time reclassified, for reasons that may not be sound if we want a strict scholarship-research distinction. Some double counting of scholarship time in the Deloitte and ABS studies seems quite possible.

Despite the seemingly on-the-high-side ABS estimates of pure research activity, if we combine the different sources (chart below) the results don’t immediately suggest a significant double count. ABS estimated current expenditure on research is equivalent to 34.5 per cent of expenditure in the Finance report. Deloitte says 52 per cent of university expenditure was on teaching, scholarship and student support.

If there was a major double count of scholarship in both teaching and scholarship (Deloitte) and scholarship as the foundation of research (ABS) the total of the two would leave little or nothing for anything else. Keep in mind that the way these cost models work we are not just talking about academic salary costs (29 per cent of all university in expenditure in 2018), but also about how the costs of multi-purpose overheads are distributed between the different activities. Instead, we have 13-14 per cent of total expenditure left for other university activities, such as engagement and commercial operations.

The lack of robust evidence on how academics actually use their time is a significant issue, and without it I cannot say how much more funding per student would be needed to make teaching-research academic employment financially stable.

Based on the feedback I have received, I am mildly hopeful that the situation is not as bad as I feared. But the fact that the double count is not leading to impossible conclusions, such as the combined Deloitte and ABS totals exceeding actual expenditure; and the fact that employment is shifting towards more specialised academic work, despite the flexibility generated by international student profits; suggests that the situation is nevertheless pretty bad.