The university coalface gets 28 cents in the dollar!

The question posed last week was how much of the money sent into the university sector at the point of DEEWRDEST actually reaches the coalface in terms of teaching and research. My best guess answer is about 28 cents in the dollar, with the rest essentially going into admin.

The way to calculate this is more properly explained in this presentation I did for the Australian Conference of Economists in 2011 (Presentation academonomics ACE July 2011), but the simplified version is the use of this formula:

Money to coalface = 100* (1- cut going to DEEWRDEST) * (1- cut going to uni admin) * (1-%time of acad on admin)

We thus break the calculation into three unknowns: the percentage creamed off by the education ministries itself, the percentage creamed off by university admin, and the percentage academics themselves have to spend on admin.

The first one of these is actually the hardest: how much do the ministries take. In their own yearly accounts, they claim that ‘overhead’ is a pittance in terms of the funds they manage, ie 1.8%. This is not just almost nothing, it is completely unbelievable since it means they count their own activities as directly valuable.

In stead of taking the account numbers on their official overhead as useful, I prefer to simply look at how many people DEEWRDEST employs. Using the administrative handbook, DEEWRDEST appears to employ 6000 staff plus consultants and outside agencies. At the going wage of around 68,500 per standard civil servant, plus super, plus the usual 30% on-costs in terms of buildings and a 30% guess as what they pay in terms of non-employees, this means DEST ‘costs’ around a billion dollars. At a total budget of 12 billion going to universities in 2011, this means one should think of DEST overhead as 8%. This does not even count all the state education ministries so is probably an underestimate.

Then the issue of the percentage going to university administrators. As I explained in a previous post, universities cook the books by counting casuals and part-time academics in their academic staff numbers, as well as people who are streight administrators with fancy titles. Instead of using the fanciful figure of 45% that they themselves produce as their estimate for how much of the wage budget goes to administrators, I looked at random staff in the phone books of 5 Australian universities, counted the 50 employed on the basis of what they did, and concluded that at least 56% of university wages gets paid to administrators and not academics. That too was being very generous as it did not count the costs of the consultants and advisers who are off the phone books of universities but definitely belong to the administrators’ camp. I thus take 56% as not just the percentage of the wage bill spent on admin but also as the most reasonable guess for how much of the costs of buildings and grounds is attributed to admin rather than academics.

Then the issue of how much time academics spend on administration. I took around 6 published studies on this issue, all based on time-use surveys. At best guess, around 30% of time of academics is spent on administrative tasks like sitting on committees, writing reports, training for the advent of cyclones, presenting in front of visitation committees, sitting on representative bodies, etc. Since it is usually the more highly paid who are doing administrative work, like being on Faculty Boards, the 30% is again an underestimate of the real percentage spent on administrative activities. By all accounts, this percentage is a lot higher than it used to be simply because of the increase in forms being sent out, compulsory training courses, etc.

What does this mean for the percentage ending up at the coalface? We can now calculate the coalface number as 100*(1-0.08)*(1-0.56)*(1-0.3)=28. So 28 cents in the dollar that comes in in DEEWRDEST actually gets spent on direct costs of teaching and research, i.e. the people and the buildings doing the work. The other 72 cents is overhead.

I would of course advocate someone to be given the resources to do this more properly, though it must be said that one clearly has to ignore much of the statistics given out by both DEEWRDEST and the universities themselves on the composition of their staff, the nature of their expenses, and the use of their time for they are not reasonable numbers. So whoever wants to do a better job than the back-of-the-envelope calculation I engaged in will need the resources to gather their own data.

All kudos to those who gave a try, many of whom gave numbers chillingly close to mine. DEEWRDEST overhead statistics for the rest.

Author: paulfrijters

Professor of Wellbeing and Economics at the London School of Economics, Centre for Economic Performance

14 thoughts on “The university coalface gets 28 cents in the dollar!”

  1. If someone were to do a more serious investigation into this sort of thing, it would be interesting to see how Australia stacks up globally. It would also be useful to know what correlation there is between the percentage taken by overhead and research output.


  2. I’m sorry Paul but you have just lost a great deal of my respect here with this sloppy diatribe. Firstly, DEST has not existed since 2008 when it merged into DEEWR. Secondly, while it is indeed a huge Department of over 6000 employees, a very large part of that goes to administering employment and workplace relations functions, to say nothing of important policy development functions. Thirdly, the finances are on the public record so you can easily work out exactly how much if you care the spend some time looking instead of blustering. Not saying there aren’t things that can be improved, but I can’t believe you let your normally high standards fall so low. Caveat: I work there, but in employment.


    1. Glad someone wants to take up the challenge! I use the word DEST because its an acronym well known in Oz academia (DEST points), which is why I use DEST/DEEWR, so that is no real issue.
      As to what people do there, you tell me what the proportion is working on the tertiary sector. I looked up the financial reports of 2 years ago and could not make heads or tails of where its employees and buildings were counted towards. Your financial statements makes no clear distinction between ministry expenses and the outward expenses. It includes all kinds of funds towards this or that group of which it is not clear what the overhead is.
      But prey, give it a try yourself if you think your financial statements are readable: have a look at the linked queensland education budget for instance (which will include a form of overhead I did not even count!) and tell me where the ministry expenses involve in tertiary education are to be found ( The equity statement on the buildings for instance will include not just the buildings in which the civil servants sit but also a lot of buildings in which teaching takes place. The ‘personel costs seem to include way more an just thhr ministry personel’. And what to make of the ‘for women’ accounts, or ‘grants towards minorities’. If you think these statements really tell you where the overhead is or even how many employees work for different purposes, go ahead and enlighten me. We can then redo that bit of the calculation. Given that I would want to include all the state civil servants, I doubt very much we are going to arrive at less than 6000 civil servants involved in higher ed, but we will see.


  3. Paul: your “phone book” methodology confuses raw admin/academic head count with actual salary costs and benefits that result from HR position classification, the individual university’s EBA, and individual contract negotiations. It may not be a representative sample. Academic Level salaries are often higher than HEW salaries and include access to research incentives and other institutional funding (such as School-based “soft funds”). This reality is missing from the public debate focus on senior administrators and consultants. One “continuing” Academic Level E professor can earn more than several “short-term contract” HEW administrators — the former may also have a defined benefit plan whereas the latter has a defined contribution plan for superannuation. Finally, there are administrators who publish academic research included in HERDC data collection — and they often don’t get access to the same research incentives.


    1. yep, certainly a possibility. Not really all that hard to find out: from the position title (which is in the phone book for some and by googling them you can find the position title of the others) you can pretty accurately work out the salaries (which are invariably on-line) except for the top brass whose salary is secret. Also, you wont know the loadings which are also secret. But you can get some idea. i haven’t done it, but my guess would be that given the low wages for casual academics that the average salaries wont be all that different, though i agree that the salaries for the administrators will be more unequal than the salaries amongst academics.


  4. DIISRTE is now the relevant department. According to its annual report, there is $34 million allocated to running the relevant bureaucracy – which covers higher ed, voc ed, international ed, student assistance and the loan scheme. Total spend is over $9 billion. Add a bit more to cover the central DIISRTE bureaucracy, but a low % of administered spending.


    1. Hi Andrew,

      yes, its that kind of estimate that is just unbelievable. If you take the on-costs cited by Matthew below, 34 million would buy you, what, 200 civil servants?
      I am perfectly happy to concede there wont be 6000 in Canberra and that it might be closer to 1000 there, though over the country as a whole it is sure again in the order of several thousands, but 200? Even if we just focus on Canberra? It is exactly that kind of published figure that makes one very weary of the financial reports: they clearly have their own method of counting overhead that does not concord with how an outsider would think of it. Which means that to get a proper estimate you would need to give someone the resources to check actual activities and numbers of concerned civil servants.


      1. Paul – I spoke to the higher education division’s annual staff meeting earlier this year and wandered around their offices. There were a few on video link from the office of teaching and learning in Sydney, but DIISRTE is essentially a Canberra only operation for funding. I’d say that the 200 estimate was around right. There would be more in the ATO administering HELP, TEQSA staff, and a few in central agencies who have higher ed responsibilities. They are not like the schools people in DEEWR dealing with thousands of schools.


  5. Hi Paul

    Regarding your estimate of the cost of the department, the 30% for on-costs (you forgot overheads) seems fairly low. The rate that gets used in the analysis that I’ve seen carried out is closer to 80-90%


    1. for ARC grants they use 28% on-costs so that is where my 30% guess came from, but i agree the real number is higher. However, the big question in terms of the Diisrte/Dest/Deewr side of things is just the number of civil servants employed in Canberra and the states involved in university and tertiary education affairs. Its a very hard number to get an estimate of because it will involve all the policy developers and might also include people from different departments (such as those involved in productivity perhaps). Really though, the department itself should calculate this number.


  6. Andrew Norton,

    you think 200 administrators in DIISRTE is about right? This is the sort of thing one can find out. At you will find for DIISRTE the following numbers:

    DIISRTE in 2012-2013 has 3256 employees. On top of that there are 90 in a standards agency (definitely overhead), 208 in the Australians Skills Quality Authority (would yuou count those as tertiary? Certainly overhead). A 1000 in IP Australia (surely a bit university-related, no?). Some 5600 in CSIRO (surely quite a few administrators in there. I am willing to bet well over 50%). 115 in the ARC (certainly overhead). 1150 in the Australian Nuclear Organisation (university-related?). And another 300 in smaller specific ones.

    So, where are the 3256 employees of DIISRTE counted as overhead? Apparently no more than 200. And none of these extra ones seem to be counted as overhead for the tertiary sector either. It is really amazing, Andrew. God knows what they count, but it aint look right to me.

    Indeed, seeing these numbers I am reverting back to my original assumption that 6000 employees employed in overhead is just about right. Indeed, it is starting to look like more than that if we would include all the state bureaucracy elements too…..


Comments are closed.