This year we held our inaugural University Strategic Decision Making conference in Melbourne on the 3rd and 4th of April for our client and non-client universities. With 14 institutions in attendance we were very pleased with the turn out.
At this point I would like to thank those that came and took two days out of their busy schedules to participate in the conference and exchange ideas and thoughts on the various financial modeling and analysis issues they all face.
So who was there?
Deakin University, Edith Cowan University, University of Western Australia, RMIT University, Swinburne University of Technology, James Cook University, Flinders University, The University of Sydney, University of Western Sydney, University of Tasmania, Charles Sturt University, University of Wollongong, University of Canberra and the L H Martin Institute.
Some of the highlights were presentations by client universities and how they are currently using their University Management Models and University Predictive Models to address a number of pressing management and decision making issues including allowing senior managers to pre-empt major disruptions and changes to operations and prepare for them well in advance. Following is a small selection from these presentations.
James Cook University discussed how they had implemented three historical models within nine months and that their primary focus is analyzing their large suite of course offerings and the financial performance of each. JCU undertook analysis of course surplus by demand as well as course surplus by operating cost to segment courses into four quadrants. This provides an excellent framework to focus management on specific strategies that could be adopted to “power up, optimize delivery or power down” specific courses. Next steps for JCU are to analyze deeper to discover the “whys”, moving towards predictive modeling and long range financial projections, including financial performance in course reviews and engaging business managers to start using the results of the analysis and execute agreed strategies.
Deakin University has been a proponent and long-time user of predictive and scenario modeling, they currently have six years of historic data in their University Management Models. The historic models are used to provide a range of data to management including the Performance Report Scorecard, presented to the senior executive, annual course reviews, faculty and school performance. The predictive model is used to inform the budget process, looking at impacts of Government changes and supporting the 10 year strategic plan.
Edith Cowan University told us how they are using the predictive model to better manage and prepare themselves for a forecasted significant drop in student numbers due to an adjustment some years ago in Western Australia to the school entry age.
Andrew Faulkner explained how the predictive model works and how it is being used by the LH Martin institute to train future university leaders. The L H Martin Institute offers a course for senior university leaders and asked Pilbara Group to provide a model that could be used to realistically represent a university and demonstrate the impact of various scenarios that leaders could confront when managing their university. Pilbara Group worked alongside Professor Bill Massy (Former vice-President for Business and Finance and former vice Provost, research, Stanford University) who provided strategic assistance and advice to the group.
Andrew offered the conference an abbreviated insight into the scenarios that were presented to the course attendees. While the model was used to show how their university looked after a number of external “shocks” were introduced (including a reduction in international student numbers and a change to government funding) resulting in spiraling deficits, course attendees were allowed to offer solutions which were worked into the model by the Pilbara Group team. Gradually and slowly options were developed to eradicate the disastrous loss to a modest surplus and then migrate into a much stronger financial position in outer years.
This session highlighted to those universities with an existing Pilbara Group predictive model the power they have at their disposal to plan and overcome problems that maybe looming over the horizon. It also demonstrated that with careful planning and a clear understanding of the financial impact of external shocks in future years, dramatic and unpalatable decisions affecting staff and students could be avoided or at least minimized.
We were also privileged to have had Ashok Vadgama from CAM-I give an excellent brief on the importance of maintaining a high level of data quality in the day to day operations of any organisation. Our other guest speaker was Professor Leo Goedegebuure from the L H Martin Institute who gave attendees an insight into his view of what challenges faced the university sector in the future.
At this conference we introduced the concept of bench marking results of historical models across client universities. Six universities agreed to participate and allow their models to be compared anonymously.
The presentation started by comparing publicly available data and looked at total FTE against student populations across Australian universities, student staff ratios, general FTE against student population, student staff ratio for general staff and a comparison of General FTE to Academic FTE across all Australian universities. Bench marking across rankings was also examined and presented but there was no great surprise in this analysis.
Andrew then moved on to compare the six universities who used the Pilbara Group costing methodology. Through the Pilbara Group Historical Model (University Management Model) the cost of research can be effectively decoupled from the cost of teaching in a robust and defensible fashion. This enables the creation of benchmarks examining the cost of teaching independently from the cost of research.
The focus of this initial exercise was on the cost of teaching with the aim of providing some initial figures that can guide the focus of more detailed benchmarks in future analyses. It was performed as a blind bench marking exercise where all of the contributing universities’ data is amalgamated to produce mean, maximum and minimum figures. These were then returned to each individual university showing where they sit within this range.
All of the teaching benchmarks were broken down by funding cluster as this provides a high level grouping that can be compared across universities regardless of internal structure. It would be quite feasible to break this down into further common categories using ASCED (Australian Standard Classification of Education) codes.
The following benchmarks were presented.
- EFTSL per Academic FTE by funding cluster.
This is effectively a Student Staff Ratio, but it is only comparing the portion of academic time that is actually contributing towards teaching, as the research component of their time has already been isolated.
- Academic Hours per EFTSL by funding cluster.
The inverse of the above benchmark using Hours rather than FTE to produce a meaningful figure.
- Total Expense per EFTSL by funding cluster.
The fully burdened cost of teaching (including apportioned university overheads).
- Overhead Expense per EFTSL by funding cluster.
The university overhead expense (non-faculty expense) per EFTSL.
- Faculty Academic Salary per EFTSL by funding cluster.
A sub component of the direct expense per EFTSL, the Academic Salary expenses incurred within the faculties.
- Faculty General Salary per EFTSL by funding cluster.
A sub component of the direct expense per EFTSL, the General Salary expenses incurred within the faculties.
- Faculty Non Salary Expense per EFTSL by funding cluster.
A sub component of the direct expense per EFTSL, the Non Salary expenses incurred within the faculties.
All of the above benchmarks examine the expense or FTE per total EFTSL within each category. In producing these bench marks it was apparent they should be broken down into further levels of detail allowing separate analysis of:
- On Campus v External v Off Shore teaching
- Undergraduate v Post Graduate Coursework v Post Graduate Research
As the current bench marks include all types of EFTSL they can be swayed by those institutions containing large numbers of external, off shore or post graduate EFTSL within a funding cluster.
Funding Cluster “8 – Dentistry, medicine, veterinary science, agriculture” has a wide range as individual universities frequently have just one or two of these sub categories resulting in quite disparate cost bases.
The final benchmark presented was:
- Class Hours per On Campus EFTSL by funding cluster.
The amount of timetabled class hours divided by On Campus EFTSL. It is effected by class sizes (smaller classes produce a higher Hours per EFTSL) and contact hours per week.
As users of ACE on-Demand are well aware, the models created utilising the university data are rich in information and provide a valuable tool for senior management. The key issue is presenting data to a range of decision makers in a variety of formats. In his session Adam looked at the two most universal reporting tools, the inbuilt OLAP client and Microsoft Excel.
Adam presented a live demonstration of linking MSExcel with ACE Models and Cubes and demonstrated how the analysis of ACE models can be conducted through this very familiar interface.
Following Adam, Brenden Russell from CALUMO presented the next generation of reporting tools. Brenden went through a range of issues to show the power and versatility of the CALUMO reporting and analysis solution.
We also covered a number of technical areas near and dear to the hearts of our users and of course us! Michelle went into a wide ranging presentation on the management of model data and five steps to better reporting, which were:
- Framing the question
- Selecting the appropriate cube
- Adding filters
- Adding dimensions
- Adding measures
She also lead a discussion on different ways of allocating revenue in the models, touching on Responsibility Centered Management and the allocation of corporate overhead and the different methods employed by a number client universities.
As you can see the conference ranged from the strategic presentations of Professor Leo Goedegebuure, Ashok Vadgama and the possibilities of detailed benchmarks across the sector through to the technical issues of data management and model interrogation. We enjoyed the interaction with both our client and non-client university colleagues and gained enormously from everyone’s input.
We at Pilbara Group would again like to thank all those that participated and presented.
We look forward to doing it again next year.