July 25, 2018

How to optimise economies of scale in universities

Keith Houghton

In 2008, distinguished economist Paul Krugman was awarded the Sveriges Riksbank Prize in Economic Sciences — commonly known as the Nobel prize in economics — for his work on international trade. 

Krugman developed an innovative approach that argued that goods and services can be produced more cheaply by virtue of economies of scale, while consumers demand a varied supply of these products. As a result, small-scale production for a local market is replaced by large-scale production for a global market, where firms with like products compete with one another and consumers are better off.

Can this well-accepted proposition on the economies of scale and consumers being better off apply to the university sector?

At its core, the economies and diseconomies of scale may be a force that shapes the higher education and research framework of the future. Indeed it may not be singular framework but plural frameworks — one each for research and education.

One might speculate that the optimal scale of research is small — it could be a cottage industry in some fields — while the scale of education might become super-sized for some, perhaps many, applications. With the potential reach of online education, might the economics of the education component of universities soon experience radical change?

The landscape has already changed with offerings such as MOOCs (massive open online courses) and HBX CORe (the Harvard Business School platform). While these two examples are very different, they have at least one characteristic in common — there is an educational product that is created once and delivered many times over a wide geographic area. The challenges with such an approach are many, validation of assessment being just one. But for this challenge, there is already a potential solution on the horizon in the form of an Australian innovation: high integrity reusable digital ID.

Much closer to home, I and some colleagues are seeking to provide a window on efficiency and productivity in Australia’s universities. This has led to certain observations on the economies of scale. The empirical analysis has been made both more simple and more complex because of the seemingly unabated push towards larger-scale institutions.

While the rationale for growth in scale is, in many instances, supported by strong evidence, the presence of smaller, highly specialised institutions of higher education — common in comparator markets — is almost non-existent in Australia. Further, while there are incentives for mergers, the incentives and frameworks for demergers are not apparent even when demergers and the creation of more specialised institutions may deliver greater efficiency and, potentially, effectiveness.

Our work has involved the creation of a measure now known as the Research and Education Efficiency Frontier Index. Based on empirical data on the performance of Australia’s 37 major public universities, we create what is known as the efficiency frontier — a two-dimensional output from an econometric model showing the most efficient performers in the sector as well as the less efficient. Importantly, the frontier is agnostic as to preferences in the education/research mix.

In the period 2011 (the base year, as it is the last year prior to the demand-driven system) to 2016 (the most recent year that data is available), productivity growth in the Australian university sector is substantial.

The average is more than 20 per cent across the six years. In all, 13 universities achieved productivity growth of 30 per cent or more — a significant achievement.

With only rare exceptions those institutions at or particularly near the efficiency frontier in 2016 are in the range 15,000 to 50,000 equivalent full-time student load. All but one of the 30 per cent-plus productivity change group are in the 15,000 to 45,000 EFTSL range, the University of New England being the exception.

Generally, the results show that the smaller universities (which also largely overlap with the regional institutions) are rarely at or even particularly near the efficiency frontier. This is true from the first year for which we have data, 2001, until 2016.

At the other end of the scale, the potential for diseconomies of scale may also exist but, based on our observations of the US and British markets, these can be moderated by organisational structure and, perhaps, culture.

In Australia, the largest four universities (by student load) also miss out on inclusion in the 30 per cent-plus productivity gain list.

The sector has been let down by three universities that have delivered negative growth (that is, negative relative to the forward-moving frontier). Indeed, in all three of these cases, their 2016 performance was negative relative to their position in 2011.

In a sector that is supported by taxpayer funding, the delivery of such significant productivity growth is a compelling argument against non-targeted funding cuts.

Could it be that the funding arrangements that have largely been in place since the unified national system was envisaged in the late 1980s have encouraged the existence of only “comprehensive” universities — even when the scale of those institutions is outside the range that enjoys economies of scale? Is the structure of our sector preparing Australia for the potential future changes to the scale of higher education?

Do we have the mechanisms to understand the efficiencies and productivity outcomes of policy decisions?

Although we have a university sector that has been largely built on decades of public funding and today still relies on billions of taxpayer dollars, we need to consider all the options that drive rational policy and performance outcomes in respect of ever-present demands for efficiency and productivity.

Australian National University emeritus professor Keith Houghton is chief academic strategist of the Higher Education and Research Group and Research Coaching Australia. He is a former dean of business and economics at the ANU.

Keith Houghton: ‘We need to consider all the options that drive rational policy and performance outcomes’. Picture: David Geraghty.