Not so simple to reward performance

The Australian government intends to fund universities based on their performance in retaining and graduating students. Up to $500 million of commonwealth grants scheme funding is proposed to improve accountability, student outcomes and graduate employment.

Performance funding could potentially provide students with better information and focus universities on teaching quality.

However, our new research reveals the full complexity, risks and contradictions of the endeavour.

Three principles are required to ensure an equitable and effective performance funding model.

First, student equity needs to be an explicit objective. There is variable institutional commitment to widening participation. One public university has a low socio-economic student participation rate of less than 4 per cent, while another is close to 40 per cent.

The most disadvantaged students are spread very unevenly across the sector. Retention rates, completion rates and graduate outcomes also differ by equity group. Performance funding objectives need to align with the broader national objectives of higher education, including legislative commitments to equity.

Second, embedding equity principles in performance funding requires the rewarding of genuine performance rather than simply outcomes. Student outputs often reflect student inputs, leaving little clarity as to the actual performance or value added by an institution. For example, the Australian universities with the highest student retention rates are typically the universities that ­attract the most academically prepared students.

Without further context, providing funding according to success or retention rates simply would reward the most selective institutions and encourage greater selectivity among others. Such a proposal would not specifically reward performance and also would undermine the enshrined principles of equity and diversity.

Between 2006 and 2009, the Australian government ran an early version of performance funding, the learning and teaching performance fund. More than half of these funds were allocated to the Group of Eight universities, suggesting the controls for student inputs were limited.

Identifying performance is complex. We examined institu­tional outputs across student success, retention, completion, satis­faction and graduate outcome metrics. The study found some correlation, for example, between retention and completion outcomes. Predictably, univer­sities that are good at retaining students also are good at ensuring their completion. However, many other potential measures of performance appear contradictory.

Short-term employment outcomes share no positive correlations with student success, retention, completion or satisfaction. Similarly, student satisfaction is negatively cor­related with all of the other measures examined. Some of these paradoxes are explained by modes of education. Online students have low retention and completion rates but high graduate outcomes. Part-time students also have strong graduate employment outcomes but low ratings on all other measures.

Designing performance metrics is difficult and prone to unintended consequences. Modes of study, for example, also are correlated with equity groups. Swinburne University’s growth in online enrolments has led to lower retention rates but increased enrolments from disadvantaged students.

Conversely, a recent sharp decline in part-time enrolments in Britain aligns with declining applications from disadvantaged students. Funding objectives must be clear to avoid a reduction in equity and access.

Flaws in performance metrics also occur through a focus on particular points in time. The wide institutional variations in graduate employment outcomes typically even out after three years and gaps in completion rates narrow after nine years. Snapshot metrics rarely capture improvement across time or nonlinear graduate pathways. Existing measures remain proxies of performance at best.

More work is needed to understand the specific institutional contribution to graduate outcomes. Similarly, research is ­required to measure the “learning gain” of students during the course of a degree. In the US, efforts to measure learning gain include the Collegiate Learning Assessment tool, a task undertaken by students at the start and completion of their degrees. In Britain, several pilot programs are being conducted to isolate learning gain for the teaching excellence and student outcomes framework.

Australian public universities also record relatively positive and similar results on most potential performance metrics. Greater variation exists among the non-university providers, which enrol both the least and most satisfied students in Australian higher education. Including all institutions within a performance funding framework would improve its ­equity and accountability.

Finally, an equitable performance funding model would be student-centred. This approach would involve including students in the design and assessment of any proposed model and ensuring inclusion of the student voice within the performance metrics themselves.

Further, models would provide clear and transparent information that students easily could access and understand, including for non-university providers. Unless models are explicitly student-centred, performance funding could exacerbate inequity, with information accessible only to the most privileged students.

Performance funding could potentially reshape institutional league tables and lead to more ­informed student choice. To be ­effective, though, the model would need to embed student equity and develop more sophisticated measures of actual performance.

Andrew Harvey is an author of the report Principles for Equity in Higher Education Performance Funding, which is available at

Originally published in The Australian - 'Not so simple to reward performance'