When university metrics backfire: what REF, TEF and NSS can learn from academic gaming behaviours

This blog was kindly authored by Dr Shahenda Shehata, Assistant Professor in Accounting at Heriot-Watt University.

Performance metrics are meant to improve universities. But when metrics become targets, they can start to reshape the behaviours they are supposed to measure. In UK higher education, frameworks such as the Research Excellence Framework (REF), Teaching Excellence Framework (TEF) and National Student Survey (NSS) play a powerful role in shaping research, teaching, reputation and resource allocation. They matter. But they also create incentives that deserve closer policy attention.

My doctoral research examined how performance measurement systems shape academic behaviour in UK business schools. Drawing on qualitative evidence from 52 interviews across 18 UK business schools, alongside document analysis and social media evidence, the research explored how academics and academic managers respond to metric-driven accountability regimes. The findings suggest that performance systems can encourage strategic responses that may look rational at an institutional level but can create unintended consequences for academic work, staff wellbeing and university governance.

This does not mean that metrics are useless or that accountability should be abandoned. Universities receive public funding, charge significant tuition fees and play a major role in the UK economy and knowledge system. They should be accountable. The problem is not accountability itself, but the way accountability systems are designed and used.

When metrics become too dominant, they can shift attention from educational and scholarly substance towards the management of measurable signals. For example, research assessment may encourage academics to prioritise journal ranking strategies over riskier or more socially relevant questions. Teaching metrics may encourage institutions to manage student satisfaction rather than address deeper questions of learning, challenge and intellectual development. Promotion systems may encourage academics to focus on what is visible and measurable, even when other academic contributions are equally valuable but harder to count.

In my research, I found that academics and managers may engage in different forms of ‘gaming’ in response to performance pressure. These behaviours include managing publication portfolios; aligning research topics with perceived journal expectations; bargaining around promotion criteria; optimising teaching indicators; and responding strategically to student satisfaction systems. Some of these behaviours may be defensive: academics are trying to survive in demanding systems. Others may be organisational: managers are trying to improve institutional performance in competitive environments. Either way, the result can be a system in which the metric becomes the target.

This matters for higher education policy because gaming behaviour can weaken the very objectives that performance systems are intended to support. If metrics encourage short-term compliance, symbolic performance or strategic presentation, they may distort resource allocation and reduce trust in evaluation frameworks. They may also intensify pressure on academic staff, contributing to stress, anxiety and burnout. In the long term, poorly designed metrics can narrow academic freedom, weaken collegial decision-making and make universities less innovative.

A better approach would not be to remove performance metrics altogether. Instead, universities and policymakers should treat metrics as diagnostic tools, not final judgements. Numbers can identify patterns and prompt questions, but they should not replace professional judgement. REF, TEF and NSS data should be interpreted alongside qualitative evidence, peer review, contextual information and meaningful dialogue with staff and students.

Second, performance systems should be designed with greater attention to unintended consequences. Before introducing or changing metrics, policymakers and university leaders should ask: what behaviour will this encourage? What will people stop doing in order to meet this target? Who benefits from this measure, and who may be disadvantaged by it?

Third, accountability frameworks should recognise the value of academic work that is difficult to quantify. Mentoring, curriculum innovation, collegial service, public engagement, intellectual risk-taking and student development all matter. Yet these activities can be overlooked when evaluation systems prioritise outputs that are easier to count.

Finally, academics should have a stronger voice in the design and interpretation of performance systems. A more participatory approach would help restore workplace democracy within universities and reduce the gap between those who design metrics and those who live under them.

The lesson from academic gaming is not that academics are resistant to accountability. Rather, it is that people respond to the incentives placed in front of them. If universities are judged mainly through narrow metrics, academics and managers will adapt accordingly.

The UK higher education sector is one of the country’s most important intellectual, social and economic assets. Its performance systems should support that value, not undermine it. REF, TEF and NSS can play a constructive role, but only if they are used carefully, contextually and democratically. Good metrics should illuminate academic quality, not distort academic purpose.