How research evaluations can damage scientific quality: 23 years of evidence from the UK Research Excellence Framework

Abstract

Periodic research evaluations are necessary to allocate funding, but they also create pressure on researchers. Between grants, short contracts and university assessments, researchers often state that evaluations leave them little time to pursue novel research agendas and publish in more challenging outlets.
Using data on 495,128 publications from the UK’s governmental university assessment, we show that evaluations affect research output. Publications just before assessment deadlines obtain substantially fewer citations and are published in lower impact-factor outlets. These trends reverse abruptly after the deadlines.
Our results highlight the importance of an appropriate length for evaluation periods.

Keywords: Research evaluations, researcher incentives, REF