Broadening the impact of university research on society (Nature 465, 416–418; 2010) should be included in the academic reward structure.

The present scientific reward system threatens to imprison academics in their ivory towers. It is ruled by bibliometric quality indicators spawned by the rise in systematic performance evaluations (L. K. Hessels, H. van Lente and R. Smits Sci. Public Policy 36, 387–401; 2009). Originally a means of communication, publication has become an end in itself.

Demonstrating the relevance to society of a research proposal helps to get it funded. But in practice, scientists are rewarded for their contribution to a field's progress, and not for its impact on society or the economy.

Hiring and promotion criteria and the evaluation of research groups and institutes need to be expanded to include factors for broader impact, alongside publication records and citation scores. Research councils could consider shifting part of their money from input funding to output funding. Cash awards might also be offered for broader-impact results (Nature 465, 398; 2010).