Much of the debate around the recent independent review of metrics has overlooked its incisive critique of the quality of university research management. establishes that metrics are substituting for real managerial skills, especially in the larger teams that are increasingly favoured, in social sciences and humanities as much as in science disciplines (¡°The weight of numbers¡±, Features, 9 July).

The report, commissioned by the Higher Education Funding Council for England, is not alone in reaching that conclusion. A by Charles E. Leiserson, a professor of computer science and engineering at the Massachusetts Institute of Technology, and Chuck McVinney, a management consultant, noted the poor quality of leadership in many science laboratories. The US National Institutes of Health has been promoting the concept of ¡°team science¡± because of concerns over delivery on large-scale research grants. Even the recent Sir Tim Hunt controversy has underscored the difference between being an eminent scientist and having the analytic skills to recognise a changing social and institutional environment, and the consequences of words or actions.

In the social sciences and humanities, the very idea of research management has barely left the starting gate, with many scholars opposed to any discussion of the concept. Management, as Raymond Williams noted, comes from a medieval Italian word for training horses: fine for lesser creatures but not for professionals. Many academics, regardless of discipline, argue that they should simply be given chunks of money, preferably public rather than private, and left alone.

Part of the problem, as the metrics report implies, is the crass version of management perpetrated in many universities. It exemplifies the worst features of the New Public Management, in which it is expected that brilliant individuals, transformational ¡°future leaders¡±, possess charismatic qualities sufficient to get the best out of subordinates, who deliver their vision. Where charisma is not enough, it can be supplemented by the micromanagement of individual performance against targets intimately linked to pay progression.

ÁñÁ«ÊÓƵ

This approach never had much of an evidence base and has failed pretty much everywhere it has been introduced. Paradoxically, universities provided the home for most of the research that discredited the model, but have been among the organisations most reluctant to abandon it. Some of this reluctance, as Leiserson and McVinney note, is attributable to the dominance of scientific thinking: human systems are expected to be measurable in the same ways as engineering systems, and ¡°soft skills¡± are dismissed. Some of it reflects the need to conform to the external demands of Treasury economists, who aspire to fit a recalcitrant world to the narrow terms of their models. Some of it comes from existing leaders convinced of their own charismatic qualities who would rather pay consultants for flattery than listen to the evidence from their business and management academics.

Research management is, however, essential in modern-day universities. This is in part because of legitimate expectations of accountability for funding, whether public or private. More important, though, is the expectation of diversity: that research will be collaborative, draw on a set of disciplines to solve complex real-world problems and draw strength from researchers¡¯ different experiences, backgrounds and orientations. The growing literature that defines the skills required has established, in particular, that transactional leadership styles valuing and engaging team members are more effective than the ¡°prophet and acolyte¡± model that has dominated in the past. Shouting at a randomly chosen graduate student or postdoc first thing every morning is not leadership.

ÁñÁ«ÊÓƵ

The new styles of leadership emphasise the skills of accountability and persuasion. Learning how to see the world from another person¡¯s perspective may be the most fundamental dimension. Engagement with the wider public is not a chore or a way to correct their ignorance but a precondition for winning their support. People outside your research area may not understand the world in the same way that you do ¨C but this does not mean their concerns are unfounded or illogical.

Research leaders need skills in five areas: as entrepreneurs, networkers, collaborators, mentors and multitaskers. They also need a strong thread of moral commitment: to respect for people, to integrity in research and organisational life, and to responsibility to the society that makes science and scholarship possible. A well-managed university lives by these values.

Research leadership demands at least as much creativity and imagination as any substantive research problem, but if all you want is to be a brilliant individual, you probably should not put yourself in the position of being a team leader. Leadership involves a measure of personal sacrifice and compromise for the benefit of others. You are the person with the existing reputation, so you can afford this better than they can. You are the public and professional face of the group and you will derive a variety of material and symbolic rewards from that.

The price is that you have to share your vision and accommodate to other people¡¯s engagements with it. This is how to encourage high performance. Annual box-ticking just isn¡¯t good enough.

ÁñÁ«ÊÓƵ

Robert Dingwall is a consulting sociologist and part-time professor of sociology at Nottingham Trent University. Mary Byrne McDonnell is executive director of the Social Science Research Council, New York. They are co-editors of The Sage Handbook of Research Management (2015).