Far from being a vague ideal, the complex and sensitive issue of maintaining integrity in science is a critical imperative. In my view, it would help to demand and monitor integrity in scientists and managers from the outset (Nature 476, 251, 262; 2011).

Most researchers know from their training that honesty is fundamental to scientific integrity. But some managers and agency officials can find themselves in difficult situations. A manager must cope with the competing pressures of supporting and protecting the scientists working on a project, ensuring the survival of the scientific institution and pleasing unforgiving political masters — possibly all under public scrutiny. Even an honest manager might fear being undermined by a rival colleague or, worse, by a scientist who is cavalier about professional ethics.

The only way to achieve scientific integrity across the board is to ensure that personal and professional values (as well as knowledge and skills) are primary criteria for the employment of both scientists and managers. These values must be demonstrated and constantly monitored, not just presumed.