Research programmes appear to represent one of the more powerful instruments through which research funders (principals) steer and shape what researchers (agents) do. The fact that agents navigate between different sources and styles of programme funding and that they use programmes to their own ends is readily accommodated within principal–agent theory with the help of concepts such as shirking and defection. Taking a different route, I use three examples of research programming (by the UK, the European Union and the European Science Foundation) to argue that principal–agent theory cannot capture the cumulative and collective consequences of the relationships it seeks to describe.