How are national science advisory systems constructed and to what extent do different systems result in different outcomes?