This paper tries to answer the question how it is possible that such a giant of sociology as James Coleman produced so little of what he considered the most important work for sociologists: good institutional design. I argue that the answer lies in the fact that the most serious problem for which Coleman pleaded for institutional design (ie to find a substitute for the vanishing primordial social organizations of family, church and neighborhood) could not even be adequately described let alone mitigated or solved by the highly simplified theory of action he borrowed from microeconomics. It is argued that a theory that would lend itself for institutional design in this area would have to be able to accommodate the considerable social influence on rationality.