AbstractThis paper describes a framework for policy analysis and design in soft, complex systems based on the transfer to such systems of control modules developed for harder (more easily quantified), simpler systems. The framework is demonstrated by outlining the development of a dynamic allocation algorithm within a system dynamics model of an engineering system and describing its application in a much larger scale management system. The potential use of this algorithm is demonstrated in two other systems, and a generalized form is presented that highlights its isomorphic qualities. It is suggested that this framework, which treats the physical and control structure of a system as a unified whole, can provide a useful contribution to policy analysis.