A Novel Approach to Policy Design Using Process Design Principles

Policy design, as it is currently practiced, relies on the manual formulation of policies. This fact, combined with an increase in the complexity of the systems under study, results in a large portion of the possible choices for action being left unexplored. The introduction of a systematic approach for exploring alternative policies using a computational methodology has the potential to help in the decomposition of complex problems into subproblems with more manageable size, and can accelerate and improve the effectiveness of the policy-making process. A new approach to facilitate policy design is proposed; it uses principles widely used in chemical engineering design combined with mathematical and artificial intelligence techniques. In particular, our ideas are based on the Hierarchical Design Method for the conceptual design of chemical processes ( Douglas, 1988) and use of techniques such as network analysis and agent-based modelling. A six-step framework has been proposed and is being implemented as a prototype decision support system (DSS). An agent-based modelling approach, due to the opportunistic nature of agents is used as the programming paradigm in the design of policies. The DSS creates a network of policy measures and extracts information from the network during the development of policy alternatives. A set of agents then combine policy measures and goals within a tree structure at different hierarchical levels during the creation of the overall policy. The results from the research are a fresh contribution to the methodological development of policies using a multidisciplinary approach and have the potential to accelerate the design of policies and improve their success rate.