Policy Process Theories and Agent Based Modelling

Recent advances in complexity theory and political science have led to a growing recognition of their inter-connections. Departing from the traditional view of policy as a top-down instrument controlled by one central authority, it is increasingly being described as a self-organization process in which self-conscious actors play different roles. Following this recognition, there is a need for better tools to analyse policy problems, using more advanced methodologies that can reflect the heterogeneous preferences and non-linear interactions involved in the policy process, such as agent-based models (ABMs).

While ABMs are already being used to produce insights to inform public policy, the process of how certain policies come to be often remains implicit within model design. Models typically assume a certain policy was assigned, and then examine its effects on different agents and the environment. In modelling terms: there is an insufficient endogenous representation of the policy process within agent-based models.

This project argues that policy process theories can become the basis for creating better informed ABMs: models that make the policy process endogenous and explicit, reflect the complexity embedded in policy making, and rely on established causal relations to describe the policy system at hand. Such models can support policy practitioners in designing more resilient policies, and provide policy scholars a useful tool to examine the inner workings of the policy world.

The main research question is:

“How can policy process theories’ conceptualization and formalization improve policy analysis in a complex systems paradigm, explicitly and endogenously simulating policy formation?”

The project further elaborates a set of epistemological questions such as

  • How is policy described and captured in terms of complexity theory?
  • What are the advantages and disadvantages of using agent-based models to simulate the policy process, and when is it appropriate?
  • Can ABMs overcome the inherent “unknowable” facets of the policy process, and is this challenge unique to computer simulation of policy?
  • What efforts have already been made to endogenously integrate insights and theories from political science and policy analysis into agent-based modelling?

An additional set of operational questions is also raised:

  • Which policy process theories are most compatible with complexity theory, and with agent-based model design?
  • How can the rationales and behaviours described in policy process theories be translated into agent-based models?
  • What information is required to build an ABM with endogenous representation of the policy process in varying degrees of specification and granularity?
  • What is the most cost-effective approach for integrating policy emergence in ABMs?
  • Does policy process simulation affect the outcomes of ABMs aiming to capture energy systems, mobility systems, and their transformation?
  • What is the additional value that policy process simulation adds to the modelling effort?

In order to meet the research goals, so far a review of existing policy-centric agent-based models was conducted, policy process theories were chosen that are compatible with agent-based modelling, a meta-model was created that uses a common language for each of the theories, the meta-model was implemented in a Netlogo code, and the resulting model was used in several case studies relating to environmental, energy and smart city policies. Finally, the project will develop a methodology to guide modellers and policy scientists in how best to use and expand the model in the future.

Promoter: Paulien Herder
Daily Supervisor: Igor Nikolic

Amit Ashkenazy

PhD candidate

Engineering Systems and Services

Energy and Industry

Research interests:
Policy analysis, policy theories, political science
Agent based modelling, complexity theory, complex adaptive systems
Environmental policy, industrial ecology, urbanism

About Amit Ashkenazy

Originally from Israel, Amit specializes in environmental policy and its application in the local, national, and the international arena. Before commencing his PhD studies on simulating policy emergence at TU Delft, Amit graduated with a master’s degree in environmental management from Yale University, and has worked as Director of the Social-Environmental Caucus in the Israeli Parliament. Amit has gained years of experience in research and consulting to government ministries, foundations and scholars on matters of environmental policy and sustainability initiatives. His PhD studies are supervised by Dr. Igor Nikolic and Prof. Paulien Herder at TPM’s Energy and Industry section.