Towards agent-based integrated assessment models: examples, challenges, and future developments. Lamperti, F., Mandel, A., Napoletano, M., Sapio, A., Roventini, A., Balint, T., & Khorenzhenko, I. Regional Environmental Change, 2018.
Towards agent-based integrated assessment models: examples, challenges, and future developments [pdf]Paper  doi  abstract   bibtex   
Understanding the complex, dynamic, and non-linear relationships between human activities, the environment and the evolution of the climate is pivotal for policy design and requires appropriate tools. Despite the existence of different attempts to link the economy (or parts of it) to the evolution of the climate, results have often been disappointing and criticized. In this paper, we discuss the use of agent-based modeling for climate policy integrated assessment. First, we identify the main limitations of cu rrent mainstream models and stress how framing the problem from a complex system perspective might help, in particular when extreme climate conditions are at stake and general equilibrium effects are questionable. Second, we present two agent-based models that serve as prototypes for the analysis of coupled climate, energy, and macroeconomic dynamics. We argue that such models constitute examples of a promising approach for the integrated assessment of climate change and economic dynamics. They allow a bottom-up representation of climate damages and their cross-sectoral percolation, naturally embed distributional issues, and traditionally account for the role of finance in sustaining economic development and shaping the dynamics of energy transitions. All these issues are at the fore-front of the research in integrated assessment. Finally, we provide a careful discussion of testable policy exercises, modeling limitations, and open challenges for this stream of research. Notwithstanding great potential, there is a long way-to-go for agent-based models to catch-up with the richness of many existing integrated assessment models and overcome their major problems. This should encourage research in the area.

Downloads: 0