Modeling economic resilience

A wide range of climatic and ecological changes are unfolding around us. These changes notably manifest themselves through an increased environmental variability, such as shifts in the frequency, intensity, and spatial distribution of weather-related extreme events. If human societies cannot mitigate these transformations, to which conditions should they adapt? To many researchers and stakeholders, the answer is resilience. This concept seems to subsume a variety of solutions for dealing with a turbulent and uncertain world. Resilient systems bounce back after unexpected events, learn novel conditions and adapt to them. Theoretical models, however, to explore the links between socioeconomic mechanisms and resilience are still in their infancy. To advance such models, the present dissertation proposes a novel conceptual framework. This framework relies on an interdisciplinary and critical review of ecological and economic studies, and it is based on the theory of dynamical systems and on the paradigm of complex adaptive systems. We identify agent-based models as crucial for socioeconomic modeling. To assess their applicability to the study of resilience, we test at first whether such models can reproduce the bifurcation patterns of predator–prey interactions, which are a very important factor in both ecological and economic systems. The dissertation then tackles one of the main challenges for the design of resilient economic system: the large interconnectedness of production processes, whereby disruption may propagate and amplify. We next investigate the role of delays in production and supply on realistic economic networks, and show that the interplay between time delays and topology may greatly affect a network’s resilience. Finally, we investigate a model that encompasses adaptive responses of agents to shocks, and describes how disruptions propagate even though all firms do their best to mitigate risks. In particular, systemic amplification gets more pronounced when supply chains are fragmented. These theoretical findings are fairly general in character and may thus help the design of novel empirical studies. Through the application of several recent ideas and methods, this dissertation advances knowledge on innovative mathematical objects, such as Boolean delay equations on complex networks and evolutionary dynamics on graphs. Finally, the conceptual models herein open wide perspectives for further theoretical research on economic resilience, especially the study of environmental feedbacks and their impacts on the structural evolution of production networks.