Participatory AgentBased Modelling for Flood Risk Insurance Sara

Participatory Agent-Based Modelling for Flood Risk Insurance Sara Mehryar, Grantham Research Institute, London School of Economics Swenja Surminski, Grantham Research Institute, London School of Economics Bruce Edmonds, Centre for Policy Modelling, Manchester Metropolitan University

Background “The Zurich Flood Resilience Alliance (ZFRA) is a multi-sectoral partnership focusing on finding practical ways to help communities in developed and developing countries strengthen their resilience to flood risk. ” 2

Flood Resilience Measurement for Communities 3

FRMC implementation: 2013 -2018 -2023 4

Second phase Focus on decision making to support enhancing community flood resilience Insurance as a mechanism to manage the cost of disaster (post-disaster economic management tool) Shift from post-disaster response tool to a pre-disaster behaviour change mechanism 5

Second phase Focus on decision making to support enhancing community flood resilience Insurance as a mechanism to manage the cost of disaster (post-disaster economic management tool) Shift from post-disaster response tool to a pre-disaster behaviour change mechanism Ø Insurance policies may have different impacts on overall flood risk reduction behaviour Ø Insurance policies can encourage more risky behaviour (e. g. , property development in high risk locations with reliance on the insurance support) Moral Hazard Ø Insurance policies can provide incentive mechanisms to trigger desirable risk reduction behaviours (e. g. risk-base pricing, deductibles, and no-claims bonuses). 6

Behavioural rules: Households’ decision in uptake of flood insurance Perception of insurance effectiveness (affordable, available, needed) Flood risk reduction measures Flood insurance • • • 7 Property level protection Flood protection measures Location of development projects Natural protection measures …….

Behavioural rules: Households’ decision in uptake of flood insurance Perception of insurance effectiveness (affordable, available, needed) Flood risk reduction measures Flood insurance • • • Insurance policies 8 Property level protection Flood protection measures Location of development projects Natural protection measures …….

Behavioural rules: Households’ decision in uptake of flood insurance Perception of insurance effectiveness (affordable, available, needed) Flood risk reduction measures Flood insurance • • Insurance policies 9 Perception of flood vulnerability • Perception of flood risk Property level protection Flood protection measures Location of development projects Natural protection measures …….

Behavioural rules: Households’ decision in uptake of flood insurance Perception of insurance effectiveness (affordable, available, needed) Flood risk reduction measures Flood insurance • • Insurance policies 10 Perception of flood vulnerability • Perception of flood risk Property level protection Flood protection measures Location of development projects Natural protection measures …….

Behavioural rules: Households’ decision in uptake of flood insurance Perception of insurance effectiveness (affordable, available, needed) Flood risk reduction measures Flood insurance • • Insurance policies 11 Perception of flood vulnerability • Perception of flood risk Social Capital: Human capital: Financial capital: Social network, institutions (norms and rules) Knowledge and awareness Individual capacity and government support Property level protection Flood protection measures Location of development projects Natural protection measures …….

Participatory ABM • Role playing game • Fuzzy Cognitive Mapping • Narrative Data Analysis 12

Role-Playing Game (RPG) • Observing decisions and actions of actors in a roughly defined setting • May reveal implicit social rules, norms, values and perceptions of actors • Flood Resilience Game: to understand how people invest in different flood risk reduction projects e. g. retrofitting vulnerable properties, uptake of insurance, building physical and natural protection measures given different level of risks, location and economic situations. 13

Fuzzy Cognitive Mapping (FCM) Mehryar, S. , Sliuzas, R. , Schwarz, N. , Sharifi, A. and van Maarseveen, M. , 2019. From individual Fuzzy Cognitive Maps to Agent Based Models: Modeling multi-factorial and multi-stakeholder decisionmaking for water scarcity. Journal of environmental management, 250, p. 109482. 14

Narrative Data Analysis (NDA) • Can be combined with other participatory methods • Interviewing • Transcribing interviews into text • Analysing text • Coding decision rules for an ABM • Possibility to gather additional information that are not achievable through FCM, RPG, or questionnaires • Developing a systematic process for translating narrative into codes requires further work Examples: • Ghorbani, A. , Dijkema, G. and Schrauwen, N. , 2015. Structuring qualitative data for agent -based modelling. Journal of Artificial Societies and Social Simulation, 18(1), p. 2. • Ghorbani, A. , Bots, P. , Dignum, V. and Dijkema, G. , 2013. MAIA: a framework for developing agent-based social simulations. Journal of Artificial Societies and Social Simulation, 16(2), p. 9. 15

Challenges & Opportunities • Applying participatory ABM by using an existing model • Dubbelboer, J. , Nikolic, I. , Jenkins, K. and Hall, J. , 2017. An agent-based model of flood risk and insurance. Journal of Artificial Societies and Social Simulation, 20(1). • Advantages: easy to use and accessible for non-modeler (NGOs, practitioners, etc. ) • Challenges: Cost, time, resources • Verifying whether game players correspond to actions in real life is challenging • FCMs are mainly based on people’s experience of what they have done 16

Thank you! Suggestions, ideas, experiences, questions? s. mehryar@lse. ac. uk 17

Agents and Actions Insurers: Estimate the property level flood risk and inform the homeowners the calculated insurance premiums. Developers: Identify optimal land, submit development proposal to the local government, built new housed and sell them in the market. Local government: May legislate or fund the 1) local flood protection measures, and 2) property level flood protection measures. Homeowners: Decide to 1) Buy and sell the house, 2) renew their insurance contracts, 3) invest in property-level flood protection measures. 18

Community Flood Resilience Different stages of flood risk management Risk reduction Prepared ness Response Recovery Provide capacity for Five capitals of community flood resilience Financial Capital Individual and government financial support 19 Human Capital Social Capital Knowledge and awareness on flood probability and exposure Social network and participation in flood risk management Physical capital Large and small scale flood protection measures Natural Capital Natural flood protection measures