What data do we need to value household flood costs? A framework and assessment for storm surge and cloudburst in Denmark

Abstract

Estimates of flood damage to households in Denmark rest mainly on insurance payouts. Payouts capture only part of the cost of flood risk. This paper identifies potential data sources for the full cost that storm surges and cloudbursts impose on households. To estab- lish which data are needed, I first develop a theoretical framework for the household cost of flood risk. The framework records costs in two ledgers, one for the household and one for the rest of society, and sep- arates resource costs from transfers. It covers costs that arise when a flood occurs, such as uninsured and non-material damage, and costs of living with the risk, such as adaptation, risk-bearing and loss aversion. For each element, I discuss how the data needed to measure it could be obtained. I find that most of the cost borne by households is not recorded anywhere today. Data availability reflects how each hazard is insured: storm surge claims are well documented, whereas cloud- burst insurance data are old and fragmented. Danish registers allow health care use, work and moves to be followed, but only if flooded households can be identified. Identifying who was flooded, when and how badly is therefore the main constraint. I recommend a recurring panel survey of flooded and unflooded households, linked to claims records and national registers.