Elsevier

Ecological Economics

Volume 68, Issues 8–9, 15 June 2009, Pages 2265-2277
Ecological Economics

Analysis
Willingness of homeowners to mitigate climate risk through insurance

https://doi.org/10.1016/j.ecolecon.2009.02.019Get rights and content

Abstract

Climate change is projected to increase flood risks in certain regions due to an increase in both precipitation and sea level rise. In addition, socio-economic scenarios project an increase in urbanization in flood prone areas, which results in a higher damage potential. The combined effect of climate and land use change on flood risks requires innovative adaptation policies to cope with rising risks. Increasingly, attention is paid to the role insurance can play in mitigating damage by providing incentives to policyholders to undertake damage reducing measures. The willingness of homeowners in the Netherlands to undertake measures that mitigate flood damage in exchange for benefits on hypothetical flood insurance policies is examined using surveys. The results indicate that many homeowners are willing to make investments in mitigation. In particular, approximately two-thirds are willing to invest in water barriers in exchange for a premium reduction and about a fifth are willing to replace floor types that are vulnerable to flooding with water resistant floor types. Furthermore, about a quarter are willing to move central heating installations to floors safe against flooding in favor of a reduction in the insurance premium. Estimates of the effectiveness of these mitigation measures to limit potential flood damage in the river delta indicate that prevented damage could be substantial, namely in the order of 1 billion euro or larger. Reductions in (absolute) flood risk due to mitigation are especially large under climate change. A probit model indicates that existing arrangements for compensating flood damage, risk awareness and perceptions, and geographical characteristics are important determinants in the decision to undertake mitigation.

Introduction

Climate change is expected to increase the frequency and severity of flooding in certain regions (IPCC, 2007). This is because a warmer climate is likely to result in a more vigorous hydrological cycle, which may cause more extreme precipitation. For the area of the Netherlands, which covers most of the delta of the rivers Rhine and Meuse, it would mean that the consequences and probability of flood risk may rise (e.g., Middelkoop et al., 2001, Aerts et al., 2006, Ward et al., 2008). Moreover, warming may result in melting of ice caps, in particular, of the Greenland and Artic ice sheets, which will contribute to sea level rise (Alley et al., 2005). The latter may increase flood risks from storm surges in low-lying delta regions that are already vulnerable to flooding, such as the Netherlands. This could have disastrous socio-economic consequences (Bouwer and Vellinga, 2007, Olsthoorn et al., 2008).1

The decay rate of greenhouse gasses is very slow and hence their presence in the atmosphere2 with the accompanying negative consequences will continue in the coming decades (IPCC, 2007, Matthews and Caldeira, 2008). Even if current climate policy would be able to stabilize greenhouse gasses to 2000 levels, then a further warming of about 0.2 °C would occur in the next two decades, while warming is expected to be twice as high than under 2000 levels if emissions increase in accordance with the SRES scenarios of IPCC (2007). This implies a clear need for adaptation policy (Stern, 2007, Pielke et al., 2007). In addition to heightened risk due to climate change, increases in wealth and population contribute to the vulnerability of societies to natural disasters (Bouwer et al., 2007).

In response to increased flood risks caused by climate change, several adaptation measures have been examined and are currently being implemented in the Netherlands (Kabat et al., 2005). These pertain to maintaining current flood probabilities by heightening of primary river dikes and improving of coastal protection. In addition, projects that create more space for rivers are undertaken, which also aim at maintaining or lowering the flood probability (e.g., Vis et al., 2003). Another strategy considered is to focus on lowering potential flood damage by developing ‘flood proof’ houses and structures, which minimizes damage during floods. For example, Aerts et al. (2008a) indicate that elevating a newly built home 5 m above sea level costs approximately in between € 10,000 and € 20,000 per house. This is a substantial investment for a homeowner but has the advantage that it completely eliminates flood risk. Other ‘flood proofing’ measures that partly reduce flood risks can be considerably less expensive, such as replacing floors vulnerable to flooding with tile floors, especially when replacement is necessary anyway after a flood or because of tear. A combination of measures that both limit damage and reduce the probability of flooding is likely to be the most effective way of preventing the occurrence of extremely large flood damages, as Aerts et al. (2008b) show in a portfolio framework of water management investments.

Insurance arrangements for flood risk may require households to undertake measures that mitigate damage or stimulate households to undertake precautionary measures voluntarily (e.g., Kleindorfer and Kunreuther, 1999, Kunreuther and Pauly, 2006). For example, households who build their house in a way that is more resistant to flooding can be rewarded with premium discounts or higher levels of coverage. These mitigation measures may limit damage during floods and thus be complementary to traditional flood protection. In addition, insurance arrangements serve a useful function in providing financial protection against the residual risk that remains after flood protection infrastructure has reduced risk to optimal levels. Practical experience suggests that individuals rarely undertake mitigation measures voluntarily (Kunreuther, 2006a). It is therefore very relevant to examine whether individuals can be stimulated to take precautionary measures in exchange for benefits, such as discounts, on their insurance policy. Although a number of studies have addressed the issue of how to create incentives for households to control and reduce flood damage through insurance, empirical analysis of the effectiveness of such incentives are rare.

International experiences suggest that mitigation measures may be an effective instrument to limit flood damage. For example, precautionary measures undertaken by (both insured and uninsured) households were very effective in limiting flood damage in Germany during the extreme flood event of the river Elbe in 2002 (Thieken et al., 2005, Kreibich et al., 2005).3 Flood damage could be limited by adapting use of buildings to flooding, which means that cellars and stories susceptible to flooding are not used cost-intensively, and by adapting interior fitting to flooding, which comprises the use of waterproofed building materials and placing of easily movable furniture on low floors. Use of buildings and interior fitting adapted to flooding were especially effective and reduced damage on buildings by 46% and 53%, and damage on contents by 48% and 53%, respectively (Kreibich et al., 2005). Thieken et al. (2006) show that insured households undertook more mitigation during the Elbe flood than uninsured ones. In particular, 28.5% of the insured households undertook at least one of the mitigation measures examined compared with 20.5% of the uninsured.4 After the 2002 flood, the city of Dresden was hit again by flooding in March 2005 and April 2006. Increased awareness of flood risk after the 2002 flood resulted in more private precautionary measures being undertaken, which limited flood damage significantly (Kreibich and Thieken, 2007). Another example is the National Flood Insurance Program (NFIP) in the USA, which by setting (compulsory) mitigation standards contributed to reduce flood losses on new structures by about six times (Pasterick, 1998). Nevertheless, the NFIP program failed to restrain development in flood plains, which may be the case because many premiums are not risk based but partly subsidized (Burby, 2001).5 For example, in the last decades considerable development of new structures took place in New Orleans, which augmented damage of hurricane Katrina (Burby, 2006).

The present study examines the willingness of Dutch households to undertake voluntary mitigation measures for insurance benefits. So far, this has not been studied. A survey was undertaken among approximately 500 homeowners in the river delta of the Netherlands. The respondents were asked if they are willing to undertake specific mitigation measures for benefits on a hypothetical flood insurance policy. The effectiveness of these mitigation measures to prevent damage caused by river flooding and reduce flood risk in the river delta under climate change is examined. In addition, we statistically analyze how perceived risks of flooding, knowledge about flooding, experience with flooding, household and geographical characteristics, and responsibilities for covering flood damage influence the willingness to undertake mitigation.

The remainder of this paper is structured as follows. Section 2 briefly discusses Dutch flood risk management and the role of damage mitigation. Section 3 explains the survey. Section 4 examines descriptive statistics of the answers to the mitigation questions. Section 5 provides a range of estimates of the potential of mitigation to limit flood damage. Section 6 discusses estimation results of a statistical analysis of the factors behind the decision of homeowners to invest in water barriers. Section 7 concludes.

Section snippets

Flood risk management in the Netherlands

The Netherlands is a densely populated country in which millions of its inhabitants live around or below sea and river water level. Many low-lying parts have been reclaimed from former lakes and are protected by so called ‘dike rings’ along the main rivers and coastal areas. A dike ring is a geographical unit bounded by a flood protection system, such as dikes (Fig. 1). It is also a separate administrative unit under the Water Embankment Act from 1996. The latter aims to guarantee a certain

Explanation of the survey

The willingness of households to undertake mitigation measures for certain benefits on hypothetical insurance policies is examined using a survey. The four mitigation measures considered are the purchase of sandbags for a premium discount, the purchase of a water resistant floor type if damage from floors vulnerable to flooding is not covered, removal of certain machines (laundry and dryer machines), and central heating boiler to higher floors for a premium discount. These particular measures

Mitigation measure 1: buy sandbags to create a water barrier

Sandbags can act as a water barrier that protects the house from flood damage if they are placed in front of doors or low windows during a flood (ICPR, 2002). This measure is only effective if water levels are low, but there is uncertainty about the exact maximum water level at which water barriers are still effective. The maximum height of waterproofing is approximately 1 m above the ground according to Kelman and Spence (2003). In contrast, the International Commission for the Protection of

Estimating the potential contribution of mitigation in limiting flood damage

It is useful to examine how much damage may be prevented by undertaking the four mitigation measures included in the survey. This in turn allows for identifying what measure is likely to be most effective in limiting damage. The last column of Table 1 indicates the proportion of homeowners who will undertake a specific mitigation measure. This information is used to estimate damage prevented by mitigation, using two approaches. First, damage prevented by the four mitigation measures is examined

A statistical model of the decision to mitigate: buy sandbags as a water barrier

The benefits received by the policyholder to undertake mitigation measures, for example premium discounts, are likely to be the main determinant in the decision whether to undertake the measure. However, other factors may influence decisions to undertake mitigation measures as well, like household and geographical characteristics. Insight into these factors is of interest for two main reasons. First, it is relevant from a social welfare perspective to know what types of households are willing

Conclusions

Climate change is projected to increase flood risk in the Netherlands because of more extreme precipitation, increased river runoff and sea level rise. This requires managing the probability of a flood through dike reinforcements and measures that limit flood damage. Experiences in various countries suggest that mitigation measures at the household level can be an effective means to limit flood damage during floods. For example, households who undertook certain measures ex-ante of the 2002

Acknowledgements

We thank our colleagues at the Institute for Environmental Studies (IVM), and especially Roy Brouwer and Sebastiaan Hess, for useful comments on our questionnaire. We are grateful to Ada Ferrer-i-Carbonell for suggestions on the econometric analysis and to Alfred Wagtendonk for assistance in preparing the GIS data. We thank Heidi Kreibich for providing information and Isabelle Seifert, Laurens Bouwer and two anonymous referees for helpful comments on this paper. This research project was

References (58)

  • BurbyR.J.

    Flood insurance and floodplain management: the US experience

    Environmental Hazards

    (2001)
  • KunreutherH.C. et al.

    A behavioral model of the adoption of protective activities

    Journal of Economic Behavior and Organization

    (1985)
  • van PraagB.M.S. et al.

    The anatomy of subjective well-being

    Journal of Economic Behavior and Organization

    (2003)
  • AertsJ.C.J.H. et al.

    Sensitivity of global river discharges under Holocene and future climate conditions

    Geophysical Research Letters

    (2006)
  • Aerts, J.C.J.H., Bannink, B., Sproing, T.A., (2008a) Aandacht voor Veiligheid.www.adaptation.nl. IVM...
  • AertsJ.C.J.H. et al.

    Dealing with uncertainty in flood management through diversification

    Ecology and Society

    (2008)
  • AlleyR.B. et al.

    Ice-sheet and sea-level changes

    Science

    (2005)
  • BlumenscheinK. et al.

    Eliciting willingness to pay without bias: evidence from a field experiment

    Economic Journal

    (2008)
  • BotzenW.J.W. et al.

    Insurance against climate change and flooding in the Netherlands: present, future and comparison with other countries

    Risk Analysis

    (2008)
  • Botzen, W.J.W. and van den Bergh, J.C.J.M., (2008b). Monetary valuation of insurance against climate change risk. FUR...
  • Botzen, W.J.W. and van den Bergh, J.C.J.M., (2009). Bounded rationality, climate risks and insurance: is there a market...
  • Botzen, W.J.W., van den Bergh, J.C.J.M. and Aerts, J.C.J.H., (2008). Report on a Survey about Perceptions of Flood...
  • Botzen, W.J.W., Aerts, J.C.J.H. and van den Bergh, J.C.J.M., (2009). Factors shaping individual perceptions of climate...
  • BouwerL.M. et al.
  • BouwerL.M. et al.

    Confronting disaster losses

    Science

    (2007)
  • BurbyR.J.

    Hurricane Katrina and the paradoxes of government disaster policy: bringing about wise governmental decisions for hazardous areas

    The Annals of the American Academy of Political and Social Science

    (2006)
  • de VriesE.J.

    Vergoeding van rampschade

    Nederlands Juristenblad

    (1998)
  • DomencichT. et al.

    Urban Travel Demand a Behavioural Analysis

    (1975)
  • DWW. (2005). HIS Hoogwater Informatie Systeem voor Rampenbestrijding bij Overstromingen. Delft: Rijkswaterstaat...
  • FEMA

    Protecting Building Utilities from Flood Damage — Principles and Practices for the Design and Construction of Flood Resistant Building Utility Systems

    (1999)
  • GrossmanP.Z. et al.

    Uncertainty, insurance and the learned hand formula

    Law, Probability and Risk

    (2006)
  • HeijC. et al.

    Econometric Methods with Applications in Business and Economics

    (2004)
  • ICPR

    Non Structural Flood Plain Management: Measures and Their Effectiveness

    (2002)
  • IPCC

    Climate change 2007: the physical science basis

  • KabatP. et al.

    Climate proofing the Netherlands

    Nature

    (2005)
  • KaplowL.

    Incentives and government relief for risk

    Journal of Risk and Uncertainty

    (1991)
  • Keijzer, X., (2008) The Effectiveness of Flood Mitigation Measures. MSc thesis Free University Amsterdam, June...
  • KelmanI. et al.

    A limit analysis of unreinforced masonry failing under flood water pressures

    Masonry International

    (2003)
  • KleindorferP. et al.

    The complementary roles of mitigation and insurance in managing catastrophic risks

    Risk Analysis

    (1999)
  • Cited by (0)

    View full text