Individual Flood Preparedness Decisions During Hurricane Sandy in New York City By prof.dr. Wouter Botzen

Transcription

1 Individual Flood Preparedness Decisions During Hurricane Sandy in New York City By prof.dr. Wouter Botzen

2 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

3 1. Individual adaptation measures in flood risk management Trends in global flood risk due to: Population and economic growth Climate change impacts: sea level rise, precipitation, storms River flood risks in 2080 under climate (RCP 8.5) and socio-economic change (SSP5) Source: Winsemius et al. (2016), Nature C.C.

4 Flood risk management approaches Traditional focus on flood-prevention infrastructure Engineering options: dikes, dams, storm surge barriers Can be cost-effective in flood-prone regions, like the Netherlands

5 Flood risk management approaches (2) Shift to integrated flood risk management approaches Both public and private actors take measures to limit flood risk Installing flood-prevention to limit risk to zero is often infeasible Household level measures can significantly limit flood damage Key research questions Which household scale measures effectively limit flood damage? Which of these measures are also cost-effective? Do households take cost-effective measures, and why (not)?

6 Assessments of Household Flood Risk Reduction Flood damage models Estimate flood damage savings from household level floodproofing measures with hypothetical flood simulations Rely on assumptions of effectiveness of risk reduction measures Useful for guiding building code policies in an area Source: Poussin et al. (2013), NHESS

7 Assessments of Household Flood Risk Reduction (2) Statistical analysis of survey data Collect data of flood damages and household flood preparedness Estimate damage savings from specific risk reduction activities Examine factors of influence on flood preparedness behavior Key findings Germany: up to 50% of flood damage savings by water-proof building materials and placing furniture on higher floors (Kreibich et al., 2005) France: similar flood damage savings for elevating homes and raising electricity installations (Poussin et al., 2015) But evidence base is small, and results are regional dependent

8 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

9 1938 Hurricane: Long Island Express Casualties 600 NY 1938 Max water level 4-6m #destroyed houses 14,000 Damage US $ 6-8 bn

10 Historical trend in building exposure to floods, in NYC Aerts and Botzen (2012) Nature C.C.

11 Irene caused $1.5 billion of costs

12 12

13 Impact Sandy in New York City (NYC) Flooded 17% of the City s land 88,700 buildings in flooded areas $19 billion of costs

14 Flood Damage Mitigation Measures Wet-flood proofing buildings Dry-flood proofing buildings 14

15 Elevation

16 Flood risk analysis framework 1000 US

17 Overview of Model Integration and Data Sources Hazard Hurricane-ADCIRC storms - Probabilities - Surge heights Source: Lin et al. (2012), Nature Climate Change Exposure NYC building stock - 33 categories - Values per census block Source: NYC Office of Emergency Management Costs - Unit costs of floodproof buildings Source: Aerts et al. (2013), ANYAS Inundation model - LiDAR elevation -Inundation depths per census block Source: Aerts et al. (2013), Risk Analysis Vulnerability HAZUS -flood damage model - Depth-damage curves - Building, contents, vehicle damage - Depends on flood-proofing Source: Federal Emergency Management Agency CBA Risk

18 Cost-benefit analysis results building codes Flood-proof existing buildings Yields $1.5 up to $3 per $1 invested Higher elevation of new constructions +6ft in coastal zone yields $3 up to $15 per $1 invested +4ft in inland zones yields $1.5 up to $9 per $1 invested Source: Aerts et al. (2014) Science

19 Behavior w.r.t. Low-Probability Risk Biases imply insufficient preparedness for floods: Underestimation of low-probability risks Threshold models Costs of information seeking Myopia and discounting of the future Charity or moral hazard of compensation

20 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

21 3. Survey flood preparedness NYC during Sandy Conducted by phone March-April 2013 Random sample, completion rate is 73% (N=1,035) Location can be linked to objective flood risk >100 questions about: Risk perceptions Flood experience and compensation Behavioral motivations for preparedness Insurance purchases Risk mitigation measures Socio-economic characteristics

22 Survey flood preparedness NYC Bronx Manhattan Queens Staten Island Brooklyn Source risk model: Aerts et al. (2013) Risk Analysis

23 4. Results flood risk perceptions High flood risk awareness 87% are aware that they live in a flood-prone area 13% were not aware Impact Hurricane Sandy 62% indicated Hurricane Sandy increased their flood risk perception Flood experience 69% were flooded during Sandy

24 Significant influence of political ideology on risk perceptions and expectations of federal relief Source: Botzen et al. (2016) Climatic Change Mediation analyses: - Democrats have higher demand for flood-proofing through higher climate change risk perceptions

25 Accuracy of risk perceptions Comparison perceived flood damage and probability (PR) with objective counterparts (OR), correct if Where error margin (EM) is 0.25, 0.5 or 0.75 Results EM=0.5 Correct Underestimation Overestimation Perceived probability 24% 25% 51% Perceived flood damage 33% 47% 19% Source: Botzen et al. (2015) JDM

26 Statistical model of perceived flood damage Underestimation when: (No) past flood damage Flood risk is below threshold level of concern Low worry for flooding Trust in government flood risk management Low education Source: Botzen et al. (2015) JDM

27 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

28 5. Flood insurance and risk reduction activities Functions flood insurance: Spread risks Attract capital from financial markets Claim reimbursements speed up recovery Regulation and incentives for risk reduction (?)

29 Economic literature about relation insurance and risk reduction Market failures can cause higher risk taking of insured Adverse selection (Akerlof, 1970, QJE) Moral hazard (Rothschild, Stiglitz, 1976, QJE) Not always supported by empirical studies Limit adverse selection due to misperceptions of risk (Botzen, van den Bergh, 2012, IER) Advantageous selection based on behavioral characteristics (de Meza, Webb, 2001, RAND J. Econ.; Finkelstein, McGarry, 2006, Cutler et al AER) Risk aversion, feelings like worry, social or private norms

30 Survey results flood insurance purchases 33% 2% 21% purchased flood insurance voluntary purchased flood insurance mandatory no flood insurance 44% don't know

31 % of respondents Relation insurance and ex ante flood-proofing 60 Purchased flood insurance voluntarily Purchased flood insurance mandatorily No flood insurance ** ** ** ** ** ** Dry proofed walls Pump and drainage Water-resistant materials Water-resistant floor Elevated utility and electric installations Note: ** indicates a significant difference at the 5% level with the no flood insurance group

32 % of respondents Relation insurance and emergency preparedness Purchased flood insurance voluntarily Purchased flood insurance mandatorily No flood insurance ** Flood shields and sand bags Move contents from floodprone parts of house Note: ** indicates a significant difference at the 5% level with the no flood insurance group

33 Models flood insurance purchases

34 Results model flood insurance (1) Explanatory variable Emergency flood preparations mandatory (n=445) voluntary (n=278) -0.09*** -0.11*** Ex ante flood preparations 0.05*** 0.05*** Perceived flood probability Perceived flood severity * Federal disaster compensation -0.21** -0.12*** Low income (<25,000) *** High education *** Note: ***, **, * indicate significance at the 1%, 5% and 10% levels, respectively.

35 Results model flood insurance (2) Purchases: Behavioral model voluntary (n=370) Emergency flood preparations -0.12*** Ex ante flood preparations 0.07*** Federal disaster compensation -0.13*** Low income (<25,000) -0.21** High education 0.18*** Flood probability below threshold of concern -0.11** Peace of mind 0.12** Norm of preparing for floods 0.18** External locus control -0.15*** Note: ***, **, * indicate significance at the 1%, 5% and 10% levels, respectively.

36 Results model flood insurance (3) Purchases: Behavioral model voluntary (n=336) Emergency flood preparations -0.12*** Ex ante flood preparations 0.08*** Low income (<25,000) High education 0.22*** Flood probability below threshold of concern -0.11* Peace of mind 0.16** Flood proofing measures Norm of preparing for floods 0.08** Flood proofing measures External locus control -0.04* Flood proofing measures Received disaster assistance -0.11*** Flood proofing measures Experienced flood damage 0.001*** Note: ***, **, * indicate significance at the 1%, 5% and 10% levels, respectively.

37 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

38 6. Conclusion Flood risk assessments and cost-benefit analysis can guide flood risk management Behavior has a large influence on hazard vulnerability Individuals base their decisions on flood risk perceptions which deviate from objective risk People view risk reduction and flood insurance as complements instead of substitutes We find that moral hazard in flood insurance is absent Behavioral and financial variables explain advantageous selection

39 Agenda 1. Context: Individual adaptation measures in flood risk management 2. Flood risk management in New York City 3. Survey household flood preparedness during Hurricane Sandy 4. Results individual flood risk perceptions 5. Flood insurance purchases and risk reduction activities 6. Conclusion 7. Ongoing work: flood preparedness in Tabasco, Mexico

40 7. Ongoing work: flood preparedness in Tabasco, Mexico Collaboration Wharton School, Red Cross, and Zurich Insurance 63 question survey 68 households in 10 communities

41 Preliminary findings Tabasco survey Flood preparedness activity Protect belongings 46% Change water purifying method 42% Evacuation plan 22% Emergency preparedness 14% % of households Community activities improve individual flood preparedness Community flood protection Availability of shelters Flood hazard maps Emergency warning systems

42 References Akerlof, G The market for lemons. The Quarterly Journal of Economics, 84(3): Botzen, W., van den Bergh, J Monetary valuation of insurance against flood risk under climate change. International Economic Review, 53: Cutler, D., Finkelstein, A. McGarry, K Preference heterogeneity and insurance markets. American Economic Review, 98: de Meza, D., Webb, D Advantageous Selection in Insurance Markets. RAND Journal of Economics, 32: Emanuel, K Climate and tropical cyclone activity. J. Climate, 19: Finkelstein, A., McGarry, K Multiple dimensions of private information. American Economic Review, 96: Kreibich, H. et al Flood loss reduction of private households due to building precautionary measures. Natural Hazards and Earth System Sciences, 5(1): Lin N., Emanuel K., Oppenheimer, M., Vanmarcke, E Physically based assessment of hurricane surge threat under climate change. Nature Climate Change, 1389: Poussin, J.K., Botzen, W.J.W., Aerts, J.C.J.H Effectiveness of flood damage mitigation measures. Global Environmental Change, 31: Rothschild, M., Stiglitz, J Equilibrium in competitive insurance markets. Quarterly Journal of Economics, 90: Winsemius et al Global drivers of future flood risk. Nature Climate Change, 6:

43 Publications presented here Aerts, J., Botzen, W., Emanuel, K., Lin, N., de Moel, H., Michel-Kerjan, E Evaluating flood resilience strategies for coastal mega-cities. Science, 344: Aerts, J.C.J.H., Botzen, W.J.W Managing exposure to flooding in New York City (NYC). Nature Climate Change, 2: 377. Aerts, J., Lin, N., Botzen, W., Emanuel, K., de Moel, H Low probability flood-risk modeling for New York City. Risk Analysis, 33: Botzen, W.J.W., Kunreuther, H., Michel-Kerjan, E Divergence between individual perceptions and objective indicators of tail risks: Evidence from floodplain residents in New York City. Judgment and Decision Making, 10(4): Botzen, W.J.W., Michel-Kerjan, E., Kunreuther, H., de Moel, H., Aerts, J.C.J.H Political affiliation affects adaptation to climate risks: Evidence from New York City. Climatic Change, 138(1),

44 Thanks for your attention! Contact information: