OECD Review on the Seine Flood Risk In Paris area. Economic impacts

Transcription

1 OECD Review on the Seine Flood Risk In Paris area Economic impacts

2 Seine Floods in Paris What would be the impact of a major flood in Paris? Cascading effects through the interruption of critical networks Macroeconomic impacts as the region represents 3 % of France national GDP

3 Methodology and general approach Integration of micro and macro approaches Listing of all the potential damages and losses Building on existing databases and tools Hydro-economic model with impact on housing and businesses Emergency planning scenario impacts on critical networks Energy, public transport, roads, water Impact of network interruption on companies Integration of micro-economic impacts to build damages and losses scenarios Modelling medium to long-term macro-economic impact at national scale

4 Methodology and general approach Macro-economic impact on the longer term DSGE model of the national economy to integrate non-linear effects and adjustments mechanisms Impact of a shock of capital destruction (public and private) and of a reduction of the output on: National growth (GDP) Employment level Public finance Impact of the French compensation insurance scheme CAT-NAT and the State liability Trimestrial time-step corresponding to the dynamic of the crisis and of the recovery process

5 Working with scenarios around the historic 191 flood Threshold effects and existing data sources CHARACTERISTICS S 1 S 2 S 3 Discharge ( / 191) 8 % 1 % 115 % Water level (Paris) 7,32 m (1924 flood) 8,12 m 8,62 m (191 flood) Duration 1 week 2 weeks 1 month Population affected Impacts on critical networks Electricity Transport Water Partial distrurbance localised during two weeks Large disturbance during 2 months in the suburbs of Paris Global distrubance 1,5 m clients without electricity Subway damaged lines closed for months 3 m clients without water supply Disturbance to economic activities 2 weeks 1-2 month 2-5 month

6 Hazard mapping through hydraulic modelling Hydraulic model with a meshing of boxes representing the flood plain along the 3 main rivers 3 flood scenarios : Water level Percentage of the box flooded Duration of the flood

7 Damages calculation Direct damages to housing Land use maps, construction costs, water depth-based damage functions Dammages to economic activities Categories of activity, density of workers in each category per land use types, database on equipments, stocks and turnover per worker and category Damage functions calculating % of damages to stocks and equipments based on water depth and flood duration (+/ 48h), proportionnate damages to business real estate, and business losses based on the turnover per worker and the duration of the submersion Direct damages to public equipments Public equipments (schools, hospitals, administrations, sport facilities) Limits: critical networks and busines interruptions caused by their interruption

8 Impacts on networks: transport

9 Impacts on networks: power

10 Impacts on networks: water 1

11 Impacts on networks: damages estimation Estimations ORSEC, experts, OECD Electricity Nb of power sub-station (2 m ) flooded and impacted Nb of distribution HV/LV (2 k ) Transport Subway: estimation of damages 1-5 Bio Train : one portion of RER 5 m, 18 months to repair Roads: 44 km of highways + 2 km of large roads, damage function 1 %, average cost 6 m for 1 km Water : 3 Bio infrastructures in the flood plain. Levels of protection exists: 3% damages in S3 only

12 Indirect impacts Spill-over effects of network interruption Electricity: key critical network Surface impacts by electricity interruption = twice the flooded area Interruption last longer than the flood Transport: nb of affected workers during the event and after the event based on reconstruction scenario Companies bankruptcy 19 SMEs in high risk zones (> 1m), one fifth in very high risk USA Small Business Administration: 25 % of SMEs go bankrupt 25% of SMEs in very high risk zones (A) and 1% in high risk (B) Turnover SMEs 1,25 m (OSEO) Impact on output

13 DIRECT AND INDIRECT IMPACTS (Billion ) CAPITAL DESTRUCTION BUSINESS LOSSES Working with scenarios around the historic 191 flood Housing Furniture Business Stocks Business Equipement Business Property SCENARIO S 1 S 2 S TOTAL PRIVATE Electricity Public Transport Roads Water Other public equipments TOTAL PUBLIC Bankruptcy - - TEMPORARY Caused by the flood Caused by electricity disruption Caused by transport disruption A , B ,

14 Macro-spill over effects A Dynamic Stochastic General Equilibrium Model Principles Private and public capital Calibration to the national economy Quarterly time-step aligned with the dynamic of the crisis and the reconstruction process The flood is modelled as a shock : Destruction of public and private capital Reduction of the output Reconstruction with the CAT-NAT insurance system and public investment 14

15 Why use a general equilibrium model? To model the effects on firms, households, the government, and their interactions To evaluate the effects of the flooding on impact and in the aftermath, i.e. the recovery Variables : Households: employment, wages & consumption Businesses: Private investment and output Government: public investment, taxe rate and public debt Principle: Maximisation of househols well-being, companies profit and public policies effectiveness

16 Results : scénario 1.2 Private Capital.5 Public Capital.2 GDP Private Investment Public Investment Employment Wages.1 Government Debt Private Consumption

17 Results: scénario 2 2 Private Capital 5 Public Capital 2 GDP Private Investment Public Investment Employment Wages Government Debt Private Consumption

18 Results : scénario 3A 5 Private Capital 5 Public Capital 5 GDP Private Investment Public Investment Employment Wages Government Debt Private Consumption

19 Results : scénario 3B 1 Private Capital 5 Public Capital 1 GDP Private Investment Public Investment Employment Wages Government Debt Private Consumption

20 Results Immediate reduction of GDP (-.15 à -2.7 %) and employment levels (-.9 à -1.7 %) in the quarter of the flooding Immediate increase of the public debt caused both by the reduction of the tax-base and the CAT-NAT reimbursement needs The speed of the recovery varies accross scenarios but does not exceed 1 year especially with the effect of reconstruction (public capital restoration) 2

21 Results : 5-year consolidation S1 Year GDP EMPLOYMENT DEBT GDP EMPLOYMENT DEBT % Bio % Jobs % Bio % Bio % Jobs % Bio S3A S2 S3B Year GDP EMPLOYMENT DEBT GDP EMPLOYMENT DEBT % Bio % Jobs % Bio % Bio % Jobs % Bio

22 Conclusion A hybrid micro-macro approach Concept of «reasonable worst case scenario» Focus on indirect effects and macro effects with a maximisation of existing data and tools Significant impacts over 5 years : 3-3 Bio of direct damages 1-6 Bio of GDP losses over 5 years.1-3 % increase of public debt 12-4 job losses in the year of the crisis Not a systemic risk «Perfect storm» 22

23 ANNEXES 23

24 Methodology Tangible Intangible Direct Impact Indirect Impact Physical damages to : - Housing - Infrastructure - Business stocks and equipments - Businesses losses - Critical network interruption - Emergency costs - Losses of lives - Health - Environment - Cultural heritage l Direct impacts: Classic approach based on hydraulic modelling and spatialised damage functions Impact indirects: Integration of interdependencies between sectors through economic modelling (I/O, DSGE) 24

25 Scenario Probabilities? With or without the dams? S1 S2 S3 Maximum water level 6,9 8,15 m 9,11 m Return period with dams 3 years 7 years 15 years Return period without dams 5 years 13 years > 3 years 25

26 Damages calculation (1) Direct damages to housing 6 different types of housing based on land use maps, Value based on construction costs per m² for each housing types, Damage functions based on water depth

27 Damages calculations (2) Dammages to economic activities 19 simplified categories of activity Evaluation of the density of workers in each category per land use types and per Equipments, stocks and turnover per worker and per category of activity Calculation per category: Damage functions : % of damages to stocks and equipments based on the water depth and flood duration (+/ 48h) Damages to real estate proportionate to damages to stock and equipments Business losses calculated based on the turnover per worker and the duration of the submersion

28 Damages calculations (3) Direct damages to public equipments Public equipments (schools, hospitals, administrations, sport faclilities) Damage functions close to the housing functions

29 Coûts en Millions d'euros CCR, Etudes techniques, 213 Comparison CCR-OCDE 2, 18, 16, 14, 12, 1, 8, 6, 4, 2, S1 Q1 S2 Q1.15 S3 Q1.4 Estimation OCDE Estimation CCR Indice de confiance CCR 29

30 Return period Mean Annual Damages 1/2 1/25 Avec réservoirs Sans réservoirs 3/1 1/5 1/ Damages (Bio ) Seine 25-5 m /// Bio France 3

31 Calculation of the CAT-NAT compensations Calculation of the deductibles: 38 for households contracts covering 1 % of the dammages 1% for business damages 3 working days for business losses of companies directly affected by the flood CAT-NAT reserves estimated at 5.7 Bio CAT-NAT. Beyond this level : State Guarantee CAT-NAT compensations (Bio ) Dammages Households Businesses Scenario S1 S2 S Total damages Business losses Bankruptcy - - A - B - Total losses: Direct effects of the flood Secondary effects

32 Representative Household Derives utility from consuming and leisure Discounts the future using a factor β Decides how much to consume (C) and work (N) as well as how much to save in government bonds (S g ) and private corporate bonds (S p ) Taking as given interest rates (r g on government debt and r p on private debt), wages (w), taxes (τ), and firms profits (Π)

33 Representative Firm Output (Y) is produced using public capital (X), private capital (K), and labour (N) Maximises an infinite stream of cash flows, using the households discount factor Taking interest rates and wages as given Can borrow by issuing private corporate bonds (B p ) to invest into their capital stock (K)

34 Government The government invests into the public capital stock (X), levies income taxes (τ) and issues government bonds (B g ) Benevolent government maximise the welfare of the representative consumer taking into account the effect of these policies on the economy, i.e. how firms and households react