Strategy Data Resultats Discussion Additional evidence Catastrophes naturelles et interruption d’activité Yanos Zylberberg CREI - Pompeu Fabra Gestion et mesure des risques environnementaux, 24 jan. 2012 1/ 18
Strategy Data Resultats Discussion Additional evidence Compta! Fonction de production: t L 1 − α Y t = A t K α t Theorie: t = dY t , t − 1 = dA t , t − 1 + α dK t , t − 1 + ( 1 − α ) dL t , t − 1 y i Y t − 1 A t − 1 K t − 1 L t − 1 Empirique: � � �� y i d i d i y i t + ε i t = β d + ˜ t − E t t 2/ 18
Strategy Data Resultats Discussion Additional evidence Stratégies empiriques Equation: y i t = β d d i t + β p p i t + β x X t − 1 + ν i + γ t + ε i t Identification: d i t = α e e i t + α p p i t + α x X t − 1 + ρ i + δ t + µ i ( 1 ) t (S) t = β d � y i t + β p p i t + β x X t − 1 + ν i + γ t + ε i d i ( 2 ) t X t − 1 = Gross capital formation, current account, exports, government consumption and reserves 3/ 18
Strategy Data Resultats Discussion Additional evidence Contribution Machine de guerre! 30000 catastrophes: exhaustif. mesures objectives. Interpretation Littérature: Strobl (2011), Noy (2009), Raddatz (2009) 4/ 18
Strategy Data Resultats Discussion Additional evidence Sources (180 pays, 1980-2006) y i t : World Development Indicators d i t : EM-DAT: The OFDA/CRED International Disaster Database e i t : Joint Typhoon Warning Center et PREVIEW Global Cyclones Asymmetric Wind speed Profile USGS et National Geophysical Data Center Gridded Population of the World (CIESIN) 5/ 18
Strategy Data Resultats Discussion Additional evidence Dataclysm (180 pays, 1980-2006) Figure: Cyclones depuis 1980 et densité de population en 2000 6/ 18
Strategy Data Resultats Discussion Additional evidence Construction Dans une zone ζ : densité d e energie e Mesure locale: e ≥ ¯ d e if e Mesure aggrégée: � 1 e ≥ ¯ e d e d ζ 7/ 18
Strategy Data Resultats Discussion Additional evidence Description - carte 100 pays, 500 cyclones and 600 tremblements de terre 8/ 18
Strategy Data Resultats Discussion Additional evidence Description - table VARIABLES Unaffected Cyclones Earthquakes Both Countries 91 46 34 39 Population (’000 000, 2006) 1’019 1’016 414 4’056 GDP per capita (2006) 14737 9496 15584 7297 Private credit (% GDP , 2006) 43.43 54.80 79.11 57.36 Agriculture (% GDP , 2006) 14.51 15.76 12.24 11.63 Manufacture (% GDP , 2006) 13.42 13.11 13.00 15.74 Industry (% GDP , 2006) 34.77 30.18 25.93 28.78 Reported losses (cyclones) 0 0 7.70 4.64 Reported affected (cyclones) 0 0 3.01 5.05 Affected (cat. 2 cyclones) 0 0 57.52 56.29 Affected (cat. 4 cyclones) 0 0 12.66 3.12 Reported losses (quakes) 0 0.48 0 0.26 Reported affected (quakes) 0 1.22 0 0.40 Affected (VI quakes) 0 14.64 0 36.67 Affected (VIII quakes) 0 0.96 0 0.91 9/ 18
Strategy Data Resultats Discussion Additional evidence Exemples Country/Year Reports Obj. measures Earthquakes Country Year Losses Affected Thresholds V VII VIII Pakistan 2005 .218 .033 .101 .014 .001 Japan (Kobe) 1995 .019 .004 .649 .100 .017 Salvador 2001 .112 .223 6.89 .293 .038 Chile (Santiago) 1985 .091 .122 7.01 .343 .112 Cyclones Country Year Losses Affected Thresholds 1 3 5 US (Katrina) 2005 .010 .002 .106 .007 .001 US (Andrew) 1992 .008 .006 .152 .034 .004 Honduras (Mitch) 1998 .729 .353 .919 .025 .001 Japan (Songda) 2004 .002 .000 .476 .109 .005 10/ 18
Strategy Data Resultats Discussion Additional evidence Cyclones 30% 11/ 18
Strategy Data Resultats Discussion Additional evidence Tremblements de terre 10% 12/ 18
Strategy Data Resultats Discussion Additional evidence VARIABLE Annual damages declared in EM-DAT (% GDP) .00248 *** Energy × density (cyclones) (.00017) s 0 .01171*** .00399*** .00552*** .00476*** .00435*** (cyclones) (.00079) (.00099) (.00097) (.00097) (.00095) s 1 .02540*** -.00417 .00307 .00336 (cyclones) (.00202) (.00275) (.00283) (.00279) s 2 .06465*** .02403*** .02185*** (cyclones) (.00419) (.00592) (.00585) s 3 .05998*** .08820*** (cyclones) (.00624) (.00668) s 4 -.14950*** (cyclones) (.01364) .00024 Energy × density (quakes) (.00016) s 0 .00078* .00052 .00022 .00040 .00041 (quakes) (.00046) (.00072) (.00070) (.00070) (.00069) s 1 .00155 .01031*** .01050*** .01084*** (quakes) (.00271) (.00379) (.00379) (.00374) s 2 -.03297*** -.03086*** -.03248*** (quakes) (.01077) (.01104) (.01091) s 3 -.02248 -.04354 (quakes) (.02690) (.02920) s 4 .10390* (quakes) (.05959) .05540 .05779 .08670 .12980 .14550 .16596 Adj. R-squared .132 p-value(quakes) .091 .158 .003 .004 .001 .000 p-value(cyclones) .000 .000 .000 .000 .000 Propensities yes yes yes yes yes yes FE country country country country country country Observations 5,171 5,171 5,171 5,171 5,171 5,171 13/ 18
Strategy Data Resultats Discussion Additional evidence Specification (S) - second stage Panel A y t − y t − 1 d t (% GDP) -0.055 -0.139 -0.388 -0.390*** -0.374*** -0.386*** (0.041) (0.322) (0.252) (0.136) (0.134) (0.104) s c , q s c , q s c , q s c , q s c , q Instruments - 0 0 , 1 0 , 1 , 2 0 , 1 , 2 , 3 0 , 1 , 2 , 3 , 4 Controls - - - - - - Observations 5,098 5,098 5,098 5,098 5,098 5,098 Panel B y t − y t − 1 d t (% GDP) -0.059 -0.008 -0.444* -0.455*** -0.444*** -0.447*** (0.107) (0.310) (0.238) (0.095) (0.093) (0.094) s c , q s c , q s c , q s c , q s c , q Instruments - 0 0 , 1 0 , 1 , 2 0 , 1 , 2 , 3 0 , 1 , 2 , 3 , 4 Controls yes yes yes yes yes yes Observations 3,986 3,986 3,986 3,986 3,986 3,986 Panel C y t d t (% GDP) -0.137** 0.054 -0.133 -0.200* -0.190* -0.222** (0.053) (0.139) (0.124) (0.107) (0.101) (0.099) s c , q s c , q s c , q s c , q s c , q Instruments - 0 0 , 1 0 , 1 , 2 0 , 1 , 2 , 3 0 , 1 , 2 , 3 , 4 Controls yes yes yes yes yes yes Observations 3,986 3,986 3,986 3,986 3,986 3,986 14/ 18
Strategy Data Resultats Discussion Additional evidence Pertes de capital: = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 15/ 18
Strategy Data Resultats Discussion Additional evidence Pertes de capital: = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 0 15/ 18
Strategy Data Resultats Discussion Additional evidence Pertes de capital: = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 0 . 35 15/ 18
Strategy Data Resultats Discussion Additional evidence Rigidités salariales: dL L = dK K = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 16/ 18
Strategy Data Resultats Discussion Additional evidence Rigidités salariales: dL L = dK K = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 . 10 16/ 18
Strategy Data Resultats Discussion Additional evidence Rigidités salariales: dL L = dK K = α Y + ( 1 − α ) dL + dA y K d d L A ���� � �� � � �� � ���� . 4 . 05 . 10 . 25 16/ 18
Strategy Data Resultats Discussion Additional evidence Complementarités sectorielles: �� 1 � 1 /ρ K ρ K = i di 0 Avec ρ = 1 / 2 et un choc concentré sur un secteur: � � ( 1 − α Y + ( 1 − α ) dL + dA K d d ) 1 /ρ − 1 = y L A ���� � �� � ���� . 4 � �� � . 10 Le résidu est petit mais ça ressemble à ce qu’on appelle de l’interruption d’activité... 17/ 18
Strategy Data Resultats Discussion Additional evidence Complementarités sectorielles: �� 1 � 1 /ρ K ρ K = i di 0 Avec ρ = 1 / 2 et un choc concentré sur un secteur: � � ( 1 − α Y + ( 1 − α ) dL + dA K d d ) 1 /ρ − 1 = y L A ���� � �� � ���� . 4 � �� � . 20 . 10 Le résidu est petit mais ça ressemble à ce qu’on appelle de l’interruption d’activité... 17/ 18
Strategy Data Resultats Discussion Additional evidence Complementarités sectorielles: �� 1 � 1 /ρ K ρ K = i di 0 Avec ρ = 1 / 2 et un choc concentré sur un secteur: � � ( 1 − α Y + ( 1 − α ) dL + dA K d d ) 1 /ρ − 1 = y L A ���� � �� � ���� . 4 � �� � . 20 . 10 . 10 Le résidu est petit mais ça ressemble à ce qu’on appelle de l’interruption d’activité... 17/ 18
Strategy Data Resultats Discussion Additional evidence 18/ 18
Merci! yzylberberg@crei.cat 19/ 18
Recommend
More recommend
