Life Expectancy and the Environment (In co-authorship with Fabio Mariani (UCL) and Natacha Raffin (Paris 1)) SALISES 11th Annual Conference Port of Spain Trinidad and Tobago March 24-26, 2010

Agust´ın P´ erez-Barahona [email protected] http://sites.google.com/site/agustinperezbarahona/ ´ cole Polytechnique (France) INRA and E

Motivation: 1.- Two-way causality between life expectancy and environmental quality: (John and Pecchenino (1994), Ono and Maeda (2001) and Jouvet et al. (2007)). 2.- Stylized facts: Environmental quality : Environmental Performance Index (EPI) (YCELP (2006)). Life expectancy : Life expectancy at birth (2005) (UN (2007)). Stylized 1 (SF1): Environmental quality is positively correlated with life expectancy. Correlation coef. 0.66 (statistically significant at the 1% level). correlation JPN

60

SDN

50

MRT

MLI ETH NER

BFA

TCD AGO

GHA

KHM GMB MDG SEN PNG LAO TGO GIN COG HTI

corr.coeff = 0.66

40

Life expectancy

70

80

ISL CHE AUS ITA CAN SWE ISR ESP NORFRA AUT NZL BEL CYP DEU FIN NLD GRC CRIIRL GBR ARE CHL CUB USA PRT DNK KOR SVN CZE MEX PAN POL ARG ECU SVK LKA OMN ALB SYR TUN MYS VENHUN COL LBN SAU BGR CHN ARMJOR ROU PRY JAM BRA GEOSLV DZA VNM PHL IRN TTO PER THA EGY MAR NIC SUR TUR HND MDA GTM DOM KGZ UZB AZE IDN UKR RUS MNG BOL TJK KAZ IND PAK BGD TKM NPL YEM MMR

GNB NGA COD MOZ

GAB

BEN

ZAF NAM KEN UGA CIV CMR TZA RWA

BDI LBR SLE

MWI CAF ZMB

ZWE

SWZ

30

40

50

60

70

EPI

1

0.03 Density

0.00

0.000

0.005

0.01

0.02

0.020 0.015 0.010

Density

0.025

0.04

0.030

0.035

Stylized fact 2 (SF2): Environmental quality and life expectancy are bimodally distributed. Hartigan’s dip test (Hartigan and Hartigan (1985)).

30

40

50 EPI (2006) bandwidth=3.9

60

70

80

30

40

50

60

70

80

Life expectancy (2005) bandwidth=4.3

Environmental trap: low life expectancy / low environmental quality. Examples of environmental traps: linked with a reduction in life expectancy. • Africa or ex-USSR: McMichael et al. (2004) and Patz et al. (2005). • Easter Island: Diamond (2005) and de la Croix and Dottori (2008). 2

Benchmark model: theoretical framework (OLG) that allows us to reproduce SF1 and SF2. a) Two-way causality: • Environmental conditions affect the survival probability: endogenous life expectancy (π(et )). • Agents invest in environmental quality depending on how much they expect to live. b) Environmental trap: • Relationship between environmental quality and life expectancy: threshold effects (convex-concave π(et )).

ecological

– Natural sciences: Cakmak et al. (1999) and Scheffer et al. (2001). – Economics: Baland and Platteau (1996) and Dasgupta and M¨ aler (2003). • Multiple steady state equilibria: initial conditions do matter. 3

1.- Preferences: U (ct , et+1 ) = ln ct + πt γ ln et+1 , with γ > 0

(1)

• πt = π(et ), such that π 0 (et ) > 0, limet →0 π(et ) = π, limet →∞ π(et ) = π and 0 < π < π ≤ 1 (medical and epidemiological literature). 2.- Budget constraint: wt is assumed to be exogenous (benchmark model). wt = ct + mt

(2)

3.- Law of motion of environmental quality: (John and Pecchenino (1994) and Ono (2002)). et+1 = (1 − η)et + σmt − βct − λQt

(3)

• β, σ, λ > 0 and 0 < η < 1. λQt > 0 (< 0): total impact of a harmful (beneficial) external effect (exogenous). et : inherited environmental quality. • Agents cannot improve their own survival probability (crucial role of early life health): Blackburn and Cipriani (2002), de la Croix and Licandro (2007), and Case and Paxson (2008).

4

4.- Agents problem: maximize (1) subject to (2), (3), ct > 0, mt > 0, et > 0, and wt and et are taken as given ⇒ FOC. 5.- Dynamics: FOC into (3) (for the sake of simplicity: wt = w and Qt = Q). et+1 =

γπ(et ) [(1 − η)et + σw − λQ] ≡ φ(et ) 1 + γπ(et )

(6)

Graphical example: et+1 φ(et )

0

e∗L

˜e

e∗H

et

5

6.- Environmental trap: A convex-concave φ(et ) might be generated by a convexconcave π(et ) (Blackburn and Cipriani (2002)). Analytical illustration: long term exposure to bad environmental conditions (Evans and Smith (2005).  π if et < ˜ e π(et ) = (7) π if et ≥ ˜ e

et+1

B0

φ(et ) = et B

A

0

e∗L ˜ e

e∗H

et

6

Main results: two-way causality between life exp. and env. quality + SF1 + SF2. 1.- Escaping the trap (1/2): medical technological progress: improvement in medicine effectiveness (de la Croix and Sommacal (2008)). – European Environment and Health Action Plan 2004-2010. 2.- Escaping the trap (2/2): reduction of harmful external effects on the environment (oil slicks, Chernobyl disaster, etc). – Implementation of international agreements that promotes a worldwide reduction of pollution: Kyoto Protocol. 3.- Welfare analysis: environmental taxes: – Inter-generational externality: external effects of agent choices on future generations ⇒ (decentralized) economy is under-investing in environmental maintenance. – Environmental taxes ⇒ optimal allocation (social planner solution). – Social planner solution may imply the elimination of the environmental trap.

7

Conclusions: 1.- A model with: two-way causality between the life expectancy and the environmental quality + multiple equilibria. 2.- Environmental trap: convex-concave π(e) • Benchmark model: low life expectancy / poor environmental quality. • Model with physical or human (education) capital accumulation: de la Croix and Doepke (2003, 2004). – Human capital: inter-generational altruism. – Low life expectancy / poor environmental quality / low physical or human capital. Extentions: 1.- Heterogeneous agents: environmental tastes. 2.- Demography: endogenous fertility and environmental quality. 3.- Fiscal policy: environmental taxes. 8