Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Global warming as an asymmetric public bad
Louis-Gaëtan Giraudet (Ecole des Ponts ParisTech, CIRED)
Céline Guivarch (Ecole des Ponts ParisTech, CIRED)
A Toxa – June 27, 2016
Global warming has asymmetric impacts
2
Identical damages ( d = - b )
Symmetric impacts ( d = b )
3
Vu
lner
abili
ty In
dex
(z-
sco
re)
Sou
rce:
Sam
son
et
al. (
20
11
)
Abatement Cost Parameter Source: Nordhaus (2015)
%: share of 2010 global GHG emissions Source: WRI (2014)
Adaptation and mitigation little correlate
Adaptation may exacerbate warming
Isaac & van Vuuren (2009)
• Growing concerns about the warming potential of air conditioning (Barrecca et al., 2016; Davis and Gertler, 2015; Auffhammer and Mansur, 2014).
• At least conceptually, agriculture could also be a concern (fertilizers, irrigation). 4
We extend the canonical dynamic game of global warming to capture three stylized facts 1. Asymmetric impacts (gainers & losers) 2. Correlations between adaptation and mitigation capabilities 3. Warming potential of adaptation (“free-driving”)
We examine the policy implications of those asymmetries for
• “Bottom-up” approaches: e.g., US-China Agreement, NDCs • Optimal solutions: GHG emission pricing
We contribute to opening up the public good model, besides • Kotchen’s “impure public goods” (2005) • Nordhaus’ “climate clubs” (2015)
5
Dynamic Games of Global Warming: Background
• Mitigation – Prevailing structure: Linear-quadratic in state, identical players. – Discussion of open-loop vs. feedback non-cooperative strategies (van der
Ploeg & de Zeeuw, 1992; Hoel, 1993) – With small enough discount rate, non-linear feedback can induce too
little warming (Dockner and Long, 1993). Restrictions subsequently placed on the result (Rubio and Casino, 2002; Wirl, 2007)
– Partial account of heterogeneity (Martin et al., 1993; Zagonari, 1998)
• Adaptation – Mitigation in period 1, adaptation in period 2 (Buob and Stephan, 2011;
Ingham et al., 2013) – Discussion of the private/public aspects of adaptation (Mendelsohn,
2000), but not of its implications for warming
Missing:
Full characterization of asymmetry of impacts Full integration of mitigation and (potentially emitting) adaptation
6
A 2-player Model of Space Heating and Cooling
7
cold region
warm region
Energy for heating (ec)
CO2 concentration
Atmospheric temperature ( T ) Climate
sensitivity Carbon intensity
Energy for cooling (ew)
Mitigation in the cold region (qc)
Carbon sink
Mitigation in the warm region (qw)
“Perfect adaptation”
“Perfect adaptation”
-
-
-
+
“Adaptive Mitigation”
,ce q T q bT
8
,we q T q dT
Controlled mitigation (energy/carbon efficiency)
“Perfect” adaptation to asymmetric, linear impacts
Energy use
Warming , ,c c w wT e q T e q T sT
2, 2i ii c w m q q
Natural sink
Technology
Retroaction
Taxonomy of Public Goods
9
damages (d)
benefits (b)
d = b
d = -b
Public bad
Public good
Asymmetric public bad
Heterogeneous public bad
0 Heterogeneous public good
Asymmetric public good
Focus on the Public Bad Corner ( d > | b | )
10
Usually studied: homogeneous & heterogeneous public bad
Our study: More general public bad (incl. asymmetric)
Optimization Problems
,,0
Minimize , e d
subject to , ,
c w
i i i rt
iq q
i c w
i i i i
e q T m q t
T e q T e q T sT
Energy expenditure
Mitigation cost
11
, , ,i i i i i i i i ie q T m q e q T e q T sT
, , ,i i i i i i i
i
e q T m q e q T e q T sT Cooperative
Non-coop., fwd-looking
,i i i ie q T m q Non-coop.,
myopic
Optimality Conditions
12
r b d s b d
Optimal mitigation
Co-state Dynamics
c c
w w
r b d s b
r b d s d
Cooperative (social optimum)
Forward-looking (Nash equilibrium)
Stable steady state d d 0 T
T T s d b
T is absent Open-loop and feedback coincide, ensuring subgame perfectness
d d 0 1i i i i iq q
d dir T
Co-state Variables at Steady State
13
d b
r s b d
0
0
c
w
b
r s b d
d
r s b d
Global public bad
(small) local public good
(large) local public bad
Social optimum (S)
Nash (N)
0 Myopic (M)
0
d b
Not shown here: Co-state variables (hence mitigation efforts) are STATIONARY
Heterogeneous Public Bad (canonical – both lose)
14
M
S
N
qw
Free-riding
Free-riding
Assumptions
0
c w
d b
qc
1:1
Asymmetric Public Bad ( = gainer & loser)
15
S
qw
qc
N
Free-riding
Free-driving
Assumptions
0
c w
d b
M
1:1
Five Regimes of Asymmetric Public Bad (d>b>0)
16
Global public good
Unstable steady state
v
i
ii
iii
iv
Mitigation ratio
Imp
act
rati
o d
/ b
w c
1
N S
d b
T T
21 1
N M
d b
1 s b
N M
d b
T T
Efficiency-Warming Tradeoffs
17
Glo
bal
War
min
g
Economic Efficiency
M
S
Incr
easi
ng
da
ma
ges
i
ii
iii
iv
v
Ni
Nii Niii
Niv
Nv
cheaper mitigation in WARM
cheaper mitigation
in COLD
Interpretation: Bilateral Agreements
18
Glo
bal
War
min
g
Economic Efficiency
M
S
Heterogeneous bads, e.g., China vs. USA
e.g., India vs. USA
e.g., India vs. Russia
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0.5 1 2
0.5 1 2 0.5 1 2 0.5 1 2
0.5 1 2 0.5 1 2
Nash Social optimum Myopia Social optimum Myopia Nash
Hicks-Kaldor: c gains, w loses
Pareto: c and w gain
Hicks-Kaldor: w gains, c loses
c gains, w loses
w gains, c loses
c and w lose
Het
ero
gen
. bad
(d
/-b
vs.
μ)
Asy
mm
etri
c b
ad (
d/b
vs.
μ)
Welfare Improvements
Homogeneous bad (canonical)
Optimal Pigouvian Prices (pki)
20
1
ii kk i
k
p
(Positive) externality
Negative externality
=>
Two externalities necessitate two prices
UNLESS
the players are myopic (λMi =0) or identical (λM
c = λMw)
Conditions under which a dynamic public bad can be under-supplied
Restrictions on Pareto improvements (esp. )
A case for differentiated emission prices, which can include (small) subsidies for GHG-intensive adaptation
21
Taking into account asymmetries in adaptation and mitigation, we find
M N
Discussion
• Climate change restricted to smooth warming – No sea level rise, ocean acidification, etc. – No stochastic events, catastrophes, etc.
• Linear-state structure
– Pertinent at the global scale, less at the local one (e.g., Dell et al., 2014) – Flexibility of open loop & strength of feedback (subgame perfectness)
• Specific technology
– Emissions induced by adaptation: what potential? – No accumulation of mitigation capital
Our model is therefore most relevant to Large countries (China + USA ~ 40% of global emissions) Sectors such as buildings, agriculture (resp. 6% and 24% of emissions) Short-term, moderate impacts
22
23
Extension: Non-linear Adaptation
24
2
i ia T T T
ia T wa T ca T
cTwT
Ad
apta
tio
n C
ost
Temperature
cw
Asym. PG Asym. PB Heterogeneous Public Good Heterogeneous Public Bad
2
with
i i
i i
i i
T Ta T T T
T T
(Burke et al., 2015) 13 CT
Quadratic-State Model
,i ie q T q a T
25
Energy use
Warming , ,c c w wT e q T e q T sT
2, / 2i ii c w m q q Technology
Adaptation 2
i ia T T T
Non-linear payoff function Open-loop and feedback no longer coincide Non-linear transition equation Numerical resolution (Kossioris et al., 2008)
I’ll emphasize a graphical exposition 26
27