American Climate Prospectus - Stanford University · CCI/IA | American Climate Prospectus: Economic Risks in the United States 3 An Independent Assessment for a Climate Risk Committee

Resources for the Future | American Climate Prospectus: Economic Risks in the United States ‹#›

10 East 40th Street, Suite 3601, New York, NY 10016 Tel: +1.212.532.1158 | Fax: +1.212.532.1162 | Web: www.rhgroup.netwww.climateprospectus.org

American Climate ProspectusEconomic Risks in the United States

Climate Change Impacts/Integrated Assessment | Snowmass, CO | July 22, 2014 (slides updated August 16, 2014)

Trevor HouserRhodium Group

Robert KoppRutgers University

Solomon HsiangUC Berkeley

Roger Muir-WoodRMS

Kate LarsenRhodium Group

DJ RasmussenRhodium Group

Amir JinaColumbia University

Paul WilsonRMS

Michael Delgado Rhodium Group

Michael MastrandreaStanford University

James RisingColumbia University

Shashank MohanRhodium Group

Prepared for the Risky Business Project

CCI/IA | American Climate Prospectus: Economic Risks in the United States 2

Overview


An Independent Assessment for a Climate Risk Committee

“If you can’t measure it, you can’t manage it”… - Mike Bloomberg

Analytical Support for the Risky Business Project (riskybusiness.org)

http://riskybusiness.org/about/cochairs/michael-bloomberg

http://riskybusiness.org/about/cochairs/hank-paulson

http://riskybusiness.org/about/cochairs/tom-steyer

http://riskybusiness.org/about/risk-committee/henry-cisneros

http://riskybusiness.org/about/risk-committee/gregory-page

http://riskybusiness.org/about/risk-committee/robert-rubin

http://riskybusiness.org/about/risk-committee/olympia-snowe

http://riskybusiness.org/about/risk-committee/donna-shalala

http://riskybusiness.org/about/risk-committee/george-shultz

http://riskybusiness.org/about/risk-committee/al-sommer

http://riskybusiness.org


Research Objectives

1. Where possible, micro-found the damage function for the United States using the highest-quality, identified, empirical measurements

2. Make the calculation transparent (and hopefully open source) 3. Make updating the calculation easy 4. Leverage state-of-the-art physical climate models to describe

the spatial structure of impacts 5. Quantify uncertainty and risk 6. Make the results interpretable and relatable to ordinary

citizens 7. Finish in 12 months


Scope of coverageFar from comprehensive


Research approach

Downscaled, probabilistic Physical Climate Projections

Impact estimates based on meta-analysis of econometric research

Complementary detailed sectoral models

Integrated Economic Analysis with CGE model, consideration of

potential adaptations

Spatial Empirical Adaptive Global-to-Local Assessment System (SEAGLAS)


Research approach

Downscaled, probabilistic Physical Climate Projections

Impact estimates based on meta-analysis of econometric research

Complementary detailed sectoral models

Integrated Economic Analysis with CGE model, consideration of

potential adaptations

Spatial Empirical Adaptive Global-to-Local Assessment System (SEAGLAS)

Note: We did not try to construct a full IAM. In particular, we made the simplifying assumption of a constant spatial and sectoral economic structure


Physical Science Projections


Multiple emissions pathways reflect different socio-economic, technological and policy futures.

200

300

400

500

600

700

800

900

1000

1775

1800

1825

1850

1875

1900

1925

1950

1975

2000

2025

2050

2075

2100

RCP8.5 RCP6.0 RCP4.5 RCP2.6 Historical

Sources of uncertainty!!

Socio-economic/Emissions!Global climate response!Regional climate response!Natural variability!Tipping points & the unknown

(“Business as Usual: Our Current Path”)

(“Small emissions reduction”)

(“Medium emissions reduction”)

(“Large emissions reduction”)

part

s pe

r m

illio

n ca

rbon

dio

xide

con

cent

ratio

ns


The global temperature response to greenhouse gas forcing is uncertain.



Temperature projections (°F) from the MAGICC simple climate model, courtesy Malte Meinshausen

-2

0

2

4

6

8

10

12

1900 1925 1950 1975 2000 2025 2050 2075 2100

RCP 8.5

RCP 6.0

RCP 4.5

RCP 2.6

Historical (10-year avg.)

Historical (annual)

-2

0

2

4

6

8

10

12

2020-2039 2040-2059 2080-2099

95%

5%

Likely Range

2080-2099


As is the regional response that accompanies global warming…

Ghosted = pattern-scaled modern surrogateNote: models/surrogates are weighted so as to match PDF, not evenly.



Probability distribution developed from simple climate model and downscaled global climate model projections with the surrogate/model mixed ensemble method of Rasmussen & Kopp (in prep.).

Used BCSD downscaled projections from the Bureau of Reclamation.


which gives rise to uncertain projections of temperature change…



Median and 1-in-20 chance summer temperature projections (°F), RCP 8.5 (high emissions)


…and of precipitation change. Sources of uncertainty!!


Median projected % precipitation change, RCP 8.5 (high emissions) in 2080-2099.!!In the faded regions, an increase and an decrease are both about equally likely.


Natural variability occurs at time scales ranging from the daily to the decadal.



2000 2020 2040 2060 2080 2100

−20

24

68

GFDL−CM3, New York City; RCP 8.5

Year

JJA

avg.

tem

pera

ture

ano

mal

y (C

)

●

●●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●●

●

●●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●●

●

●

●●

2090 2092 2094 2096 2098 2100

515

2535


Year

mon

thly

avg

. tem

pera

ture

(C)

●

●

●

●

●●

● ●

●●

●

●

●

●●

●

●

● ●

●●

●

● ●

● ●

●

●

● ●

●

Jul 03 Jul 10 Jul 17 Jul 24 Jul 31

2530

35


July 2090

daily

avg

. tem

pera

ture

(C)

avg. daily realization

monthly avg: 30.3 C

●

●

●

●

●

● ●●

●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

● ●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●●

●

●

● ● ●

●●

●●

●

●

● ●

● ●

●

●

●

●

●

●

●

●

●

●

●

● ●●

●●

●

● ● ●

●

●

●

●●

●

●●

●

●●

●

●

●

●●

● ●●

●

●

●●

●

●

●

●

●

●

● ●

●

●

●● ●

●

●

●

● ●● ●

● ●

●

●

●

●

●

●

●

● ●●

●●

●

● ●

● ●●

●

● ●

●●

● ●

●

●

●●

●●●

●

●●

●●

●

●

●

●

●

●

●

●●

● ●

●

●●

●●

● ●

●

● ●

●

●

●

●

●

●

● ●

●

● ●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

● ●

●

● ●●

●

●

● ●

●

●●

●

●

●

●

●

●●

●● ● ●

●

●●

●

●

●

● ●

●●

●

●

●

●

●

● ●

●

● ●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

● ●

●

● ●●

●●

●

●

●

● ●●

●●

●●

●●

● ● ●●

●● ●

●●

●

● ●

● ● ● ●●


Meta-analysis of Econometric Research

CCI/IA | American Climate Prospectus: Economic Risks in the United States

Distributed Meta-Analysis System

15


Distributed Meta-Analysis System



1. Nationally representative

2. Analyze recent time-periods in US history

3. Robust to unobserved factors that differ across spatial units (jurisdictions, counties, or states)

4. Identify responses to high-frequency climatic variables (days or weeks)

5. Identify responses to the full distribution of temperature and rainfall measures

6. Account for temporal displacement

7. Account for seasonal patterns and trends in the outcomes

8. Ecologically valid (not laboratory studies)

Our criteria for inclusion


• Agriculture (maize, soy, cotton, wheat)!• Schlenker and Roberts (PNAS, 2009) - Cotton, Soy, Maize!• Hsiang et al. (2013) - Wheat!• McGrath and Lobell (ERL, 2013) - Carbon fertilization!• Fisher et al. (AER, 2012) - Crop storage!

• Labor productivity (extensive margin only) • Graff Zivin and Neidell (JLE, 2014)!

• Heat- and cold-related mortality (age resolved) • Deschenes and Greenstone (AEJ, 2011)!• Barreca et al. (2013)!

• Crime (violent and property) • Jacob et al. (JHR, 2007)!• Ranson (JEEM, 2014)

• Electricity demand (residential) • Auffhammer and Aroonruengsawat (CC, 2011)

19

Sectors where we apply this approach


Impact functions – agriculture

0 10 20 30 40

−0.08

−0.06

−0.04

−0.02

0lo

g(yie

ld)

0 0.5 1 1.5−1

−0.5

0

0.5

1

Precipitation (m)

log(

yield

)

0 10 20 30 40

−0.08

−0.06

−0.04

−0.02

0

log(

yield

)

0 0.5 1 1.5

−0.5

0

0.5

1

Precipitation (m)

log(

yield

)

0 0.5 1 1.5

−0.5

0

0.5

1

Precipitation (m)

log(

yield

)

0 10 20 30 40−0.06

−0.04

−0.02

0

log(

yield

)

0 0.5 1 1.5

0

0.5

1

Precipitation (m)

log(

yield

)

0 10 20 30 40

−0.04

−0.02

0

0.02

log(

yield

)

0 0.5 1 1.5

−0.5

0

0.5

Precipitation (m)

log(

yield

)

0 10 20 30 40

1

2

3

log(

yield

)

−2 0 2 4 6 8 100

0.1

0.2

0.3

0.4

Yield (% change)6 8 10 12 14 16

0

0.1

0.2

0.3

0.4

Yield (% change)6 8 10 12 14 16 18

0

0.1

0.2

0.3

0.4

Yield (% change)6 8 10 12 14 16 18

0

0.1

0.2

0.3

0.4

Yield (% change)

0 10 20 30 40−0.06

−0.04

−0.02

0

Maize vs. TDD (East) Maize vs. P (East) Maize vs. TDD (West) Maize vs. P (West)

Soy vs. P (East) Soy vs. TDD (West) Soy vs. P (West)

Cotton vs. TDD Cotton vs. P Wheat vs. TAVG

Maize vs. 100ppm CO2 Soy vs. 100ppm CO2 Cotton vs. 100ppm CO2 Wheat vs. 100ppm CO2

Soy vs. TDD (East)

Temperature (oC)

Temperature (oC) Temperature (oC)

Temperature (oC)


log(

yield

)


Impact functions – other sectors

−10 0 10 20 30 40

−1.5

−1

−0.5

0

Crim

e ra

te (%

chan

ge)

0 5 10 15 20 25 30

0

0.2

0.4

Precipitation (mm)Cr

ime

rate

(% ch

ange

)

−10 0 10 20 30 40

−1

−0.5

0

0.5

Crim

e ra

te (%

chan

ge)

0 5 10 15 20 25 30−0.3

−0.2

−0.1

0

0.1

Precipitation (mm)

Crim

e ra

te (%

chan

ge)

0 10 20 30

−100

−50

0

50

Min

utes

0 10 20 30

−60

−40

−20

0

20

Min

utes

−10 0 10 20 30−1

0

1

2M

orta

lity r

ate

(% ch

ange

)

−10 0 10 20 30

−5

0

5

10

15

Mor

talit

y rat

e (%

chan

ge)

−10 0 10 20 30

−1

0

1

Temperature (oC)

Mor

talit

y rat

e (%

chan

ge)

−10 0 10 20 30

0

5

10

Mor

talit

y rat

e (%

chan

ge)

−10 0 10 20 30−5

0

5

10

15

x10 −4

Mor

talit

y rat

e (%

chan

ge)

Property Crime vs. TMAX Property Crime vs. P Violent Crime vs. TMAX Violent Crime vs. P

High−risk Labor vs. TMAX Low−risk Labor vs. TMAX

Mortality vs. TAVG

Mortality (0−1) vs. TAVG Mortality (1−44) vs. TAVG

Mortality (45−64) vs. TAVG Mortality (65+) vs. TAVG

x10 −3

x10 −4

x10 −4

x10 −3


Temperature (oC) Temperature ( C)


Temperature (oC)


Direct impacts


Agriculture example: Dose-response

23

−2 0 2 4 6 8 100

0.1

0.2

0.3

0.4

Yield (% change)

Maize vs. 100ppm CO2


Agriculture example: Median in RCP 8.5

24

(assuming current economy)


Agriculture example: Expected number of events

25


Distribution of impacts: RCP 8.5, median

Figure 8.3: CRCP 8.5 med

largest incpeople) becthese regioclimate inexperienceminimizinRockies, Anorthern M

Climate impact oian

reases in mortcause the numbons increases n these locatie many letg the gains frppalachia, Nor

Midwest and N

on heat and cold

tality rates (deber of high temsubstantially, ions is too wthally cold rom warming. rthwest, northortheast (excep

-related mortalit

aths per 100,0mperature days

but the baseliwarm for the

days initialIn contrast, t

hern Great Plainpt New York C

ty

00 s in ine em lly, the ns, ity

and Nerates fogains frlosses dwe compopulatthese gnationa

ew Jersey) expeor the medianrom a smaller during additio

mpute changes tion growth fr

gross national al aggregate m

erience net redn RCP 8.5 pronumber of col

onal hot days. in total morta

rom the 2010 Cpatterns trans

mortality (Figur

HEALTH

ductions in moojection becauld days outweiFor a sense of

ality assuming Census and se

slate into chanre 8.3, right col

H 66

ortality use the

gh the f scale, the no

ee how nges of lumn).

Figure 8.3: CRCP 8.5 med

largest incpeople) becthese regioclimate inexperienceminimizinRockies, Anorthern M

Climate impact oian

reases in mortcause the numbons increases n these locatie many letg the gains frppalachia, Nor

Midwest and N

on heat and cold

tality rates (deber of high temsubstantially, ions is too wthally cold rom warming. rthwest, northortheast (excep

-related mortalit

aths per 100,0mperature days

but the baseliwarm for the

days initialIn contrast, t

hern Great Plainpt New York C

ty

00 s in ine em lly, the ns, ity

and Nerates fogains frlosses dwe compopulatthese gnationa

ew Jersey) expeor the medianrom a smaller during additio

mpute changes tion growth fr

gross national al aggregate m

erience net redn RCP 8.5 pronumber of col

onal hot days. in total morta

rom the 2010 Cpatterns trans

mortality (Figur

HEALTH

ductions in moojection becauld days outweiFor a sense of

ality assuming Census and se

slate into chanre 8.3, right col

H 66

ortality use the

gh the f scale, the no

ee how nges of lumn).


Coastal impacts

27


Energy demand

28

% increase in annual residential + commercial energy expenditures

Impact function calibrated against RHG–

81 AMERI

Figure 10.2: Percent

demand of212 GW by

While mosthe time (das well as consumers

CAN CLIMATE PRO

National Chang

f 8 to 95 gigaw2040-2059 and

st of this capacduring peak dem

operating coss. The resulting

OSPECTUS

e in Electricity D

watts (GW) by d 223 to 532 GW

city would onlymand periods)sts are passed g electricity pri

Demand

2020-2030, 73 W by 2080-2099

y operate part , the capital coon to electric

ice increases le

to .

of sts ity

ead

to a likeand com2.1 to 7(FigureSoutheincreas

ely 0.1 to 2.9% mmercial ener7.3% by 2040-e 10.4). The east, Great Plase in 2080 -209

increase in totrgy costs on av2059, and 8 togreatest incre

ains and South99 of 12 to 28%

tal annual residverage by 2020o 22% by 2080eases occur ihwest, with a

%, 9 to 30%, an

dential 0-2039, 0-2099 in the a likely d 11 to


National aggregate impacts


Total cost and sectoral breakdown differ by region

RCP 8.5, median case, 2080-2099

-6% -4% -2% 0% 2% 4% 6% 8% 10% 12% 14%

USA

Connecticut Delaware

Maine Maryland

Massachusetts New Hampshire

New Jersey New York

Pennsylvania Rhode Island

Vermont West Virginia

Alabama Arkansas

Florida Georgia

Kentucky Louisiana

Mississippi North Carolina South Carolina

Tennessee Virginia

Illinois Indiana

Iowa Michigan

Minnesota Missouri

Ohio Wisconsin

Kansas Montana Nebraska

North Dakota Oklahoma

South Dakota Texas

Wyoming

Arizona California Colorado

Nevada New Mexico

Utah

Idaho Oregon

Washington

Coastal (Historical Activity)

Agriculture

Labor

Energy

Crime

Mortality (Market)

Mortality (VSL)

NORT

HEAS

T SO

UTHE

AST

MID

WES

T GR

EAT P

LAIN

S SO

UTHW

EST

NW


Direct costs and benefits% of GDP, RCP 8.5, 2080-2099

Figure 13.23Percent

At a state lvaries grerelatively hagriculturathe Midwewhole. Laband relativwhich arenegative fo

The combilate centuhistorical hwith a 1-inthan 4.8%costs consiThe likely routput, anA number likely combnational avstates likely

3: Direct costs an

level, the relateatly (Figure 1

high in the Nal impacts playest and Great Pbor productivitvely evenly s

e large, are por others.

ined likely direury, using ma

hurricane actin-20 chance of. The Southeaiderably higherange for Florid for Mississip

of Midwest abined impactsverage, while my see relatively

nd benefits unde

tive importanc13.24). Coasta

Northeast and Sy a considerab

Plains than for ty costs are mepread, while

positive for s

ect cost of thesarket costs foivity, is 0.8% f costs less thaast sees likely er than the couida is 3.2% to 8pi is 2.7% to 8.7nd Great Plain considerably most Northeaslittle direct cos

er RCP 8.5 as a s

ce of each impal damages raSoutheast, wh

bly larger role the country a

eaningfully-sizmortality cos

some states a

se six impacts r mortality ato 3.3% of GD

an 0.4% or mocombined direunty as a who

8.9% of econom7% (Figure 13.2ns states also s

higher than tst and Northwests.

share of GDP, 20

act ank hile

in s a

zed sts, and

in and DP, ore ect

ole. mic 25). see the est

At the hurricageograpsame, increascosts riand Miand 6.4South athis mproject

Finallyapproacthe numviewed and benunquanreport,of ‘dee(Weitzmin Chap

080-2099

upper bound ane activity anphic distributiobut the mag

ses (Figure 13.2ise to 1.4% to 5ississippi likely

4% to 19%. Theand in the Nor

measure, withtions for eight N

y, note that becch and many kmbers present

d as a comprehenefits (Pindyckntified impacts while some o

ep’ structural man, 2009; Hepter 15.

DIRECT CO

of our estimand the VSL for

on of damages gnitude of the

26). National .7% of econom

y direct costs re difference betrthwest and Noh net benefitNorthwest and

cause we are takknown impactted in this chaensive portrait k, 2013; Stern, s are discusse

of methods for uncertainties

eal and Millne

OSTS AND BENEFI

ates, using promortality costremains rough

e likely directaverage likely

mic output. In Frise to 10.1% ttween impacts ortheast growsts in the m

d Northeast stat

king an enumes are not quanapter should nof all economi2014). Many of

ed in Part 4 oevaluating thein climate im

r 2013) are disc

TS 116

ojected ts, the hly the t costs

direct Florida to 24%

in the s using

median tes.

erative ntified, not be ic costs f these of this e costs

mpacts cussed


Comparison to benefit-cost integrated assessment models

!1.00%&

!0.80%&

!0.60%&

!0.40%&

!0.20%&

0.00%&

0.20%&

0.40%&

FUND&3.8&(direct)&ENVISAGE&(CGE)&ACP&(direct,&2040!2059)&

U.S.$costs$at$~2.2°C$above$pre3industrial$

Ecosystems/SeHlements&

Water&

Tourism&

Crime&

Health&(VSL)&

Labor/Health&

Coastal&

Energy&Demand&

Agriculture&

Total&

Total&ACP&sectors&


Comparison to benefit-cost integrated assessment models

!3.00%&

!2.50%&

!2.00%&

!1.50%&

!1.00%&

!0.50%&

0.00%&

0.50%&

1.00%&

FUND&3.8&(direct)&ENVISAGE&(CGE)&ACP&(direct,&2080!2099)&

U.S.$costs$at$~4.1°C$above$pre4industrial$

Ecosystems/SeFlements&

Water&

Tourism&

Crime&

Health&(VSL)&

Labor/Health&

Coastal&

Energy&Demand&

Agriculture&

Total&

Total&ACP&sectors&


Considering mitigation and adaptation


Large mitigation benefit for mortality

35

-30

-20

-10

0

10

20

30

40

50

60

70

80

8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6

1-in-20

1-in-20

Likely Range

NORTHEAST SOUTHEAST MIDWEST GREAT PLAINS SOUTHWEST NORTHWEST USA

Change in temperature-related deaths per 100,000 individuals, 2080-2099


Smaller mitigation benefit for coastal impacts in 21st century

36

Changes in average annual hurricane and coastal flood damage, 2080-2099Billion 2011 USD

$0

$5

$10

$15

$20

$25

$30

$35

8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6 8.5 4.5 2.6

1-in-20

1-in-20

Likely Range

NORTHEAST SOUTHEAST MIDWEST GREAT PLAINS SOUTHWEST NORTHWEST


Thought experiment: benefits of adaptation

37

What if T/P-yield relationship in the East evolves toward that in the West?


Market adaptation via general equilibrium effects

38

-2%

-1%

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

Dir

ect

RH

G-M

US

E

1-in-20

1-in-20

Likely

Range

NORTHEAST SOUTHEAST MIDWEST GREAT PLAINS SOUTHWEST NORTHWEST USA


Uncertainty and inequality


Sources of uncertainty

model weather statistical interaction

0 .5 1 1.5 2

Variance (% baseline)

violent

property

CrimeC

0 .2 .4 .6 .8


low risk

high risk

LaborD

0 200 400 600 800 1,000


CO2

ferti

lizat

ion

no C

O2

ferti

lizat

ion

oilcrop

grains

cotton

oilcrop

grains

cotton

AgricultureA B

0 100 200 300 400

Variance (deaths/100k)

all ages

age >64

age 45-64

age 1-44

age <1

Mortality


Death is the most unequal impactRCP 8.5 median, 2080-2099, valued at VSL

35 40 45 50 55 60 65 70 75 80 85

−3500

−3000

−2500

−2000

−1500

−1000

−500

0

500

1000

1500

ALAR AZ

CA

CO CT

DE

FL

GA

IA

ID

ILIN

KSKY

LA

MA

MD

ME

MI MN

MO

MS

MT

NC

ND

NE

NH

NJ

NM

NV

NY

OH

OK

OR

PA

RI

SC

SD

TN

TX

UT

VA

VT

WAWI

WV

WY

State income per capita (thousand dollars)

Media

n m

ort

alit

y gain

at V

SL (

dolla

rs) Average of 10 wealthiest

states: $75 gain/capita (-1.1 deaths per 100,000) !Average of 10 poorest states: $1900 loss/capita (27 deaths per 100,000)


Valuing risk and inequality

VALUING RISK AND INEQUALITY 126

change; they may be smaller if individuals who are initially richer are harmed relatively more.

Unlike the risk premium, it is possible for the the inequality premium to be negative if climate change reduces initial wealth disparities, which can happen if climate change imposes sufficiently larger damages on wealthy populations than on poorer populations. In this case, the unequal distribution of climate impacts would lower the perceived cost of those impacts relative their expected value. Thus it is not obvious ex ante that accounting for inequality aversion will necessarily increase the perceived cost of climate change.

We note that we do not resolve differences in damage across counties within a state—accounting for such differences would likely increase the inequality premium—although cross-state impacts tend to be more unequal than cross-county impacts within each state. We also do not account the distributional impacts of climate change with a state by income or demographic group, which are likely more important than differences across counties.

For both agriculture and mortality, the inequality premium can be significant. Strong inequality aversion alone can increase the value of agriculture losses by up to 40% (see first column of Table 15.1, where RRA=0), although the macroeconomic effects described in the preceding chapter dampen the inequality of direct agricultural impacts to some extent.

Strong inequality aversion can more than double the value of mortality losses, adding a 150% premium for an IA of 10 (see first column of Table 15.2, where RRA=0). The large magnitude of the inequality premium for mortality arises because the mortality increase is highest in some of the nation’s poorest states and least in some of the richest. Among the poorest ten states, the per capita median mortality cost is $1,900 per person under RCP 8.5 in 2080-2099 (with losses exceeding $2,000 in Florida, Mississippi, Alabama and Arkansas); among the ten richest states, the average is a gain of $75 per person (with gains exceeding $500 in North Dakota, Wyoming, Massachusetts and Minnesota).

PUTTING IT TOGETHER

Above, we separately analyzed the risk and inequality premiums for two types of impacts. However, we can combine both risk aversion and inequality aversion to compute an inequality-neutral, certainty equivalent damage (see Technical Appendix). This value is the hypothetical cost that, if shared equally among all

individuals with certainty, would have the same welfare impact as the actual unequal distribution of state-specific risks. The combined inequality-risk premium is the difference between this hypothetical cost and expected damages; it is the cost of having unequal economic risks imposed by climate change. Calculating this premium helps us conceptualize how inequality in expected losses, inequality in the uncertainty of losses, and inequality in baseline income together increase the perceived value of climate change damages.

Table 15.1: Combined inequality-risk premiums for agricultural impacts, 2080-2099 RCP 8.5, Premium as percentage of expected losses for maize, wheat, cotton, and soy output

!! !! !! ""#!

!$%%!&'(!)*+,!-./)+*'0! 1*23!)*+,!-./)+*'0!%%4!

!! !!!! 5! 6! 7! 8! 9! :5!

;<#

!

1*2

3!*

0/

=>

-?*

@A!

@'?/

)-0B/!%

%4! 5! 5C! 5C! :C! :C! 6C! 6C!

6! :DC! :DC! :7C! :7C! :EC! :EC!

7! 6:C! 6:C! 66C! 66C! 6DC! 6DC!

$%%

!&'

(!@

'?/

)-0

B/

!F'

)!

*0/

=>

-?*

@A!

8! 68C! 6GC! 6GC! 69C! 69C! 6HC!

9! D6C! DDC! DDC! D7C! D7C! DEC!

:5! 75C! 7:C! 7:C! 76C! 76C! 76C!

Table 15.2: Combined inequality-risk premiums for mortality impacts, 2080-2099 RCP 8.5, Premium as percentage of expected losses, applying value of a statistical life

!! !! !! ""#!

!$%%!&'(!)*+,!-./)+*'0! 1*23!)*+,!-./)+*'0!%%4!

!! !!!! 5! 6! 7! 8! 9! :5!

;<#

!

1*2

3!*

0/

=>

-?*

@A!

@'?/

)-0B/!%

%4! 5! 5C! DC! 8C! :5C! :7C! :9C!

6! 69C! D6C! D8C! 7:C! 78C! E6C!

7! E9C! 87C! G5C! G8C! 9DC! H:C!

$%%

!&'

(!@

'?/

)-0

B/

!F'

)!

*0/

=>

-?*

@A!

8! H5C! HGC!:57

C!

::D

C!

:6:

C!

:D:

C!

9!:6:

C!

:6H

C!

:D9

C!

:79

C!

:EH

C!

:G5

C!

:5!:E5

C!

:EH

C!

:G5

C!

:9:

C!

:HD

C!

658

C!

VALUING RISK AND INEQUALITY 126

change; they may be smaller if individuals who are initially richer are harmed relatively more.

Unlike the risk premium, it is possible for the the inequality premium to be negative if climate change reduces initial wealth disparities, which can happen if climate change imposes sufficiently larger damages on wealthy populations than on poorer populations. In this case, the unequal distribution of climate impacts would lower the perceived cost of those impacts relative their expected value. Thus it is not obvious ex ante that accounting for inequality aversion will necessarily increase the perceived cost of climate change.

We note that we do not resolve differences in damage across counties within a state—accounting for such differences would likely increase the inequality premium—although cross-state impacts tend to be more unequal than cross-county impacts within each state. We also do not account the distributional impacts of climate change with a state by income or demographic group, which are likely more important than differences across counties.

For both agriculture and mortality, the inequality premium can be significant. Strong inequality aversion alone can increase the value of agriculture losses by up to 40% (see first column of Table 15.1, where RRA=0), although the macroeconomic effects described in the preceding chapter dampen the inequality of direct agricultural impacts to some extent.

Strong inequality aversion can more than double the value of mortality losses, adding a 150% premium for an IA of 10 (see first column of Table 15.2, where RRA=0). The large magnitude of the inequality premium for mortality arises because the mortality increase is highest in some of the nation’s poorest states and least in some of the richest. Among the poorest ten states, the per capita median mortality cost is $1,900 per person under RCP 8.5 in 2080-2099 (with losses exceeding $2,000 in Florida, Mississippi, Alabama and Arkansas); among the ten richest states, the average is a gain of $75 per person (with gains exceeding $500 in North Dakota, Wyoming, Massachusetts and Minnesota).

PUTTING IT TOGETHER

Above, we separately analyzed the risk and inequality premiums for two types of impacts. However, we can combine both risk aversion and inequality aversion to compute an inequality-neutral, certainty equivalent damage (see Technical Appendix). This value is the hypothetical cost that, if shared equally among all

individuals with certainty, would have the same welfare impact as the actual unequal distribution of state-specific risks. The combined inequality-risk premium is the difference between this hypothetical cost and expected damages; it is the cost of having unequal economic risks imposed by climate change. Calculating this premium helps us conceptualize how inequality in expected losses, inequality in the uncertainty of losses, and inequality in baseline income together increase the perceived value of climate change damages.

Table 15.1: Combined inequality-risk premiums for agricultural impacts, 2080-2099 RCP 8.5, Premium as percentage of expected losses for maize, wheat, cotton, and soy output

!! !! !! ""#!

!$%%!&'(!)*+,!-./)+*'0! 1*23!)*+,!-./)+*'0!%%4!

!! !!!! 5! 6! 7! 8! 9! :5!

;<#

!

1*2

3!*

0/

=>

-?*

@A!

@'?/

)-0B/!%

%4! 5! 5C! 5C! :C! :C! 6C! 6C!

6! :DC! :DC! :7C! :7C! :EC! :EC!

7! 6:C! 6:C! 66C! 66C! 6DC! 6DC!

$%%

!&'

(!@

'?/

)-0

B/

!F'

)!

*0/

=>

-?*

@A!

8! 68C! 6GC! 6GC! 69C! 69C! 6HC!

9! D6C! DDC! DDC! D7C! D7C! DEC!

:5! 75C! 7:C! 7:C! 76C! 76C! 76C!

Table 15.2: Combined inequality-risk premiums for mortality impacts, 2080-2099 RCP 8.5, Premium as percentage of expected losses, applying value of a statistical life

!! !! !! ""#!

!$%%!&'(!)*+,!-./)+*'0! 1*23!)*+,!-./)+*'0!%%4!

!! !!!! 5! 6! 7! 8! 9! :5!

;<#

!

1*2

3!*

0/

=>

-?*

@A!

@'?/

)-0B/!%

%4! 5! 5C! DC! 8C! :5C! :7C! :9C!

6! 69C! D6C! D8C! 7:C! 78C! E6C!

7! E9C! 87C! G5C! G8C! 9DC! H:C!

$%%

!&'

(!@

'?/

)-0

B/

!F'

)!

*0/

=>

-?*

@A!

8! H5C! HGC!:57

C!

::D

C!

:6:

C!

:D:

C!

9!:6:

C!

:6H

C!

:D9

C!

:79

C!

:EH

C!

:G5

C!

:5!:E5

C!

:EH

C!

:G5

C!

:9:

C!

:HD

C!

658

C!

c⇤i =�X

j

pjc1�⌘i,j

�1/(1�⌘).

X

i

Niv(c⇤i ) =

X

i

Niv(c0i (1� f))

certainty-equivalent consumption ci* given by

inequality-neutral equivalent % loss f given by


Take-aways

• By 2020-2039, median projected average summer temperatures in DC match and the expected number of dangerously humid summer days exceed those of Mississippi today.

• By 2080-2099 under RCP 8.5, the NE, SE, and MW south of the Mason-Dixon lines have median projected summer T hotter than Louisiana today, and even north of M-D line have more expected dangerously humid days than Louisiana.

• Mortality & Labor are largest costs in the US; Energy & Coastal impacts sizable.

• Median projected increase in deaths under RCP 8.5, 2080-2099, is about 10 per 100,000, similar to current traffic death rate.

• Cost of Crime ≈ cost to Agriculture (small in $).


Take-aways

• Largest sources of uncertainty continues to be driven by physical climate.

• Nonlinear response functions " South and Midwest lose most in the sectors we quantified.

• Inequality impact on welfare is large (likely exceeds risk effect), with mortality the largest source of inequality (RCP 8.5 2080-2099 median, mortality in wealthiest 10 states falls ~1/100,000 and in poorest 10 states rises ~30/100,000).

• Mitigation benefits largest and most certain for labor, mortality, energy, and crime. Agriculture benefits less clear because of carbon fertilization; coastal because of slow response of the system.


Potentially generalizable innovations in SEAGLAS

• DMAS.berkeley.edu – broadly applicable to statistical analyses, lowers the cost of cross-disciplinary communication through automation.

• Framework for probabilizing GCM projections • Approach to micro-founding damage functions • Coupling of downscaled GCMs and impact functions in a

coherent probabilistic framework to quantify risk and inequality

http://DMAS.berkeley.edu

Resources for the Future | American Climate Prospectus: Economic Risks in the United States ‹#›

10 East 40th Street, Suite 3601, New York, NY 10016 Tel: +1.212.532.1158 | Fax: +1.212.532.1162 | Web: www.rhgroup.netwww.climateprospectus.org

American Climate ProspectusEconomic Risks in the United States

Climate Change Impacts/Integrated Assessment | Snowmass, CO | July 22, 2014

Trevor HouserRhodium Group

Robert KoppRutgers University

Solomon HsiangUC Berkeley

Roger Muir-WoodRMS

Kate LarsenRhodium Group

DJ RasmussenRhodium Group

Amir JinaColumbia University

Paul WilsonRMS

Michael Delgado Rhodium Group

Michael MastrandreaStanford University

James RisingColumbia University

Shashank MohanRhodium Group

Prepared for the Risky Business Project

Documents

American Climate Prospectus - Stanford University · CCI/IA | American Climate Prospectus: Economic Risks in the United States 3 An Independent Assessment for a Climate Risk Committee