The Social Cost of Carbon –An Overview

David Anthoff

Energy and Resources Group

University of California, Berkeley

1

Outline

• What is the SCC?

• How is it used?

• Current research efforts

• Some important issues in policy use

2

DefinitionAn economic measure of the harm caused by emitting one extra ton of CO2.

3

0

5

10

15

20

25

30

2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2080 2085 2090 2095 2100

CO2 emissions

Base run

Marginal runI=

4

0100200300400500600700800

pp

m

CO2 concentrations

Base run Marginal run

0

1

2

3

4

5

2010202020302040205020602070208020902100

°C a

bo

ve p

re-i

nd

ust

rial

Temperature

Base run Marginal run

050

100150200250300

bn

$

Climate damage US

Base run Marginal run

050

100150200250300

bn

$

Climate damage Europe

Base run Marginal run

050

100150200250300

bn

$

Climate damage China

Base run Marginal run

0

5

10

15

20

25

30

CO2 emissions

Base run Marginal run

5

Marginal damage

USA Europe China …

2010 $0.20 $0.03 $0.12 …

2011 $0.20 $0.04 $0.12 …

2012 $0.30 $0.05 $0.13 …

2013 $0.40 $0.05 $0.14 …

2014 $0.50 $0.06 $0.15 …

2015 $0.60 $0.07 $0.16 …

2016 $0.70 $0.08 $0.17 …

2017 $0.80 $0.09 $0.18 …

…. … … … …𝑆𝐶𝐶 = 𝔼

𝑡

𝑟

𝑤𝑡𝑟 ×𝑀𝐷𝑡𝑟𝑆𝐶𝐶 =

𝑡

𝑟

𝑤𝑡𝑟 ×𝑀𝐷𝑡𝑟

1 ~ ,

6

Integrated assessment models

7

Economy

Population

Atmospheric chemistry

Climate change

Impacts

Emissions

Concentrations

Physical impact

$ Damage

Mortality

8

Sources of published SCC estimates

• DICE/RICE Bill Nordhaus

• PAGE Chris Hope

• FUND Richard Tol & me

Surveys, “meta-studies”

Parametric variations

Simplified versions

9

Outline

• What is the SCC?

• How is it used?

• Current research efforts

• Some important issues in policy use

10

Academic use

• Estimates have been published since the mid-1990s

• A Nobel Prize has been won for this kind of work (Nordhaus)

• A large variety of different papers• New SCC estimates

• Investigations into discounting, equity or risk preferences

• Used as an input into policy evaluations

• IAMs used to find optimal policies (or robust policies, or strategies)

11

Policy use

• United Kingdom: 2000s• CBA of individual projects/regulations

• Later ex-post analysis of major national climate policy

• United States: 2010-present• Regulatory impact analysis

• German: 2007 (?) – present

12

A short history of the SCC in the US – Part 1

• Everything happening within existing regulatory framework (Clean AirAct, 2007 Supreme Court pollutant ruling, 2009 endangermentfinding, EO 12866)

• Obama administration convenes Interagency Working Group on theSCC 2010 official SCC estimate• Three IAMs: DICE, PAGE and FUND• Five standardized socio-economic and emissions scenarios• Standardized climate sensitivity• 2.5%, 3% and 5% discount rate

• 2013 update (uses newer model versions)

• ~2015 White House commissions NAS report

13

Figure 2: Net benefits with and without valuing CO2

1,000,000

100,000

■ Net benefits w ith CO2

■ Net benefits w ithout CO2

10,000

~ ~ ~ 0 N C: 1,000 0

~

100

10 -

1 Observations in order of increasing net benefits (n = 50)

14

0

5

10

15

20

25

30

35

40

45

Option 1 - state Option 1 - regional Option 2 - state Option 2 - regional

bill

ion

20

11

$

Proposed rule for existing power plants - benefits and cost in 2020

Complience Cost Health benefits Climate benefits (5% dr) Climate benefits (3% dr) Climate benefits (2.5% dr)

Source: US EPA (2014), Table ES-8

■ ■ ■ ■ ■

15

Federal SCC

States:• Minnesota• Colorado• Maine• Nevada• Illinois• New York• California

International:• Canada• …

Various court cases.

16

Used in dozens of federal regulatory impact assessments (including Clean Power Plan rule)

CLIMATE

LEADERSHIP

COUNCIL

THE CONSERVATIVE CASE FOR CARBON DIVIDENDS How a new climate strategy can strengthen our economy, reduce regulation, help working-class Americans, shrink government & promote national security

James A. Baker, III

Martin Feldstein

Ted Halstead

N. Gregory M?.nlsiw

Henry M. Paulson, Jr.

George P. Shultz

[hom?sStephenson

RoB Walton 16

A short history of the SCC in the US – Part 2

• Trump issues EO that rescinds IAW SCC estimates during first monthsin office

• “Regulatory Impact Analysis for the Review of the Clean Power Plan:Proposal”

• Still uses the Social Cost of Carbon concept

• Two key changes:• Discount rate (3% and 7% replace 2.5%, 3% and 5%)• Domestic SCC

• New SCC estimates: $6/tCO2 and $1/tCO2

• Nothing else as changed on the SCC side

17

Outline

• What is the SCC?

• How is it used?

• Current research efforts

• Some important issues in policy use

18

Two large scale efforts underway

• Climate Impact Lab (Houser, Greenstone, Hsiang, Kopp et al.)• Develop new damage estimates of climate impacts using modern

econometrics

• RFF Social Cost of Carbon Initiative (network of partners, RFF, UC Berkeley, Harvard, Princeton, U Washington, PennState and others)• Implement the National Academies recommendations

19

RFF SCC Initiative I

• Provide a modular computational platform for integrated assessment models (Mimi.jl)

• Develop probabilistic projections for relevant socio-economic variables, like GDP, population, emissions intensity• Statistical approach spearheaded by Stock, Muller and Watson• Expert elicitation spearheaded by Cooke

• Create state-of-the-art natural science components (Berkeley, Penn State and Oxford collaboration)

• Integrate existing damage estimates from the literature (and hopefully Impact Lab results once they are out)

• Develop a discounting module compatible with the NAS recommendations (Pizer, Newell & Prest)

20

RFF SCC Initiative II

• We see our initiative very much as providing a platform forcollaboration

• We want other groups to create modules on the Mimi.jl platform

• We can provide support and help to other groups

• We want to provide results and tools for policy makers

21

Mimi.jl

• Modular platform for integrated assessment models

• Some goals:• Open source and free

• Fast

• Easy to use

• Decentralized workflow

• Transparency in research

• Some non-goals:• We don’t aim to be the best platform for every type of IAM

22

Existing models on Mimi.jl

• MimiFUND.jl

• MimiDICE2010.jl

• MimiDICE2013.jl

• MimiRICE2010.jl

• MimiPAGE2009.jl

• Mimi-SNEASY.jl

• Mimi-FAIR.jl

• Mimi-MAGICC.jl (CH4 parts)

• Mimi-HECTOR.jl (CH4 parts)

• Mimi-CIAM.jl

• Mimi-BRICK.jl

• AWASH

• PAGE-ICE

• Local air pollution

• NICE

• NICER

23

Groups using/collaborating

• UC Berkeley (my lab)

• RFF

• Penn State (Klaus Keller’s lab)

• Princeton University (Marc Fleurbaey and Rob Socolow’s group)

• RFF Europe (formerly known as FEEM) (planned)

• A proposal out of Duke (planned)

• LSE (James Rising)

• RAND Corporation

24

Outline

• What is the SCC?

• How is it used?

• Current research efforts

• Some important issues in policy use

25

Equity

Count US damages“Domestic SCC”

Count world damages“Global SCC”

Count equity adjusted world damages

“Equity weighted SCC”

Trump administration Obama administration UK (but incorrectly), Germany

Only the harm experienced within US borders is accounted for.

Harm anywhere in the world is accounted for. A dollar loss in the US receives the same weight as a dollar loss in the poorest region in the world.

Harm anywhere in the world is accounted for. A dollar loss to someone poor is weighted more than a dollar loss to someone richer.

26

Robustness

• Deep (structural) uncertainty is pervasive and (in my opinion) won’t go away• If someone tells you anything else, I would get VERY suspicious

• There will be multiple damage functions, multiple scenarios, multiple models, and there won’t be agreement which of these is the “correct” one

• Being transparent about these uncertainties is really important use multiple models, show differences, show how policies fare under different ones

27

SCC in a world of existing climate targets

• In purist world, you use the SCC (or IAMs) to find an optimal target

• We don’t live in a purist world, and to my knowledge economic climate analysis has never been used in that way in real world policy design

• Even if you have a political or scientific target, you can use the SCC to quantify the benefits of individual policies

• (and you can use marginal abatement cost estimates to check whether you are implementing cost-effective policies)

28

Funding

• The SCC is one of the most important numbers to estimate in climate policy

• As far as I know, none of the primary policy users has every provided any research funding for SCC work• Right now foundations like the Sloan Foundation are stepping in

• These are early days for the SCC and we need a massive research effort to improve it

29

Conclusion

• SCC is now deeply embedded in the policy fabric in the US and othercountries

• While there are two major efforts underway to improve the SCC, theresearch field is tiny compared to the natural science side of things does not reflect the importance of this work

30

Thank you!www.david-anthoff.com

anthoff@berkeley.edu

31

Model choice & discount rate& some other sensitivities

32

Source: Nordhaus (2014)

Table 3. Estimates of the Social Cost of Carbon for 2010 from US Interagency Working

Group and Comparison with Alternative Model Estimates

Model and Scenario

B. Estimates of 2015 SCC from

US Working Group, 2013:

DICE..-2010

PAGE

FUND

Average

5%

11.0 20.2

2.7

11.3

Constant Discount Rate on Goods

4.2%* 4%* 3%

31.4 58.6

17.3

35.8

2.5%

48.1

85.3

30.4

54.6

33

Waldhoff et al. (forthcoming)

73 89 so

45 ■ Change w/o CO2 Fertilization

40 ■ With CO2 Fertilization

35

30

N 25 0 u

I ... ' 11).

20

15

10

5

0

34

Discounting

35

Discounting & aggregation

• Opportunity cost approach: we don’t want to waste resources, but try to stay out of distributional judgements (some variant of Kaldor Hicks)

• Social Welfare approach: we take an ethical stance, encoded as a particular Social Welfare Function

36

Opportunity cost approach

• We should use the rate of return of other investments as our discountrate (what else could we have done?)

• 1) So we need a forecast of interest rates

• 2) And we need to figure out which interest rate is the relevant one

• For 1), we could in principle use many different methods, but in theIAM literature with a neo classical growth core, the interest rate isdetermined endogenously (using the Ramsey rule). Note that theRamsey rule might appear normative, but it is NOT in this approach

• Question 2) has received a lot less attention, but there is a smallliterature on the climate beta

37

Social Welfare approach

• You start with a Social Welfare Function, and derive your weights/discount factors from that

• A particular case of SWF that exhibits aversion to inequality between individuals runs under the term “equity weighting” in the climate literature

• Core idea is that of declining marginal welfare of consumption

38

Damage function

39

𝐷𝑡𝑟 =𝑌𝑡𝑟𝑌0

𝜖

× 𝑌0 × 𝑓 𝑇𝑡𝑟

𝐷𝑡𝑟 = 𝑌𝑡𝑟 × 𝑓 𝑇𝑡𝑟

𝑓 𝑇 = 𝛼𝑇2( )

( )

( )

40

Anthoff et al. (under review)

$30/tC

$75/tC

800

700

600

500

400

300

200

100

0

2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2080 2085 2090 2095 2100

-100 ~--------------------------------------

- standard Impacts - Impacts l inear in temperature

Impacts independent of GDP - 1 mpacts proportional to GDP

Figure 6. The social cost of carbon as a function of the time of emission with the standard model, w ith impact functions linear in temperature, and with impacts independent and proportional to GDP.

41

Uncertainty

42

0.4------...---........... --........... ---T----.........---....,,......----r-----

0.35

00 0.3 § . ,_ ~

..52 0 ,25 ::s 8 ,.,,..... rt) O' 1 2-~ . 0 C: .g 0.15 u cd ~

~ 0.1

0 .. 05

. . . . . . ■ • ■ ■ ■ •

t ■ 'I ■ ■ t ■ ,0 ■ ■ • ■ ■ ■ ■ 0 ■ ■ ■ ■ I ■ t ■ ■ I ■ t ■ ■ I ■ I ■ ■ I ■ t ■ ■ I' ■ ■ ■ ■ t, ■ I ■ ■ t, ■ I ■ ■ ■ ■ I ■ ■ " ■ t ■ ■ ■ ■ ■ ■ ■ t, ■ t ■ ■ ■ ■ I ■ ■ ■ ■ t ■ ■ ■ ■ t ■ ■ ■ ■ ■ ■ ■ t, ■ I ■ ■ ■ ■ I ■ ,0 ■ ■ t ■ ■ ■ ■ I ■ ■ ■ ■ I ■ ■ ■ ■ 0 ■ ■ ■ ■ t ■ . . - . - -• • • • ■ • • • a • • •

■ I ■ ■ ■ ■

■ • • • ■ • . . . . . . ■ • • • ■ • . . - - .. -• • ■ ■ ■ ■

■ I ■ ■ ■ ■ . ' . . .. . • • • ■ ■ •

• ■ ,I ■ ■ I ■ ~ ■ ■ I, ■ '1 : 8 ■ I , ■ .. ■ ■ I , ■ ,I ■ ■ I, ■ ,I ■ ■ I, ■ 'I :- ■ M ■ ,I ■ ■ I, ■ ■ ■ ■ I, ■ I ■ ■ I' ■ I ~ ■ I' ■ I ■ ■ .. ■ I ■ ■ I, ■ ■ ■ ■ I' ■ I ■ ~ ■ ■ I ■ II II ■ I ■ ■ .. ■ I ■ ■ ■ ■ ■ ■ : II ■ I ■ II II ■ I ■ >I ■ ■ I ■ ■ .. ■ I ■ ■:■ ■ I ■ ■ II ■ 0 ■ .. II ■ I ■

• • ■ • .. • ■ • • • ■ ■

• • • • ■ •

- . - - . -■ • • ■ • ■

- . - - .. -

--Di -count Rate 2 .. 5%1

...._____, 3 ,. 0%1 5~0%1

• • • • ■ • ■

.......... .. L. . -, .. l .. ~.~P~At.A.V~r.~;Q~. -~--$.1.~ ................. I .................. _,t _ ... ... ... . .... . ..... ..... .... ... ... .. l ........... .. .... . II

I

• I • • • • • •

··· ·· ., .. •· · -; -· · ·· · ·· 1·· ·· ·-·~· ·- · ·· ··· .. ·· · ·· •·· ~- ·:· ·3 · ·0;0;r ·_Av·e·r·a·g·· et~ .. $_•4·3 .... · .. i .. · .... · ·· .. · .... · .. · · .. ·· ..... ... ·· · .,. \· ··· ·· ·-· ·· .. . ·· .. l I I \ I . o :- , , - . !- , \ i • • • • • • e

: : : I : : : : • • • ■ • ..

• ■ ■ ■ ■ " ■ •• ■ ■ ■ •:• , , ■ ■ , ■ ■ I • •• ■ ■ " ■ ■ ~ ■ " ■ 'I ■ ■ r • • • • • • •• • • ,. • " ~ • "I • • • r • , • • r • • • • • • • • ~ ,. • , • " " • , • " r • , • • " • • • : ,. • , • " " • , • " r. • , • " • • , • •:" • , • " " • , • " " • , • • " • , • • • , • " • • , • • " • , . • " " • , , • • ~ • r • • • • r • • • • r • " " • r •

: I . i I l • 2.5% ~veragel = $64 .

..... ...... ·-· . ' .. . . . ', . . ' ......... ' .... ' ~- .1 ... ". ' ......... ' -~ .. , ... ' .... ..... '.; .. ' .... ' .... ..... ' .. : .. ' .... ' .... ' .... .... ' ......... ..... .. ";.' .... '. ' .. '.' .. '. . . : I : I - -

: r--J1■-■1 : I : I : : : ---•r---,1 : : : 1 : : : th ..... .. ,... ... .. . ...... -:-- ......... 'i" i "" i ............ ( ................ ( ···· .. ··· .. ··· .. ·-(····· .. ........... ······· i··3:.0~"'.915"" ' -

I

128,

0 I...Cn__.ll.1Jl_JL..JL_Jlll.1LJL--1C::n::c:l'.!c::::I.Ji::::::::::ll=~='..l:..LI-IJ:....il.U"-.IL.....ll.L....JJ~:::::l..-il.l::I..C::U~::::u:::l:ala.-..~~

0 20 40 60 80 100 120 140 160 Social Cot of Carbon in 2010 [2007$

43

Anthoff & Tol (2013)

Curva ure cooliing energy

Climate ensi ivity

Income ela icity cooling energy

Ber ch ark migra ion ( or er Soviet Union o South Asia)

Benchmark agriculture CO2 fertiliza ion (Japan and .JUu,, • ._

Korea)

ncome elastici

AEEI grow h (Aus ralia and ew Zealand)

Benchmark migration (Latin America to North Africa)

-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06

St ndardized regression coefficient

0.08 0.10 0.12

Fig. 2 The ten mo. t impm1ant parameter. that determine the ocial co t of carbon and their . tandardi . ed regre_. ion coeffici nt for a 1 % pure rate of time pr £ ~ nee and a 1.5 rate of rj, k a ers ion . 90 % confidenc interval a. rror bars 44