cost of carbon: discounts, fungibility and agricultural ghg offset projects bruce a. mccarl...
TRANSCRIPT
Cost of Carbon:Discounts, Fungibility and
Agricultural GHG Offset projects
Bruce A. McCarlDistinguished and Regents Professor of Agricultural Economics
Texas A&M University
Presented atClimate Change Segment of Advanced Resources Class
College Station, Feb 2011
Fungibility
A number of concepts have arisen that are likely to differentially characterize the contribution of alternative possible offsets within the total regulatory structure. These involve:
PermanenceAdditionalityLeakageUncertainty GWP
General concern price may differentiate based on characteristics like a grading standard
Why is Permanence an Issue
Volatility of sequestered carbon - sequestered carbon can be rapidly released back to the atmosphere on reversal of practices, fire etc.
Saturation /New Equilibrium - differential rates of accumulation over time and a long run decline to a near zero rate of net sequestration
Sustainability of Practices – crop rotations and herbicide resistance plus land diversion
Contract duration and liability terms - project payment terms, liability and duration influence offset value including leasing
Uncertainty – how much carbon is sequestered and retained (not entirely a permanence issue but closely related)
Saturation/ New Equilibrium of Sequestration in Ag Soils and Forests
Figure 2. Cumulative Carbon sequestration in a Southeastern U.S. pine plantation
Source: Data Drawn form Birdsey (1996)
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Age (years)
Me
tric
ton
s C
/ac
C in trees and understory
C in soil and litter
Total C
West and Post 2002 Soil Organic Carbon Sequestration by Tillage and Crop Rotation: A Global Data Analysis Soil Science
Society of America Journal 66:1930-1946 (2002)
Note saturation by year 20
0
20
40
60
80
100
120
140
160
180
200
0 5 10 15 20 25
Experiment Duration (year)
Abs
olut
e ch
ange
in th
e an
nual
am
ount
of c
arbo
n se
ques
tere
d w
ith
a ch
ange
from
CT
to N
T (
%)
Birdsey et al, USFS, FORCARB
Note saturation by year 80
Soil C sequestration over time after a change from conventional to zero-tillage
operations
Permanence and The Price of Carbon
The question is if such items are allowed whether the permanence concerns associated with sequestration may alter the value or quantity of the resultant carbon offset in the market place
Offsets are not fungible – from an offset purchaser’s point of a view, an impermanent sequestration asset may be worth a different amount than permanent offset
General concern price may differentiate based on permanence characteristics like a grading standard
FungibilityGrading standard
#2 Yellow corn, CD plywood, Long staple cotton
Receive a price premium/discount depending upon product characteristics and consumer cost of using
GHG offset (especially carbon sequestration) may have consumer cost effects being not fully claimable permanence (plus additionality, leakage, and uncertainty)
Smith, G.A., B.A. McCarl, C.S. Li, J.H. Reynolds, R. Hammerschlag, R.L. Sass, W.J. Parton, S.M. Ogle, K. Paustian, J.A. Holtkamp, and W. Barbour, Harnessing farms and forests in the low-carbon economy: how to create, measure, and verify greenhouse gas offsets, Edited by Zach Willey and Bill
Chameides, Durham, NC: Duke University Press, 229 p, 2007.
McCarl, B.A., "Permanence, Leakage, Uncertainty and Additionality in GHG Projects", Paper developed as input to book Quantifying Greenhouse Gas Emission Offsets Generated by Changing Land Management, A b ook developed by Environmental Defense, 2006.
http://agecon2.tamu.edu/people/faculty/mccarl-bruce/papers/1149.pdf
How to derive Permanence Discount?
Permanence discount can arise in terms of a Price discount where like three grades of gasoline one needs to pay more for a more permanent item
Quantity discount where the saleable quantity is reduced to a quantity that one can permanently count on or is discounted for its impermanent terms
Either way one gets less than would be paid for a more permanent asset like capturing methane from a landfill or a manure lagoon and immediately burning it up
Kim, M-K., B.A. McCarl, and B.C. Murray, "Permanence Discounting for Land-Based Carbon Sequestration", Ecological Economics, vol. 64, issue 4, 763-769, 2008.
How to derive Permanence Discount?
Effective price (PE) today for impermanent offset – carbon sequestration
Cost of impermanentEquivalent Price = __________________________
Quantity receivedWhat is cost --
Quantity of offsets over time times carbon price
Plus if leasing the cost of obtaining replacement credits when the contract expires
Plus cost of maintaining practice that is not related to carbon (ie money to maintain practice and monitor carbon
Note last two terms zero for permanent assets as GHG is destroyed
Cost of Carbon -- Breaking it down : QGHGO Nominal quantity of offsets
The quantity of GHGE offsets that can be claimed before discounts is the difference from the time path of
net GHE emissions and increments in sequestration
by the target entity before and after possibly discounted back to the current period.
This consists of changes ino Direct sequestration- amount of additional carbon stored in soil
or in multiyear standing plants or trees over time. o Methane and nitrous oxide emissions.o Fossil fuel emissions due to changes in input usageo Carbon releases in input manufacturing.
Fungibility- How do we derive price discount?
tT
tt
tT
ttt
t
tt
DiscfsetInYearQuantityOf
DiscOtherCostfsetInYearQuantityOficeOffset
isc
isc
OtherCost
)1(/
)1(/)*Pr( Ton CurCostPer
)D(1
fsetInYearQuantityOf fsetTodayQuantityOf
)D(1
fsetInYearQuantityOf tInYearPriceOffse t ostOfOffsePresValueC
fsetTodayQuantityOf
tostOfOffsePresValueC Ton CurCostPer
0
0
T
0tt
T
0tt
t
Note I have a non constant price variant
Fungibility- How do we derive price discount?
imperfectperfect
perfect
t
t
t
t
tT
0tt
tT
0tttt
perfect
TonCurCostPerTonCurCostPer implies
OffsetPrTonCurCostPer
QOffsetffsetClaimQuanO
0OthCost
0Buyback
0PDiscount
QOffsetQOffset
Disc)(1/ffsetClaimQuanO
Disc)(1)/OthCostBuyback*OffsetPrPDiscount)(1*QOffset*OffsetPr( TonCurCostPer
To derive price discount for permanence etc add some terms (Pdiscount, buyback and claimable offsets) then equate a perfect perpetual offset with an imperfect one
Fungibility- How do we derive price discount?
tT
0tt
tT
0ttt
tT
0tt
tT
0tttt
imperfectperfect
tt
t
t
t
tT
0tt
tT
0tttt
timpermanen
Disc)(1/QOffset
Disc)(1)/et/PriceOffsMainCostBuyback( =nt PermDiscouimplies
Disc)(1/QOffset
Disc)(1)/OthCost)BuybackQOffset*PDiscount)((1*OffsetPr(OffsetPr
TonCurCostPerTonCurCostPer
QOffsetffsetClaimQuanO
quantityoffset offunction anot is that paid is cemaintainen if 0OthCost
reversesprojectiforleasingif0Buyback
twithvariesQOffset
PDiscount)(1*OffsetPrdiscountedisoffsetnonpermtopaidOffsetPr
Disc)(1/ffsetClaimQuanO
Disc)(1)/OthCostBuyback*ffsetPrQOffset*discount)(1*OffsetPr( TonCurCostPer
OP
Permanence case
Fungibility- How do we derive price discount?
tT
0tt
tT
0ttt
Disc)(1/QOffset
Disc)(1)/et/PriceOffsMainCostBuyback( =nt PermDiscou
Permanence case
When is discount zeroNo BuybackNo Maintenance cost
25 year lease with 100% buyback – 48% price discount Maintenance at 10% of cost -- 36%
How Big is the Discount?
Agricultural soil carbon sequestration
25 year lease with 100% buyback – approximately 49% price discount
Maintenance cost at $5/acre – approximately 36% price discount
Afforestation
Harvest year 20 without reforestation – 52%
Harvest year 20 with reforestation – 23%
Harvest year 50 without reforestation – 20%
Harvest year 50 with reforestation – 7%
Do we include items that are not permanent
Contract terms like leasing and liability can handle non permanent characteristics
Trading is designed to allow cheaper offsets to be used than can be developed by the initial emitter. So the question is are these impermanent items cheaper despite permanence?
Many things are impermanent
Policies that reduce fossil fuel use leave the resource in the ground as a potentially volatile source of emissions
CCS is impermanent as it puts items in places where they could be released from and which requires maintenance/monitoring costs
Sequestration strategies may provide a bridge to a lower cost emission reducing future where CCS or other things can be cheaply used
Sequestration has other benefits in water quality, erosion and is non competitive with food in some ways
Carbon Sequestration and Uncertainty
Quantity of land-based carbon sequestration is subject to uncertainty - buyers may incorporate uncertainty in their offering prices
Uncertainty is used to describe
Phenomena such as statistical variability, lack of knowledge or surprise
Lack of confidence in a single value
This paper presents an empirical confidence interval based uncertainty discount approach
Then we apply to an East Texas case
Kim, M-K., and B.A. McCarl, "Uncertainty Discounting for Land-Based Carbon Sequestration", Journal of Agricultural and Applied Economics, forthcoming, 2009.
http://agecon2.tamu.edu/people/faculty/mccarl-bruce/papers/1121.pdf
Issues on Uncertainty– Shortfall Penalty
Purchaser of carbon credit faces risk of having the quantity sequestered falling below the claimed level (shortfall penalty)
Example: US SO2 trading - penalty for excess emissions of SO2 is set at $2000/ton annual adjustment factor which is 3-5 times offset price
Substantial interest on behalf of the purchaser directed toward ensuring that the potential offset credits acquired can be safely relied upon to exceed the environmental commitments
Sources of Uncertainty
Climate and other factors like pests, fire etc.
Aggregation induced sampling error at a regional scale
Carbon pool measurement error, and
Inter-temporal variation in the duration and permanence of carbon sequestered in the future
Uncertainty Issue (3) – Appropriate Distribution
Site level - Carbon distribution for a year highly variable according to a Canadian soil scientist I once heard talk. But site level is not right basis
Spatial aggregation - aggregation of multiple sites (or farms) generating carbon credits
A contract for 100,000 tons may require 800 US farms of an average farm size of 500 acres
Temporal aggregation - multi-year contracts with the same group of carbon credit suppliers
Project commitments spanning over a number of years expected due to the slow change characteristics of carbon
Empirical RiskRegion Sorghum Rice Soybean
Brazoria county, TX 21.4 14.2 23.1
TX crop reporting district 9 17.0 7.4 18.1
State of TX 10.4 7.5 15.6
US 8.8 5.2 7.0
Region Sorghum Rice Soybean
Brazoria county, TX 5.1 5.3 8.7
TX crop reporting district 9 2.9 2.3 5.4
State of TX 3.3 2.2 3.9
US 1.3 2.0 2.5
Over 5 years
Within one year
Moral : Aggregation over time and geography reduces riskDon’t use numbers from crop modelers
Empirical Uncertainty Discount
The uncertainty discount from a confidence interval approach Discount = Multiplier * Relative risk * Carbon/Yield relation
We get a CV of 6.3% resulting in the uncertainty discounts of 10.2% for a 95% confidence level and 7.9% with a 90% confidence level
Kim, M-K., and B.A. McCarl, "Uncertainty Discounting for Land-Based Carbon Sequestration", Journal of Agricultural and Applied Economics, forthcoming, 2009.
http://agecon2.tamu.edu/people/faculty/mccarl-bruce/papers/1121.pdf
Fungibility - Uncertainty
* CV
CV * Z yDiscUncertaint α
Discount given a
Coefficient of Variation (CV) of
Confidence Level
Multiplier
from Normal
Distribution
Z
5% 10% 80% 0.84 4.21% 8.42% 85% 1.04 5.18% 10.36% 90% 1.28 6.41% 12.82% 95% 1.64 8.22% 16.44% 99% 2.33 11.63% 23.26%
Cost of Carbon -- Breaking it down : Additionality (ADD)
Under the KP concern that activities only receive credit if they would not have been done in the absence of GHGE offset incentives.
If under BAU land would have left farming to grasslands that project for such a transformation should receive a discount.
Need projection of without project baseline along with assumption on whether reduction in baseline would also receive credit.
Raises issue current offset practices may be discontinued to become eligible to start them anew and qualify for offset payments
This discount should be higher at low carbon prices than at higher prices
Additionality and Baselines Premise: Only reductions/sequestration
other than what would have occurred without the project are considered additionalCrediting programs may exert
additionality requirement for offset crediting (e.g., CDM and JI/Kyoto Protocol)
To determine additionality, you need a project baseline
Project Baselines A quantification of what would occur (over time) on the
project land if no project were adopted Practice adoption (e.g., tillage adoption) Carbon consequences (dynamic)
Basic approaches Project-specific
Structured case study of project conditions Performance standard
Analysis of cohort group data to determine prevailing practices and GHG performance
Estimating Baseline for Ag Soil C Tillage Project I
Performance Standard Approach:
1. Obtain data on tillage practice adoption
2. Segment relevant cohort group (e.g., farmers in a specific region, state, district, county,…)
3. Estimate probability of no-till adoption for relevant cohort group
4. Predict adoption forward over project period
5. Quantify adoption to soil carbon effects using biophysical models (e.g., Century, EPIC) or emission factors (IPCC)
6. Combine (4) and (5) to quantify soil C baseline over time
Estimating Baseline for Ag Soil C Tillage Project II
Project-specific Approach:
1. Barriers test: determine if there are any legal, institutional, technological, or market barriers to no-till adoption on the project site If so, then the project is completely additional
2. Compare no-till to conventional tillage on traditional economic grounds for the project site If conventional-till is more profitable without carbon
payments, it should be the baseline and the no-till project is additional
If no-till is more profitable without carbon paymentsÞ No-till is baselineÞ Project is not additional
Baseline selection issues Value of extra precision
Landscape heterogeneity Uncertainty reduction for buyer and seller
Cost of extra precision Data collection Analysis
Incentive compatibility Do subjective evaluations leave too much
discretion and opportunity to game the system ? Adverse selection: Will set standards favor
entry of “bad” (non-additional) projects? Updatable?
Fungibility - Additionality
* CV
tT
0tt
tT
0ttt
Disc)(1/QuanOffset
Disc)(1/AdditionalProportion*QuanOffset ityDiscAdditional
tOffsetsWithProjec
fsetsBaselineOf- tOffsetsWithProjec AdditionalProportion
Texas Rice Case – 67% acreage reduction in 15 years12% price discount when converting to grass, 4% to trees
Cost of Carbon -- Breaking it down : Leakage (LEAK)
GHG offsets can be undermined if leakage or slippage occurs.
Actions to reduce net emissions may alter current or anticipated production levels, in turn creating alterations in market conditions (e.g. price effects) that can induce emission increases elsewhere
The sectors, particularly agriculture are highly competitive. Elements of the markets are global with commodities moving freely throughout the world. As such leakage is certainly a concern. Localized projects will stimulate additional economic activity domestically or internationally. Findings range from 20-80 percent, can use Murray et al formula.
Fungibility - Leakage
* CV
pr
ot
C P)] (1*E - [e
C * e LeakingProportion
LeakingProportion-1 L
eakDisc
e is the price elasticity of supply for off project producers.
E is the price elasticity of demand for commodity produced.
Cot is GHG emissions per unit of increased commodity production outside project.
Cpr is GHG offsets per unit of reduced commodity production in project.
P is relative market share and is quantity of commodity produced by project divided by market amount produced.
Murray, B.C., B.A. McCarl, and H-C. Lee, "Estimating Leakage From Forest Carbon Sequestration Programs", Land Economics, 80(1), 109-124, 2004.
Fungibility - LeakageInternational
US Only US and Annex B Countries
All Countries
$10 $100 $10 $100 $10 $100 U.S.
Production of Traded Crops 99.60 93.47 99.87 97.09 100.52 105.11
All Production 99.33 97.53 99.93 97.43 99.47 98.59
Exports 98.84 81.77 99.93 97.65 102.19 126.92
Production of traded commodities in
rest of world
Global production 99.96 99.60 99.95 99.44 99.98 99.71
Annex B Countries (excluding U.S.) 100.36 102.66 99.51 92.31 99.61 99.25
Non-Annex B Countries 100.32 112.22 100.49 120.13 96.89 57.60
Note All datat are index nubers of production in a categoryParticipating production is offset by production elsewhere
Scope of Participation
Fungibility - Empirical
* CV
Perm Add Leak Uncer All SaleableRice to crops 30% 12% 32% 10% 62% 38%Rice to pasture 50% 4% 17% 10% 64% 36%Rice - trees(pulp) 30% 1% 16% 10% 48% 52%Rice - trees (saw) 10% 1% 16% 10% 33% 67%
Not additive
Beaumont through Columbus Texas area has historically produced rice. In 1985, 600,000 acres. In 2000, 214,000 acres. Policy, environment and markets are applying pressure. Today, many rice producers are in quest of new opportunities. Trees, other crops and pasture provide possible alternatives to some.
LeakDisc)(1*AddDisc)(1*UncerDisc)(1*PermDisc)(1*eOffsetpricrsetProducePricetoOff
Is this a problem – in a model
Not always
Full coverage eliminates leakageMulti period is handled in fasomAdditionality handled by dynamic baseline
Uncertainty is not
What issues might IAM modelers consider? Fungibility - Aggegate
0
100
200
300
400
500
0 50 100 150 200 250 300 350 400
Car
bo
n P
rice
in
Do
llar
s p
er T
CE
Emission Reduction in MMT of Carbon Equivalents
Total Greenhouse Gas Emission Reductions from Agriculture and Forestry
Discount for Saturating SinksNo Sink Discounting
0
100
200
300
400
500
0 10 20 30 40 50 60 70 80
Car
bo
n P
rice
in
Do
llar
s p
er T
CE
Emission Reduction in MMT of Carbon Equivalents
Soil Carbon Sequestration
Discount for Saturating SinksNo Sink Discounting
0
100
200
300
400
500
0 50 100 150 200
Car
bo
n P
rice
in
Do
llar
s p
er T
CE
Emission Reduction in MMT of Carbon Equivalents
Carbon Sequestration from Trees
Discount for Saturating SinksNo Sink Discounting
0
100
200
300
400
500
0 20 40 60 80 100 120 140 160
Car
bo
n P
rice
in
Do
llar
s p
er T
CE
Emission Reduction in MMT of Carbon Equivalents
Biofuel Offsets
Discount for Saturating SinksNo Sink Discounting
Is this a problem – with projects
Always
Partial coverage virtually insures leakageMulti period needs to be handled when
buyback or maintenanceAdditionality depends on rules
Uncertainty is there
SO WHAT
Fungibility can be a problemOpportunities are not perfect substitutesProjects may aggravate problemModelers will lose hair over payment schemes
Big Holy Trinity
Cost of Carbon Including the gap
Private cost
Carbon will cost money to produce, sell, and measure, govt may help
Not all carbon may be saleable
where DISC = (1-ADD)*(1-LEAK)*(1-UNCER)*(1-PERM)
DISC*QGHGO
)GCMTCPAIC PDC(tonpercostPrivate
Cost of Carbon
Private cost
PDC – Cost producer incurs to switch from current practices (estimated by models we have looked at)
PAIC - Cost to get producer to adopt above PDC in terms of incentive to get trained bear extra risk etc.
MTC - Transactions cost to assemble, measure, monitor, certify, sell, carbon
GC - Government cost share
DISC*QGHGO
)GCMTCPAIC PDC (tonpercostPrivate
Cost of Carbon Breaking it down
PDC – Cost producer incurs to switch from current practices
Incremental difference in
o Revenue implications of altered yield (change in yield times relevant price over all commodities).
o Cost implications of altered input usage (change in input usage times relevant price) over all inputs
o Cost implications of new equipment requirements and the salvage value for displaced equipment
Has been done by most who have looked at cost of carbon
Cost of Carbon -- Breaking it down PAIC - Cost to get producer to adopt above PDC
Offset of PDC marginal cost is not sufficient for adoption
Incremental incentive to overcomeo substantial learning time, o increased risk o transition investment costs
Kurkalova, Kling, and Zhao find a PAIC of about $3 per acre would be needed for Iowa corn and soybean farmers $12/ tonne using West and Post's 0.25 tonne/ acre
Lewbowski, Plantinga and Stavins compute logit probability ofafforestation given incentives
What could we do
Cost of Carbon -- Breaking it down MTC- Transactions cost
Carbon must be marketed and cost arises
o Assembly Costs
o Measurement, and monitoring
o Certification
o Enforcement
o Additional adoption cost incentive estimates
o Management of adverse outcomes
Cost of Carbon -- Breaking MTC downAssembly cost
Emitting entities emit large quantities of GHGs. (How much from a 100 megawatt coal power plant)
It not economically efficient for a purchaser in quest of 100,000 tons to deal with a single farmer.
100,000 tonnes at 0.25 tons per acre = 400,000 acres400 acres/farm =>1,000 farmers (avg U.S. is 460 acres)
This implies role for brokers who aggregate producers and sell permits. Cost arises in such process.
Crop insurers get 25% or 1/3 above what is paid to farmersMinn water quality 50%
Cost of Carbon -- Breaking MTC downMeasurement, and monitoring
Conveyance will require measurement and monitoring to establish offsets are being produced and continue.
This requires the development of a low cost measurement and monitoring approach that involves a sampling based scheme integrating field level measurement, computer simulation, and remote sensing on some dynamic and geographical basis.
Mooney et al propose a scheme and assert that (1) efficiency depends on the price of C credits. (2) costs largest in areas with greatest heterogeneity (3) in case study cost <= 3% of the value of a C-credit.
Cost of Carbon -- Breaking MTC downCertification
Certain bodies may develop and certify offset quantity estimates for practices and then monitor that the practices continue.
For example a government rating could be established that indicates the number of offset credits from a tillage change under a set of circumstances.
Costs of obtaining such a certification as borne by private parties and by the government would be relevant cost
components.
Cost of Carbon -- Breaking MTC downEnforcement
Contact enforcement may require hearings and setup of an enforcement entity
Further this may arise between traders or within an assembly group
Some estimate is needed of costs that will be encountered for the enforcement of permit contractual obligations.
Cost of Carbon -- Breaking MTC downAdditional adoption cost incentive estimates
Cost may well be encountered involving education and training of producers on how to alter their practices so that they most efficiently produce greenhouse gas offsets.
These costs need to be estimated in a way so that one does not double count the producer adoption incentive terms as discussed above
Cost of Carbon -- Breaking MTC downManagement of adverse outcomes
Certain classes of offsets are volatile and subject to uncertainty including possible destruction by extreme weather events, fires, floods etc.
Brokers and or contracts may include procedures to insure against certain types of failures.
These may involve within contract enforcement mechanisms, insuranceplanned safety margin where more offsets are produced
then are sold allowing make up for shortfalls
Cost of Carbon -- Breaking it down How do we get at this? - Alternatives
1. Examine existing but immature marketsGEMCO, Overseas. California, RGGIChicago climate exchange
2. Examine similar enterprisesWetlands , OilmanForester who consolidatesSO2, Water qualityCoops, Mortgage sales
3. Engineer something
4. Simulate a world
Cost of Carbon -- Breaking it down GC- Government cost share
Government may have an active role in some of the
assemblymeasurement and monitoringproducer education sale
thus may offset some proportion of the total costs.
Cost of Carbon -- Breaking it downRationale for Public cost bearing
Suppose social cost for 2 projects is such thatsc1>sc2
so society favors project 2
But the order of the private costs is reversedpc1<pc2
so private groups prefer project 1
Thus we have a potential market failure, what can the government do?
Bear cost of (pc2-pc1)*QGHGO*Disc for activity 2. As long as that number does not exceed difference in inefficiency cost plus co benefits
Cost of Carbon -- Breaking it down
Public cost
ф – Govt inefficiencyCB – Co-benefits
where DISC = (1-ADD)*(1-LEAK)*(1-UNCER)*(1-SATV)
DISC*QGHGO
)CBGC*φMTCPAIC PDC (tonpercostSocial