Formulas for Quantitative Formulas for Quantitative Emission TargetsEmission Targets

Jeffrey Frankel, Harpel ProfessorJeffrey Frankel, Harpel ProfessorKennedy School of GovernmentKennedy School of Government

Harvard UniversityHarvard University

For workshop on Architectures for Agreement:For workshop on Architectures for Agreement:Addressing Global Climate Change in the Post Kyoto WorldAddressing Global Climate Change in the Post Kyoto World

Harvard University, March 14, 2008Harvard University, March 14, 2008

Organized by Joe Aldy and Robert StavinsOrganized by Joe Aldy and Robert Stavins

A constructive approachA constructive approach

• “You’re Getting Warmer: The Most Feasible Path for Addressing Global Climate Change Does Run Through Kyoto.”

• Features of Kyoto worth building on– Politics: Quantitative limits maximize national sovereignty– Economics: Market mechanisms

• What successor to the 2008-2012 regime, to build on what is good about Kyoto & fix what is lacking?

Desiderata for the next stage, Desiderata for the next stage, requirements for the next multilateral treatyrequirements for the next multilateral treaty

• More comprehensive participation– specifically getting US, China, et al, to join

• Efficiency• Dynamic consistency (credible century path)

• Equity re poor countries• Compliance – no country will stay in if it means

huge economic sacrifice in any given period• Robustness under uncertainty

ProposedProposed Architecture for Architecture for Quantitative Emissions TargetsQuantitative Emissions Targets

• Unlike Kyoto, my proposal seeks to bring all countries in realistically, & to look far into the future.

• But we can’t pretend to see with as fine a degree of resolution at century-long horizon as at 5-year horizon.

CenturyCentury--long horizon: building long horizon: building confidence within a LR frameworkconfidence within a LR framework

• Need a sequence of negotiations, fitting within a common flexible framework

• Threshold for participation: 1 ton of emissions per person?

• How to set sequence of quantitative targets? – a decade at a time. – <= Fixing targets a century ahead is impractical.– Key contribution of this paper: refine my proposal

how the framework would allocate targets

Persuading developing countries to Persuading developing countries to join a system of quantitative targetsjoin a system of quantitative targets

• In LR, as t -> ∞ , developing countries would notionally achieve their equity-based demand for equal emissions /capita.

• Even in SR, quantitative emissions commitments for developing countries, if set carefully, can address everyone’s concerns.

• Principles to guide formulation of such targets: – Targets ≈ BAU path in 0th period => gains from trade– Progressivity, when cuts begin in 1st period, – Protection against inadvertent stringency.

The decade horizon:The decade horizon:formulas for setting targetsformulas for setting targets

• A nested sequence of formulas for emissions. • Formulas would be very general for the distant

future, & become increasingly specific as each budget period draws close.

• At short horizons, the formula for emission limits is phrased as cuts from expected BAU path.

• The BAU baseline entails “growth targets” in emissions for such countries as China & India => we are talking about cuts relative to the baseline, not cut in their absolute levels of emissions.

Formula for targeted reductions would Formula for targeted reductions would include among its determinants:include among its determinants:

• 1990 emissions• emissions in the year of the negotiation• population• income, and • perhaps a few other special variables

– whether the country in question has coal – or hydroelectric power,– though 1990 emissions numbers can capture these.

Threshold for accepting targetsThreshold for accepting targetsPerhaps year in which emissions/cap reach 1 ton CPerhaps year in which emissions/cap reach 1 ton C

20472032-

203720322022-

2032201720122007-

12

20372025202020152010"2007""2002"

IndiaChinaLat.

Amer.Australia+KoreaUS

Japan+Can+NZEU 27

Year in which country agrees to commit to targets: BAU immediately & cuts in subsequent 5-year budget period

Proposed properties of formulasProposed properties of formulasin 0in 0th th period of membershipperiod of membership

• In the year in which a country joins for the 1st

time, agreed target path = estimated BAU path. – => country not asked to take on any economic cost

the moment it joins – Poor joiner can only gain in SR, by the right to sell

permits, giving it an economic incentive to join– Gets China, India etc. tied in (a pre-requisite for US) ,

long before they are asked to make sacrifice– world environment benefits as well, even within the

joiner’s 0th period, by pre-empting leakage.

00thth period, period, continuedcontinued

• For the US, 2008-2012 would (retroactively) be considered 0th budget period, i.e., period in which joiner commits no more than to limit its emissions to BAU.

Proposed properties of formulasProposed properties of formulasin 1in 1stst budget period of each joinerbudget period of each joiner

• Targets embody cuts

• Late joiners’ formula for cuts relative to BAU:– Each 1% diff. in income/cap vs. 1997 EU income

=> target = 0.14% greater emissions abatement from BAU, vs. EU cut agreed under Kyoto.

– This rule was inferred from targets accepted at Kyoto by EU et al (OLS estimates: .11%-.17%)

– Call it the “Progressivity Factor”

The targeted reductions from BAUThe targeted reductions from BAUagreed to at Kyoto were progressive agreed to at Kyoto were progressive

with respect to income.with respect to income.

Formula in 1Formula in 1stst budget periodbudget period

• also puts some weight on 1990 emissions• In current example, weight = .206

• vs. .794 on BAU & Progressive Factor

• Why give weight to 1990 levels? • In part to assure Europeans that US is not being

unduly rewarded for having dropped out of Kyoto & “ramping up” its emissions over 1990-2008 to a higher level from which to negotiate future reductions.

22ndnd budget period:budget period:more serious cuts relative to BAUmore serious cuts relative to BAU

20572042-2047

2037-2042

2032-2037

2022-2027

2017-20222017

IndiaChinaLatinAmer.

Austra.+KoreaUS

Japan +Can +

NZEU 27

Proposed formulas Proposed formulas in 2in 2ndnd budget periodbudget period

• Key off of EU commitment (made in Jan. 2008)to a target 20% below 1990 emissions

• To assign analogous targets to others:– Again, Progressive Factor gives: cuts vs. BAU =

EU's 2008 commitment +/-.14 * log (country income/cap / EU 2007 income/cap)

– As at period 1, put weight (.206) on 1990 emissions.– But now also put weight on a global per cap. average,

which can be used from here on to set overall stringency– For now, 1 ton/cap; also equals 2002 global baseline– Initially, weight on global average = .09;

=> less weight on recent history.

33rdrd budget periodbudget period

2057-2062

2052-2057

2047-2042

2042-2047

2032-2037

2022-2027

2022-2027

IndiaChinaLatinAmer.

Austra.+KoreaUS

Japan +Can +

NZEU 27

Proposed formulas Proposed formulas in 3in 3rdrd budget period. Etc.budget period. Etc.

• Key off of EU commitment (= interpolatn. 2020 - 2050)to a target 25% below 1990 emissions

• To assign analogous targets to others:– Again, Progressive Factor gives: cuts vs. BAU =

EU's commitment +/- .14 * income differential vis-à-vis EU 2007

– As before, put weight (.206) on 1990 emissions.– Weight on global average 1 ton/cap raised to = .16.

Proposed formulas Proposed formulas in 8in 8thth budget period.budget period.

• Key off of EU commitment to a year-2050 target 50% below 1990 emissions

• To assign analogous targets to others:– Again, Progressive Factor gives: cuts vs. BAU =

EU's commitment +/- .14 * income differential – As before, put weight (.206) on 1990 emissions.– Weight on global average 0.85 ton/cap raised to .83.

References can be found at:References can be found at:http://ksghome.harvard.edu/~jfrankel/GEI.htmhttp://ksghome.harvard.edu/~jfrankel/GEI.htm

Next stepsNext steps1. Look at economic & environmental

impact of the proposed formulas

2. Introduce uncertainty

Next stepsNext steps

1. Look at impact of this proposed example of the formulas, with help of FEEM’s Witch model

1. Economic 2. Environmental3. Experiment with some different parameters4. Design a user-friendly version anyone can play with.

Next stepsNext steps1. .

2. Introduce stochastic growth processes1. Impose rule that anyone losing X% of GDP will drop out2. Try to account for the effect of dropouts on remaining members3. Try to account for the effect of possible future breakdown on

expectations of firms deciding long-term investments.4. Hypothesis: X more seriously determines stringency of 1st half-

century path than does the discount rate in cost-benefit analysis5. Seek robustness, by trying

1. indexed decade targets or 2. escape clause or safety valve.

Addendum: Desiderata for the next stage, Addendum: Desiderata for the next stage, requirements for the next multilateral treatyrequirements for the next multilateral treaty

• More comprehensive participation– specifically getting US, China, et al, to join

• Efficiency• Dynamic consistency (credible century path)

• Robustness under uncertainty• Equity re poor countries• Enforceability

We need to get developing countries We need to get developing countries into the system for 3 reasons.into the system for 3 reasons.

(1)Source of the big coming ∆ emissions, according to Business-as-Usual path (BAU), i.e, technical experts’ forecast absent a climate change agreement. • China, India, & other poor countries = 2/3 global CO2 emissions this century. • Without their participation, emissions abatement by industrialized countries will

not do much to mitigate global climate change (2) Leakage: If Kyoto is implemented without developing countries, their

emissions are likely to rise even faster than BAU: • Relocation of carbon-intensive industries from countries with emissions

commitments to non-participants, or • increased consumption of fossil fuels by non-participating countries in response

to declines in world prices. • Damage could be ¼ ton rise in poor countries per ton abated in the rich.

(3) Opportunity for US & other rich countries to buy relatively low-cost emissions abatement from poor countries • is crucial to keep the economic cost low• would make rich countries more likely to comply with their commitments.

Efficiency: minimize the economic costs Efficiency: minimize the economic costs of achieving a given environmental goal.of achieving a given environmental goal.

• “Who flexibility”– Esp. via international trading of permits– sharp cost reductions estimated from trading

• 80%, if key developing countries are in (CEA, 1998). • ≈ independent estimates of Edmonds, et al (1992, 97)

• “When flexibility”– averaging over budget window (e.g., five years)– possibility of banking credits.

• “What flexibility”– substituting across 5 GHGs, not just CO2, or – between the GHGs & sequestration through forestation. – reduces both E (econ. cost) & vulnerability to fluctuations in C price.

What about full costWhat about full cost--benefit optimizationbenefit optimizationas favored by economists?as favored by economists?

• Integrated Assessment Models (e.g., Nordhaus, 1994; Manne-Richels, 1997; & Hammett, 1999): optimal path => shallower cuts in earlier years, deeper cuts late in century.

• True, scrapping coal-fired power plants today is costly, while credibly announcing stringent goals 50 years from now would be cheaper, by giving time to plan ahead.

• But benefit-cost maximization, though obviously right in theory, is the wrong logic in practice:– uncertainties are so big as to make cost/benefit estimates

worthless – Optimal path not credible <= Governments cannot bind their

successors.

Dynamic consistencyDynamic consistency• If the designers of a treaty specify a path for steep reductions

in the distant future,– their successor governments won’t pay the high economic costs to

follow through when the time comes. – Those making investment decisions today realize this inconsistency, &

act accordingly. – Clinton would have been laughed out of town if he had announced in

1997 that his Climate Change proposal was to do nothing until 2050 & then institute tough emission cuts in the second 1/2 of the century .

• Important to stand by precedents – In order to avoid the incentive for countries to stay out of an agreement

as long as possible – or to ramp up emissions in the meantime, – so as to be able to negotiate from a higher base .– That is why the 1997 treaty chose 1990 emissions as base-line.

Robustness under uncertaintyRobustness under uncertainty

• Cooper, Nordhaus, etc.: global carbon tax more robust under certainty.

• A carbon tax is my 1st choice for US domestic implementation, if politically acceptable.

• But not feasible internationally.• I still favor adding an escape clause or

safety valve to the quantitative system.

EquityEquity• Developing countries have contributed only 20

% of the CO2 accumulated in the atmosphere over past 150 years.

• They do not have the ability to pay for emissions abatement as rich countries do.

• Their governments have higher priorities.• India: “equity requires setting targets at equal

amounts per capita.”⇒ huge transfers, however. Not realistic. <= They are not in that strong a bargaining position.

EnforceabilityEnforceability

• No country will accept huge economic costs– Ex ante, won’t agree to treaty– Ex post, won’t abide by it.

• We want to set targets in future rounds– so that ex ante, costs re BAU are small– and ex post, costs are robust with respect

to unknown future developments