The Stern Review Stern Review & Its Critics & Its - - PowerPoint PPT Presentation

The The “ “Stern Review Stern Review” ” & Its Critics & Its Critics

Implications for the Theory & Implications for the Theory & Practice of BCA Practice of BCA

Daniel H. Cole

• R. Bruce Townsend Professor of Law (Not Economics)

Indiana University School of Law−Indianapolis

Outline of Paper Outline of Paper

Summarizes Stern Review and its findings Summarizes Stern Review and its findings Summarizes criticisms of Stern Review Summarizes criticisms of Stern Review Discusses implications for BCA theory and Discusses implications for BCA theory and practice practice

What the Stern Review Does What the Stern Review Does

Assesses costs and benefits of climate Assesses costs and benefits of climate change to 2200 (presuming BAU) change to 2200 (presuming BAU) Assesses costs and benefits of GHG Assesses costs and benefits of GHG mitigation/climate stabilization mitigation/climate stabilization Discusses policy options Discusses policy options (All this in (All this in only 600 pages!

• nly 600 pages!)

)

The Stern Review The Stern Review’ ’s Model s Model

PAGE2002 IAM PAGE2002 IAM

  Designed to assess not only mean expected

Designed to assess not only mean expected values of harm but values of harm but “ “high harm high harm” ” scenarios to scenarios to 2200 2200

  79 random variables in Monte Carlo analysis

79 random variables in Monte Carlo analysis

Including large-scale discontinuities Including large-scale discontinuities

  Inputs based on IPCC TAR

Inputs based on IPCC TAR

  Assumptions (values for Ramsey equation)

Assumptions (values for Ramsey equation)

ρ

ρ = 0.1% = 0.1%, , η η = 1, = 1, g g = 1.3%, = 1.3%, r r = 1.4% = 1.4%

The Stern Review The Stern Review’ ’s Findings s Findings

Cut emissions 25% by 2050 to stabilize Cut emissions 25% by 2050 to stabilize global climate at 500-550 global climate at 500-550 ppm ppm CO CO2

2eq

eq

  Cost: 1% of global annual GDP (

Cost: 1% of global annual GDP (≈ ≈ $1 $1 trillion/year) trillion/year) “ “forever forever” ”

  Benefits (costs of climate change avoided):

Benefits (costs of climate change avoided): 5% of global GDP 5% of global GDP “ “forever forever” ”

Recommended policies: Increase price of Recommended policies: Increase price of carbon by taxes or tradable quotas carbon by taxes or tradable quotas

Controversial Aspects of Stern Review Controversial Aspects of Stern Review

(1) Damage estimates greatly exceed those (1) Damage estimates greatly exceed those

• f other
• f other BCAs

BCAs

  E.g.,

E.g., Nordhaus Nordhaus & Boyer (2000); & Boyer (2000); Mendelsohn Mendelsohn et al. (1998); et al. (1998); Tol Tol (2002) (2002)

(2) Recommends earlier and more extreme (2) Recommends earlier and more extreme cuts in GHG emissions cuts in GHG emissions (3) Underlying (1) and (2): exceptionally low (3) Underlying (1) and (2): exceptionally low discount rate discount rate

Critiques of the Stern Review Critiques of the Stern Review

The Critics The Critics

  Nordhaus

Nordhaus, , Tol Tol, , Yohe Yohe, , Mendelsohn Mendelsohn, , Dasgupta Dasgupta and Weitzman and Weitzman

The Critiques The Critiques

  problems in damage estimates

problems in damage estimates

  ρ

ρ = 0.1% is too low = 0.1% is too low

  η

η = 1 is too low = 1 is too low

  r

r = 1.4% is too low = 1.4% is too low

  What about the

What about the g g? ?

Problems with Stern Problems with Stern’ ’s Damage Estimates s Damage Estimates

3 standard deviations higher than earlier 3 standard deviations higher than earlier estimates ( estimates (Yohe Yohe & & Tol Tol 2007) 2007) Assume pessimistic scenarios ( Assume pessimistic scenarios (Yohe Yohe & & Tol Tol 2007) 2007) Do not take adaptation into account ( Do not take adaptation into account (Mendelsohn Mendelsohn 2007) 2007) Assume large Assume large nonmarket nonmarket damages ( damages (Mendelsohn Mendelsohn 2007) 2007) Do not specify valuations of human life and other Do not specify valuations of human life and other nonmarket nonmarket effects effects No sensitivity analysis No sensitivity analysis

Is Is ρ ρ = 0.1% Too Low? = 0.1% Too Low?

Yes ( Yes (Nordhaus Nordhaus 2006; 2006; Yohe Yohe & & Tol Tol 2007) 2007)

  Much lower than any other climate change BCA (but see

Much lower than any other climate change BCA (but see Cline 1992) Cline 1992)

  People observed to use must higher rate (Weitzman 2007)

People observed to use must higher rate (Weitzman 2007)

  Explains Stern

Explains Stern’ ’s higher damage estimates s higher damage estimates At At ρ ρ = = 3% Stern 3% Stern’ ’s damage estimate reduced by 80-90% s damage estimate reduced by 80-90%

No No

  Discounting is generally unethical (Stern 2006; Ramsey

Discounting is generally unethical (Stern 2006; Ramsey 1928) 1928)

  Climate change is special case

Climate change is special case

Potential for reductions in consumption (

Potential for reductions in consumption (Dasgupta Dasgupta 2006; 2006; 1999) 1999) Like foreign aid (Schelling 2006) Like foreign aid (Schelling 2006)

Is Is η η = = 1 Too Low? 1 Too Low?

Yes ( Yes (Dasgupta Dasgupta 2006) 2006)

  Even after adjusting for risk, later generations

Even after adjusting for risk, later generations will be much wealthier than current generation will be much wealthier than current generation

  η

η = 1 implies ridiculously high savings rate = 1 implies ridiculously high savings rate (97.5%, assuming 4% rate of return on risk-free (97.5%, assuming 4% rate of return on risk-free investments) investments)

No ( No (Quiggin Quiggin 2006) 2006)

  Dasgupta

Dasgupta’ ’s s assumed rate of return on risk-free assumed rate of return on risk-free investments is unreasonably high; 1-2% investments is unreasonably high; 1-2%

• bserved in practice
• bserved in practice

It It’ ’s the s the r, r, smarties smarties! ( ! (Weitzman 2007) Weitzman 2007)

It is the combination of It is the combination of ρ ρ, , η η, , and and g g that matters for that matters for the Ramsey equation the Ramsey equation

  If

If ρ ρ = 2%, = 2%, η η = 2, and = 2, and g g = 2%, then = 2%, then r r = 6% = 6%

  If

If ρ ρ = 0%, = 0%, η η = 3, and = 3, and g g = 2%, then = 2%, then r r = 6% = 6%

Stern Review Stern Review’ ’s low s low r r is problematic is problematic

  “

“[ [S]trongly S]trongly against mainstream economics against mainstream economics” ”

  Ignores how people are observed to behave

Ignores how people are observed to behave

  Paternalistic: imposes analyst

Paternalistic: imposes analyst’ ’s own values s own values

Stern Stern’ ’s s r r may end up being may end up being “ “more right than wrong more right than wrong” ”

  Uncertainty over

Uncertainty over r r is the is the “ “most important uncertainty in the most important uncertainty in the economics of climate change economics of climate change” ”

Long-run uncertainty could reduce value of Long-run uncertainty could reduce value of r r from from ≈ ≈ 6% to 6% to ≈ ≈ 2% 2% (close to Stern figure) (close to Stern figure)

Or, Is It the Or, Is It the g g? (Weitzman 2007) ? (Weitzman 2007)

Climate change damages may not correlate well Climate change damages may not correlate well with aggregate economic activity with aggregate economic activity Low probability, high temperature increases could Low probability, high temperature increases could lead to lead to “ “low-g low-g” ” disasters disasters

  Knightian

Knightian uncertainty: we don uncertainty: we don’ ’t know how much we don t know how much we don’ ’t t know about know about “ “tail probabilities tail probabilities” ”

Weitzman recommends combining gradualism of Weitzman recommends combining gradualism of Nordhaus Nordhaus with more study of the with more study of the “ “low-g low-g” ” events on events on which predominate Stern which predominate Stern’ ’s (2006) analysis s (2006) analysis

Implications for Practice Implications for Practice

Special problems for Special problems for BCAs BCAs at the frontiers of at the frontiers of scientific knowledge and economic understanding scientific knowledge and economic understanding The political nature of The political nature of BCAs BCAs and the importance and the importance

• f sensitivity analyses
• f sensitivity analyses

The eternal issue: discounting The eternal issue: discounting Issues in Damage Estimation Issues in Damage Estimation

  The importance of transparent

The importance of transparent nonmarket nonmarket (including (including human life) valuations human life) valuations

  Taking seriously expected damages beyond the mean

Taking seriously expected damages beyond the mean

BCA as a dynamic, BCA as a dynamic, interative interative process process

Of Confidence and Caveats Of Confidence and Caveats

Ex ante Ex ante BCAs BCAs about climate change are about climate change are bound to be wrong bound to be wrong

  Long time horizons, incomplete scientific

Long time horizons, incomplete scientific information, high uncertainty information, high uncertainty

Stern (2006) overly confident of relatively Stern (2006) overly confident of relatively radical conclusions and recommendations radical conclusions and recommendations BCA authors should be parsimonious BCA authors should be parsimonious about conclusions and recommendations about conclusions and recommendations

  The greater the uncertainties, the greater

The greater the uncertainties, the greater should be the modesty of conclusions should be the modesty of conclusions

The Politics of The Politics of BCAs BCAs

Stern (2006) was as much a political Stern (2006) was as much a political document as an economic analysis document as an economic analysis But all But all BCAs BCAs are (more or less) political are (more or less) political

  Subjective elements (e.g., parameter values)

Subjective elements (e.g., parameter values) inevitably informed by author inevitably informed by author’ ’s political or s political or ideological predilections ideological predilections

Emphasizes importance of sensitivity Emphasizes importance of sensitivity analyses analyses

  Stern (2006) initially failed to include one

Stern (2006) initially failed to include one

On Discounting On Discounting

The choice of parameter values greatly The choice of parameter values greatly influences the outcome of influences the outcome of BCAs BCAs (Duh!) (Duh!)

  Choice of

Choice of ρ ρ alone is not critical; alone is not critical; r r is what counts is what counts

No consensus on a No consensus on a “ “best practice best practice” ” for choice for choice

• f discount rate (
• f discount rate (Portney

Portney & & Weyant Weyant 1999) 1999)

• It is unethical

It is unethical

• BCA authors cannot avoid

BCA authors cannot avoid imposing their own values imposing their own values

• Climate change is special case

Climate change is special case

• People do it

People do it

• It is

It is “ “conventional conventional” ”

• BCA authors shouldn

BCA authors shouldn’ ’t t impose their own values impose their own values

Anti discounting Anti discounting Pro discounting Pro discounting

Even If Climate Change is a Even If Climate Change is a “ “Special Case Special Case” ”

Stern wrongly ignored HM Treasury Stern wrongly ignored HM Treasury’ ’s s discounting schedule ( discounting schedule (Green Book Green Book) )

  Consequences

Consequences

creates impression that something fishy is going on creates impression that something fishy is going on unjustifiably erodes authority of policymakers unjustifiably erodes authority of policymakers

  Options

Options

use use Green Book Green Book values in alternative calculations values in alternative calculations explain why explain why Green Book Green Book values are inappropriate for values are inappropriate for the case of climate change the case of climate change

Toward a Toward a “ “Best Practice Best Practice” ” of Calculating

• f Calculating

under Alternative Discount Rates under Alternative Discount Rates

Becoming common in US regulatory Becoming common in US regulatory analyses analyses Same effect from independent Same effect from independent BCAs BCAs on

• n

same issue using different discount rates same issue using different discount rates

  e.g.,

e.g., Nordhaus Nordhaus & Boyer (2000) v. Stern (2006) & Boyer (2000) v. Stern (2006)

  The

The “ “marketplace marketplace” ” of

• f BCAs

BCAs

  But Stern (2006) carries inordinate weight

But Stern (2006) carries inordinate weight

imprimatur of the UK government imprimatur of the UK government Stern was policy-maker Stern was policy-maker

Estimating Damages Estimating Damages

BCA damage estimates must be based on BCA damage estimates must be based on

  Transparent and calculable valuations

Transparent and calculable valuations

  Consistently applied parameter values

Consistently applied parameter values

Problems with Stern Review Problems with Stern Review’ ’s damage s damage estimates estimates

  No way to assess accuracy of damage estimates

No way to assess accuracy of damage estimates

  Calculations can

Calculations can’ ’t be replicated t be replicated

Low-probability, high-value events at the tails Low-probability, high-value events at the tails

• f damage distribution curves should be
• f damage distribution curves should be

treated seriously (Weitzman 2007) treated seriously (Weitzman 2007)

BCA as a Dynamic, Iterative Process BCA as a Dynamic, Iterative Process

No BCA is an island; all No BCA is an island; all ex ante ex ante economic economic analyses are provisional analyses are provisional The importance of engaging and The importance of engaging and responding to critics responding to critics

  Stern added sensitivity analysis as postscript

Stern added sensitivity analysis as postscript

  Stern welcomed disputation

Stern welcomed disputation

Stern: Stern: “ “one contribution to [a] discussion

• ne contribution to [a] discussion”

”

Conclusion Conclusion

“ “We must be transparent and clear. If you take We must be transparent and clear. If you take little account of the interests of future generations little account of the interests of future generations you will care little about climate change. But ethical you will care little about climate change. But ethical positions cannot be dictated by policy analysts positions cannot be dictated by policy analysts…” …” (Stern 2006, postscript) (Stern 2006, postscript) Every [BCA] is an exercise in subjective Every [BCA] is an exercise in subjective

• uncertainty. If, as the
• uncertainty. If, as the Stern Review

Stern Review puts it, puts it, ‘ ‘climate climate change is the greatest externality the world has change is the greatest externality the world has ever seen, ever seen,’ ’ then a cost-benefit calculation of what then a cost-benefit calculation of what to do about it is the greatest exercise in Bayesian to do about it is the greatest exercise in Bayesian decision theory that we economists have ever decision theory that we economists have ever performed performed” ” (Weitzman 2007) (Weitzman 2007)