New version page

MIT 15 023J - Effect of Uncertainty and Learning on Decisions

Upgrade to remove ads
Upgrade to remove ads
Unformatted text preview:

Course so far…OutlineMotivation I The “Question”Why Does Uncertainty Matter?Why Does Uncertainty Matter?What is “Risk-Aversion”?Risk-Averse Utility FunctionWhy Does Learning Matter?Arrow-Fisher IrreversibilityIrreversibilities in Climate ChangeDICE Model of Climate and EconomyDICE (Continued)Focus of this Study: Climate SensitivityCurrent Uncertainty in Climate SensitivityAlternative Analysis FrameworksEffect of Learning on Efficient PolicyEffect of Learning under Cost-BenefitHedging Policy Under Uncertainty Cost-Benefit CaseWhat Determines the Optimal Hedge?Decision-Making under UncertaintyIllustration: Optimal Carbon Tax for Temperature Target of 2oCRealized Temperature Change for 2 Degree TargetQUESTION: What should we do now if we are uncertain?Relative to Best Policy in Each Case…If We Learn in 2020 Simulating Learning: Stochastic ProgrammingIf We Learn in 2030If We Learn in 2040If We Never LearnSummary – Effect of Learning Later with 2o targetEffect of Learning for Several Temperature TargetsDecision under uncertainty with partial learningEffect of Partial Learning on Optimal HedgingSummaryMIT OpenCourseWare http://ocw.mit.edu15.023J / 12.848J / ESD.128J Global Climate Change: Economics, Science, and PolicySpring 2008For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.Effect of Uncertainty and Learning on DecisionsMort WebsterMIT Joint Program on the Science and Policy of Global Change15.023-12.848-ESD.128April 23, 2008Course so far…• Projecting climate impacts• Costs of GHG emissions reductions• Analysis under certainty• Descriptions of uncertainty– Probability distributionsOutline• Motivation: Why uncertainty matters• Effect of learning on decisions• Application of decision analysis to climate policyMotivation IThe “Question”Since climate change is so uncertain, shouldn’t we just wait until we know more?Why Does Uncertainty Matter?Revenues ($ Million)0 20406080100Probability Density0.000.050.100.150.200.250.30If $10 million investment is required, would you do it?What about $60m?μ= $30mWhy Does Uncertainty Matter?• If no learning is possible and no risk aversion– Make decision based on expected value• If you can learn and revise along the way– May want to do more or less at first (hedging)• If you are risk-averse– You care about more than mean outcomesWhat is “Risk-Aversion”?Choose:A) A gamble with 50% chance of paying $100and 50% chance of paying $0orB) Pay $49 for sureRisk-Averse Utility FunctionMonetary outcome (Billion $)02468100.00.20.40.60.81.0UtilityRisk-NeutralRisk-AverseWhy Does Learning Matter?• Why would you do something different today if you can learn tomorrow?• Answer: if the outcome is irreversibleArrow-Fisher Irreversibility• Problem: – Two time periods t=1, 2– Total forest area = 100– Cost of cutting forest C(x)– Benefits of cutting forest B(x) - UNCERTAIN– Choose x1, x2to Max (B-C).•What isx1if you can’t learn?•What isx1if you can learn between t1and t2?Irreversibilities in Climate Change• GHG Concentrations– Temperature Change– Climate Damages• Capital Stock / Economic InvestmentsDICE Model of Climate and Economy• Simple Model of Economy:– Capital and Labor Make GDP– GDP split between Savings and Consumption–CO2Emissions from Production• Simple Climate Model– Carbon Cycle – Emissions to Concentrations– Radiative Forcing– Energy BalanceDICE (Continued)• Policy Variable: – Reduce Emissions by Fraction (μ)– Abatement Costs C = f(μ)– Damage Costs D = f(ΔT)• Maximize Discounted Utility (Consumption)• Two Types of Analysis– Cost-Benefit (balance abatement costs with damage costs)– Cost-Effectiveness (set absolute constraint)Focus of this Study:Climate Sensitivity• One of the critical uncertainties•Defn:– Amount of global mean temperature change from a doubling of CO2at equilibrium.• Meaning: – Represents net effect of feedbacks in the atmosphere.Current Uncertainty in Climate SensitivityClimate Sensitivity (oC)0246810Probability Density0.000.050.100.150.200.250.30Alternative Analysis Frameworks• Cost-Benefit– Use damage function to monetize climate impacts– Find economically efficient path of abatement• Cost-Effectiveness– Pick some target (e.g., concentration or temperature stabilization)– Find least cost way to meet targetEffect of Learning on Efficient Policy2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)050100150200Perfect InformationNever LearnComplete Resolution in 2040CS=8.0CS=5.0CS=3.0CS=1.5Effect of Learning underCost-BenefitTime When Learning Occurs2020 2030 2040 2050 2060 neverOptimal Tax in 201517.017.217.417.617.818.0Hedging Policy Under Uncertainty Cost-Benefit Case2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)050100150200Perfect InformationNever LearnWhat Determines the Optimal Hedge?Mean of Climate Sensitivity (oC)234567Optimal Carbon Tax in 2015 ($/ton)101520253035Answer: The Shape of the Probability DistributionDecision-Making under UncertaintySome Simple Starting Points:1) What should you do if you KNOW?2) What should you do if you will NEVER LEARN?3) What should you do if you don’t know, but WILL LEARN at time T?4) What should you do if you don’t know and will reduce your uncertainty at time T?2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400CS = 1.5CS = 3.0CS = 5.0CS = 8.0Illustration: Optimal Carbon Tax for Temperature Target of 2oC2000 2020 2040 2060 2080 2100 2120 2140Global MEan Surface Temperature Change (oC)0.00.51.01.52.02.53.0CS = 1.5CS = 3.0CS = 5.0CS = 8.0Realized Temperature Change for2 Degree TargetQUESTION: What should we do now if we are uncertain?• Wait (do nothing until we know more)?• Implement Highest Tax (worst case)?• Implement Lowest Tax (best case)?• Something in the Middle? – Where in the middle?Relative to Best Policy in Each Case…2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400Perfect InformationIf We Learn in 2020 2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400Perfect InformationLearn in 2020Simulating Learning:Stochastic Programming2000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400CS = 1.5CS = 3.0CS = 5.0CS = 8.0If We Learn in 20302000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400Perfect InformationLearn in 2020Learn in 2030If We Learn in 20402000 2020 2040 2060 2080 2100Carbon Tax ($/ton)0100200300400Perfect InformationLearn in 2020Learn in


View Full Document
Download Effect of Uncertainty and Learning on Decisions
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Effect of Uncertainty and Learning on Decisions and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Effect of Uncertainty and Learning on Decisions 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?