Environmental Economics Lectures 11, 12 Valuation and Cost-Benefit Analysis

Transcription

1 Environmental Economics Lectures 11, 12 Valuation and Cost-Benefit Analysis Florian K. Diekert April 9 and 30, 2014 Perman et al (2011) ch CON 4910, L11&12. Slide 1/ 28

2 Preview lecture 11 and 12 A Cost-Benefit Analysis (CBA) CON 4910, L11&12. Slide 2/ Choosing A over B or a heuristic definition of CBA 2. Cost-benefit analysis in a static and certain world Tests for marginal and non-marginal projects CBA and social welfare functions 3. Dynamic aspects 4. Accounting for uncertainty, risk, and irreversibility B Valuation 1. Theory Categories of environmental benefits WTP and WTA 2. Practice Stated preference methods (in part. contingent valuation ) Revealed preferences: Travel cost method, Hedonic pricing, Production function based techniques

3 A heuristic definition of cost-benefit analysis Carry out a project if E [NPV ] > 0 NPV = Net-Present Value = T 0 δt NB t NB t = Net-Benefits at time t = N i WTP i C In politics, relevant question is whether NPV A > NPV B. CON 4910, L11&12. Slide 3/ 28

4 Choosing A over B CON 4910, L11&12. Slide 4/ 28

5 Choosing A over B Formidable moral-philosophical problem. The Pareto-principle is nice, but impracticable: There will almost always be some that lose from a policy choice. Alternative: Kaldor-Hicks criterium (recall lecture 1): Situation A is better than situation B iff there is some hypothetical distribution of resources that could make everyone better of under A than under B. Drawback: Ignores distributional aspects! Advantage: CON 4910, L11&12. Slide 5/ Gives complete ordering 2. Guide the way to actual Pareto improvement 3. In random draw over many similar projects, it makes everyone better-off on average 4. Focus on increasing the pie, rather than fighting over how to allocate it

6 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) CON 4910, L11&12. Slide 6/ 28

7 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) Suppose the project is small relative to the rest of the economy. CON 4910, L11&12. Slide 6/ 28

8 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) Suppose the project is small relative to the rest of the economy. Let Q be evaluated at the efficiency prices P. CON 4910, L11&12. Slide 6/ 28

9 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) Suppose the project is small relative to the rest of the economy. Let Q be evaluated at the efficiency prices P. Then, go ahead if NB = N i P i Q i > 0 CON 4910, L11&12. Slide 6/ 28

10 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) Suppose the project is small relative to the rest of the economy. Let Q be evaluated at the efficiency prices P. Then, go ahead if NB = N i P i Q i > 0 This rests on a first-order Taylor approximation: Choose some vector X close enough to Q. Production possibility set is: {X PX PQ} Now Y = X + Q and {Y PY = P(X + Q) PQ + P Q = PQ + NB} PPF has been pushed out if and only if NB > 0 CON 4910, L11&12. Slide 6/ 28

11 Marginal project in a time-less and risk-less word Consider a project which implies the following change of the quantity vector : Q = ( Q 1, Q 2,..., Q N ) Suppose the project is small relative to the rest of the economy. Let Q be evaluated at the efficiency prices P. Then, go ahead if NB = N i P i Q i > 0 This rests on a first-order Taylor approximation: Choose some vector X close enough to Q. Production possibility set is: {X PX PQ} Now Y = X + Q and {Y PY = P(X + Q) PQ + P Q = PQ + NB} PPF has been pushed out if and only if NB > 0 Does not hold for non-marginal projects CON 4910, L11&12. Slide 6/ 28

12 CBA and social welfare functions CON 4910, L11&12. Slide 7/ 28

13 CBA and social welfare functions Define social welfare function: W = w ( U 1 (x 1, E),..., U N (x N, E) ) Project: environmental improvement de at cost per person of C i = dx i dw = ( Ui w i dx i + U ) i x i E de i = ( U i w i C i + U ) i/ E de x i U i / x i i = i w i U i x i (WTP i C i ) U If welfare weights w i i x i are the same for all, standard CBA ranks projects according to social welfare CON 4910, L11&12. Slide 8/ 28

14 CBA and social welfare functions w i are a purely normative choice U i x i (marginal utility of an extra unit of income) are descriptive, but not observable U The assumption that w i i x i is the same for all cannot be verified. Unweighted utilitarianism: W = U U N w i = 1 for all i. If U i x i j s. If U i x i CON 4910, L11&12. Slide 9/ 28 > U j x j : equal w i implies larger weight on i s WTP than > U j x j : standard CBA (equal weight on everyone s WTP) implies w j > w i Standard CBA systematically favors those who care little about money at the margin

15 Dynamic aspects CON 4910, L11&12. Slide 10/ 28

16 Dynamic aspects Comparisons between different people living at the same time is difficult: still harder to compare utility of different people living at different times! Especially worrisome when the points in time are far apart: Will preferences stay the same? Which discount rate to use? How much richer will future generations be? Long-lived stock pollution problems are extremely sensitive to the discount rate. At a 1% discount rate, we would be willing to pay ca NOK to avoid a damage of 1 million NOK in a century. At a 3% discount rate we would pay up to 52 thousand NOK, at a 10% discount rate, we would not even pay 73 NOK. CON 4910, L11&12. Slide 11/ 28

17 Uncertainty, risk, and irreversibility CON 4910, L11&12. Slide 12/ 28

18 Uncertainty, risk, and irreversibility Irreversibility is a buzzword. People rather have the option to change course in the future: Introduce quasi-option value (following the notation in Traeger,REE, Define the PV under anticipated learning by: [ { } ] V l (0) = u 1 (0) + E max x 2 {0,1} [ u 2 (0, x 2, θ) ] V l (1) = u 1 (1) + E u 2 (1, 1, θ), the PV under the possibility of postponement: { [ ]} V p (0) = u 1 (0) + max E u 2 (0, x 2, θ) x 2 {0,1} [ V p (1) = u 1 (1) + E u 2 (1, 1, θ) ], and the PV of a now or never decision: [ ] V n (0) = u 1 (0) + E u 2 (0, 0, θ) [ ] V n (1) = u 1 (1) + E u 2 (1, 1, θ) CON 4910, L11&12. Slide 13/ 28

19 Uncertainty, risk, and irreversibility Using these definitions, we see how the NPV decision rule has to incorporate the QOV while the full value of sophistication additionally incorporates the SOV. We have: NPV = V n (1) V n (0) > 0 QOV = (V l (0) V l (1)) (V p (0) V p (1)) = V l (0) V p (0) FVS = V l (0) V n (0) = V l (0) V p (0) + V p (0) V n (0) }{{}}{{} QOV SOV CON 4910, L11&12. Slide 14/ 28

20 Uncertainty, risk, and irreversibility Decision rules under uncertainty 1. maximin 2. maximax 3. minimax regret 4. assign subjective probabilities CON 4910, L11&12. Slide 15/ 28

21 Review Part A: Cost-Benefit Analysis 1. A heuristic definition of cost-benefit analysis 2. Detour: The moral of choosing A over B 3. Cost-benefit analysis in a static and certain world Tests for marginal and non-marginal projects CBA and social welfare functions 4. Dynamic aspects 5. Accounting for uncertainty, risk, and irreversibility CON 4910, L11&12. Slide 16/ 28

22 Preview Part B: Valuation 1. Theory Categories of environmental benefits WTP and WTA 2. Practice Stated preference methods The method of contingent valuation (CV) Discussion: (http: // Revealed preferences Travel cost method Hedonic pricing Production function based techniques CON 4910, L11&12. Slide 17/ 28

23 Valuation: Theory CON 4910, L11&12. Slide 18/ 28

24 Categories of environmental benefits In the old edition of the Perman book (2003), there were a lot of acronyms, summarized in two equations: EC = UV + EV + OV + QOV TV = CUV + NCUV + NUV CON 4910, L11&12. Slide 19/ 28

25 Willingness-to-pay & willingness-to-accept Environmental valuation theory 1 Assume that quantity/quality of environmental good e can be treated as an argument in a wellbehaved utility function. The individual cannot choose level of e. y is income. u = u(y, e) e0 e1 At A, WTP for e improvement = BC is Compensating Surplus, CS At A, WTA in lieu of e improvement = DA is Equivalent Surplus, ES e1 e0 At B, WTP to avoid deterioration = BC is Equivalent Surplus, ES At B, WTA compensation for deterioration = DA, Compensating Surplus, CS Figure: ECON 4910, L11&12. Slide 20/ 28 Slide taken from Perman s webpage

26 Willingness-to-pay & willingness-to-accept CON 4910, L11&12. Slide 21/ 28

27 Compensating and Equivalent variaton Compensating variation is the change in income that would compensate for the price change Equivalent variation is the change in income that would be equivalent to the price change CON 4910, L11&12. Slide 22/ 28

28 Compensating and Equivalent variaton Compensating variation is the change in income that would compensate for the price change Equivalent variation is the change in income that would be equivalent to the price change Two necessary conditions for environmental quality changes to be approximately identified by price changes in demand functions: Non-essentialness Weak complementarity CON 4910, L11&12. Slide 22/ 28

29 Valuation: Practice CON 4910, L11&12. Slide 23/ 28

30 Stated preference methods: Contingent Valuation Excerpt of NOAA-panel recommendation: Sample from entire affected population Personal interview Pre-testing WTP instead of WTA Referendum format Reminder of undamaged substitutes No-answer option available CON 4910, L11&12. Slide 24/ 28

31 Contingent Valuation: Discussion CON 4910, L11&12. Slide 25/ 28

32 Revealed preference methods 1. Travel cost method CON 4910, L11&12. Slide 26/ 28

33 Revealed preference methods 1. Travel cost method 2. Hedonic pricing CON 4910, L11&12. Slide 26/ 28

34 Revealed preference methods 1. Travel cost method 2. Hedonic pricing 3. Production function based techniques CON 4910, L11&12. Slide 26/ 28

35 Review lecture 11 and 12 A Cost-Benefit Analysis (CBA) CON 4910, L11&12. Slide 27/ Choosing A over B or a heuristic definition of CBA 2. Cost-benefit analysis in a static and certain world Tests for marginal and non-marginal projects CBA and social welfare functions 3. Dynamic aspects 4. Accounting for uncertainty, risk, and irreversibility B Valuation 1. Theory Categories of environmental benefits WTP and WTA 2. Practice Stated preference methods (in part. contingent valuation ) Revealed preferences: Travel cost method, Hedonic pricing, Production function based techniques

36 Preview next (and last) lecture 1. Private contributions to public goods (Nyborg & Rege, 2003) Discuss theoretical predictions and empirical evidence for various alternative behavioral models Special focus on the role of public policy ( crowding out / crowding in ) 2. Course synopsis CON 4910, L11&12. Slide 28/ 28