by Molly Macauley*, Joel Darmstadter*, Jhih-Shyang Shih*,

1. Measuring the Contribution to the Economy of Investments in Renewable Energy: Estimates of Future Welfare Gains by Molly Macauley*, Joel Darmstadter*, Jhih-Shyang Shih*, Emily Aronow*, David Austin**, and Tom Bath*** We thank the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, for their support. * Resources for the Future **now at the U.S. Congressional Budget Office *** independent engineering consultant http://www.rff.org/disc_papers/PDF_files/0205.pdf

2. Theoretical Framework Cost-index based measure of expected consumer welfare gains from innovation Use of counterfactual and its own technological change Adoption rate External effects Uncertainty (data, forecasts) Discount rate Our Application Electricity generation technologies: --Renewable (PV, solar thermal, biomass, wind, geothermal) --Fossil (Combined cycle gas turbine -- conventional and advanced designs) Two geographic regions (CNV, MAPP) Weibull (“fast,” “slow”) Carbon dioxide, thermal effluent Time period: 2000-2020 Approach

3. Derived Demand for Renewable Energy Technologies:Illustration of Net Surplus Change

4. Derived demand for renewable energy technologies: illustration of net surplus change with external costs

5. Derivation of estimating relationships C*dt= E* (udt, Pdt, Wdt) E* (udt, PI , WRE) and C*I = E* (uI, Pdt, Wdt) (1) E* (uI, PI , WRE) ½ ln (C*dtx C*I )=½ (sdt+sI) ln (Wdt/ WRE) (2) (Bresnahan, AER 1986) WRE =  WI + (1- ) Wdt (3)

6. Relationship between expenditures, cost index

7. Using the index to estimate the present value of consumer surplus Interpretation of the index: “how much better off are we (that is, society in general) as a result of investment in renewables, taking into account the alternative (conventional technology) and differences in the social benefits and costs between renewables and conventional technology?”

8. Model Framework Private Generation Costs PV, ST, GEO, BIO, Wind, CCGT, A-CCGT (DOE/EIA (2000), DOE/EPRI (1997), authors’ adjustments) • Externality Costs • Carbon (CCGT) (Krupnick et al. (1996)) • Thermal H2O (CCGT, Biomass, ST) (Authors’ estimates) Private and Social Generation Costs • “Market Conditions” • Adoption rates • Electricity prices (DOE/EIA (2000)) Cost Indices Private Consumption Expenditures (DOC (2001), authors’ forecasts) • Aggregate Consumer Surplus • Discount rate • Triangular and normal distributions combined with Monte Carlo draws characterize uncertainty.

9. Diversity of renewable energy resources in the United States.

10. Electricity Market Module Supply Regions Source: EIA

11. Weibull Adoption Rate Curves

12. Examples of External Effects • Carbon • Water • Land use • Avian and other ecological resources • Lifecycle (such as manufacturing) • Distinguishing pecuniary from technological effects

13. The present value of benefits from 2000 to 2020 for Wind Class 6 from scenario 1 for CNV

14. Key: % Confidence Interval 95% 75% 50% 25% 5% Discounted incremental net benefits from 2000 to 2020 for Wind 6 from scenario 1 for CNV

16. CNV Region Largest “loss” Fast adoption of PV w/ A-CCGT and no ext. $11.9 B Largest “gain” Fast adoption of WC6 w/ C-CCGT and ext. $4.6 B. ~ $556/household MAPP Region* $4.9 B. $4.4 B. ~ $600/household * No HGT, ST Illustrative Results (45 scenarios)MedianPresent Value, 2000-2020, 5% discount rate

18. Other Results • Externalities -- Carbon and thermal externalities enhance relative surplus associated with renewable technologies -- Thermal externality reduces relative surplus associated with ST, BIO (even though externality is also linked with CCGT) • Innovation -- Innovation in CCGT reduces median relative surplus associated with renewables by ~ 5%

19. Other results, continued • Uncertainty -- Uncertainty can lead to values +/- 20-40% of median at the 5% and 95% confidence intervals • Portfolio -- Equal portfolio weights create negative surplus estimates -- Variable portfolio weights can yield positive surplus estimates but smaller than single technology surplus estimates

20. Conclusions • Limits -- Pairwise or exogenously specified portfolio comparisons rather than optimization -- Data gaps (external effects) and assumptions (“GenCo”; deregulation; state/local policies) -- Gross not net of public and private investment or other expenditures to attain cost goals, adoption rates

21. Conclusions, continued • Findings -- Large differences among technologies -- Regional differences -- Adoption push -- Externality internalization -- Useful estimates result from model

22. Conclusions, continued • Contribution -- Offers conceptually grounded measurement approach; alternative to data-intensive econometric models; appeal of “cost index” analogy with Consumer Price Index; tool for program managers -- Allows for uncertainty, externalities, policy simulation -- Could extend to include “green preferences” (data? Are they verified?); state and local policies -- Could extend to NRC-defined benefits including commercialization, knowledge, option values