Robert Topel. The Prospec+ve Social Value of a New Good: Hydrogen Fuel-Cell Electric Vehicles

Transcription

1 The Prospec+ve Social Value of a New Good: Hydrogen Fuel-Cell Electric Vehicles Robert Topel Booth School of Business and Energy Policy Ins+tute University of Chicago May, 2016 Preliminary and incomplete. For discussion purposes only. Do not quote.

2 Goal of This Project Calibrate the prospec+ve social value of FCEVs Value to consumers plus other social gains, such as reduced CO2 emissions Applica+on today is for the US, but could be extended to other countries Rely on technical/cost/market assump+ons from a variety of sources

3 FCEVs: Four Sources of Prospec+ve Social Gain Consumers Surplus what buyers would gain directly from the product innova+on. Two components FCEVs have dis+nct characteris+cs that may be valued. Buyers environmental concerns, driving characteris+cs, being unique Possible direct fuel cost saving rela+ve to ICVs Reduced carbon emissions Social value depends on social cost of carbon and net world-wide reduc+on in emissions due to FCEV penetra+on Monopsony gain to impor+ng countries from lower oil price Reduced US demand for oil reduces the world price of oil The US is a net importer of oil net gain from lower price Oil Security gain from reduced oil use Oil supply shocks affect na+onal income in propor+on to oil usage.

4 Consumers Surplus O\en studied ques+on: what is the social welfare gain from a par+cular new product innova+on? What do consumers gain from the new good? Cell phones, CT scans, mini-vans, etc. Almost all exis+ng studies are ex-post conceptual experiments: A\er the innova+on occurs and people use (buy) the product, how much would they have to be paid to do without it? Consumers Surplus (CS): The amount of money that consumers would require as compensa+on for removing a par+cular good from their set of choices CS exists because of product differen+a+on the good offers something of value that other available goods do not Cell phones v telephones Mini-vans v sta+onwagons

5 Consumers Surplus: Basics The social gain from a new good is the difference between its value to users and their cost of acquiring it Consumers Surplus: the area below demand (value) and above price (cost) on all units purchased Es+mated from knowledge of how buyers respond to varia+ons in price, and assump+ons about the shape of the demand curve $/Q Consumers Surplus ($/yr) Expenditure = PxQ* Q* Price V(Q) Q

6 Benefits Must Offset Costs of Building, Maintaining and Opera+ng a Fueling Network Costs of supplying H include produc+on (either centrally or distributed) from several possible feedstocks (e.g. natural gas), plus capital costs of infrastructure Chicken & Egg network externality drivers will only purchase if there is a fueling network of sufficient density Sta+ons will only be built if there are enough drivers May require subsidies and coordinated investment by FCEV producers to succeed

7 Externali+es and Market Failure : The Case for Public Interven+on Externali+es: benefits or costs from individuals ac+ons that are borne by others Here, externali+es exist on both the demand and supply sides Demand: Choices by consumers FCEVs reduce carbon emissions FCEVs many improve energy security FCEVs would reduce the price of oil benefit to all oil buyers Supply: Network effects may impede building a network Even if a FCEV plahorm of vehicles and sta+ons would be socially beneficial, these externali+es mean that public interven+on (subsidies, taxes or regula+ons) may be necessary to create/maintain the plahorm Private investment and private choices may fail to produce the efficient outcome

8 Step 1: Measuring Prospec+ve Consumers Surplus FCEV plahorm for consumers does not exist cannot rely on direct evidence from purchasing decisions Make inferences from related vehicle innova+on that consumers valued rela+ve to alterna+ves HEVs: similar aiributes displacing ICVs Use HEVs as benchmark for FCEVs Several studies of HEV demand and value Kahn (2007), Beresteanu and Li (2011), Sallee (2008), Gallagher and Muehlegger (2011), Furlong (2012a, 2012b). Evidence that HEVs provide differen+ated value due to: Perceived green characteris+cs lower emissions and less use of fossil fuels. (Kahn, Furlong (2012b)) Fuel cost savings HEV demand strongly related to price of gasoline. (Beresteanu and Li, Furlong (2012a))

9 Step 1: HEV Benchmark Example Evidence in Beresteanu & Li (2011) and Furlong (2012) indicate conserva+ve es+mate of surplus/hev of about C =$7000. This is a present value per vehicle This must consist of a component due to product differen+a+on (C*) and a component due to fuel cost saving (F) over the discounted expected life (L) of a vehicle: * C HEV = C HEV + F HEV F HEV = P G [G IC G HEV ]L = P G G IC 1 m IC m HEV L L = a=0 S(a)e (β r)a da Es+mate L from EPA/NHTA data on vehicle survival rates S(a). Use this to es+mate F HEV from mileage advantage of HEVs vs. ICVs: m IC < m HEV Boiom line: At $2.60/gallon and 12,000 miles per year, F $3,800 Then C* = C F = $3200 in surplus per HEV aiributable to product differen+a+on. Use this as benchmark es+mate of C* for FCEVs

10 Step 1 (con+nued): Consumers Surplus from FCEVs Calculate expected fuel cost saving (if any) from FCEV vs ICV F FC = [P G G IC P H K FC ]D = P G G IC [1 P H P G K FC G IC ]D Use range of DOE forecasts of dispensed price P H, including costs of capital for fueling network. Then surplus per FCEV equals the HEV benchmark plus F FC * C FC = C FC + F FC P P H G price/k hydrogen price/g gasoline K = required kilograms for M miles G = = = required gallons for M miles * = C HEV + P G G IC [1 P H P G K FC G IC ]D = $ P G G IC [1 P H P G K FC G IC ]D

11 Step 1: (Con+nued) Absent substan+al subsidies, two reasons FCEVs will generate less surplus and (much) lower sales than HEVs The acquisi+on cost of FCEVs ( $60K in 2016) is substan+ally higher than typical HEV ( $30K) DOE (2012) es+mates delivered cost of H = $4.49/kg. At that price H is cost compe++ve w/ gasoline on $/mile basis The current price of H ( $10- $13/kg in 2016) is higher than DOE projec+ons, so fuel costs will ini+ally be higher than ICVs Large subsidies will be required to achieve market penetramon on the scale of HEVs, even assuming equal valuamons by consumers: Vehicle: $30K/vehicle in 2020, assumed declining over +me Fuel: $23K/vehicle in 2020 (present value), declining over +me

12 Step 2: Valuing Reduced Carbon Emissions FCEVs will displace some quan+ty of gasoline consump+on If H is produced with zero emissions (e.g. with carbon sequestra+on) then FCEVs would be ZEVs well to wheels (WTW) If not ZEV, then must calculate net displacement of emissions in US Assume SMR produc+on from natural gas, variable carbon footprint Next must es+mate how US displacement affects world emissions Extreme example: If world oil supply is fixed, then FCEVs cannot reduce emissions because the same amount of oil is used with or without FCEVs Overall impact on emissions depends on how world oil markets adjust. If not ZEVs, FCEVs could increase world emissions Also require a price of carbon the social benefit of a one ton reduc+on in world carbon emissions E.g., EPA now uses $47/ton as present discounted value to the world of a one ton reduc+on in current emissions Es+mate is highly uncertain, based on 4 Integrated Assessment Models

13 Step 3: Monopsony Benefits From Reduced Oil Prices The US accounts for roughly 20% of world oil demand and imports 30-50% of its oil consump+on of 6.7 billion bbl/yr Even a small reduc+on in US demand will reduce the world price of oil, crea+ng a wealth transfer from oil producers to oil consumers. US (Japan, EU) benefits because it is a net importer of oil. Effect on world oil price from change in US demand: D sgas ηoil % Δ P = k % ΔD D S D η η + η [ ] oil us US gas oil oil s gas = share of gasoline in value of barrel η j D = demand elasticity for product j η j S = supply elasticity for product j %ΔD US = percentagechange in demand for oil in US k US = US share of world oil use

14 Step 3 Monopsony Gain (con+nued) The reduc+on in US demand is the amount of oil consump+on displaced by FCEVs in the US The elas+ci+es of demand and supply are taken from exis+ng econometric literature on world oil and gasoline markets The reduc+on in price applies to net oil imports to the US US buyers gain more than US producers lose

15 Step 4: Security Benefits of Reduced Oil Demand Oil is an input to US na+onal income. Supply and/or price shocks cause corresponding income changes. Consumers & firms pay more for gasoline and other oil uses, reducing welfare & percapita income Magnitude of income impact from a given shock has declined over +me as oil share of na+onal income has declined. Security cost per barrel depends on Frequency (arrival rate) of oil supply shocks Typical size of supply shocks Sensi+vity of na+onal income to a typical shock Current policy es+mate from NRC/EPA/NHTA is that each barrel of US consump+on generates a security cost of about $7. So I assume reduc+on in use generates a $7/barrel benefit.

16 Framework Summary: Net Social Benefits of FCEVs Depend On: 1. Consumers Surplus from new vehicle type Product differen+a+on unique characteris+cs valued by consumers. HEVs as benchmark Fuel cost saving compared to alterna+ve vehicles (projected cost of H produc+on in here) FCEVs and Hydrogen will require subsidies to be compe++ve. These must be subtracted from gross surplus 2. Social value of possibly reduced carbon emissions Net effect on aggregate (world) emissions +mes social cost of carbon. May be either posi+ve or nega+ve 3. Monopsony effect of reduced US oil demand on world oil prices US is net importer: net wealth transfer from foreign producers 4. Energy security benefit of reduced domes+c oil consump+on Smaller impact of supply shocks on US economy

17 Simula+ons of Social Value The net social benefits of FCEVs depend on a number of assump+ons about future markets and consumers valua+ons The simula+ons that follow provide illustra+ve examples of social benefits under various assump+ons: Consumers value of FCEVs Social cost of carbon Vehicle and fuel costs Vehicle and fuel subsidies Rela+ve emissions of H vs. gasoline Elas+ci+es of demand/supply in oil markets

18 Base Case (Not Realis+c!) FCEVs immediately available at same price as HEVs ( $30,000) Hydrogen immediately available at $4.50/kg Fueling network is not a constraint Demand for FCEVs grows to equal share of HEVs by 2035

19 Base Case Assump+ons: Vehicle & Fuel Costs Roughly Equivalent to HEVs 1. New car market share. At similar cost as HEV, penetra+on will grow like HEVs: 0.25% in 2020 grow to 5.8% by Ini+al (2020) Price of FCEV: $30, P H = $ CS/vehicle at full HEV cost parity: $7000 Elas+city of demand for vehicles: Expected vehicle life: 14.8 years 6. Expected Discounted life: 9.3 years Consumer discount rate: Gasoline share of refined cost: LR elas+city of demand for gasoline: LR elas+city of demand for oil: LR elas+city of supply of oil: kg/gal for given miles: Price of gasoline in 2020: $ No subsidies to vehicle and fuel achieve cost parity with HEVs 14. Lbs. CO 2 /gallon: Carbon cost/mt: $ Growth rate of carbon cost: Rela+ve emissions: (e/k)/(e/g): Security benefit per barrel: $ Annual miles/vehicle: 15, US oil import share: US annual oil consump+on: 6.75 billion bbl. 22. US daily oil consump+on: 18.5 million bbl. 23. Crude price/bbl 2020: $ Growth rate of crude oil price: World oil consump+on/day: 90.3 million bbl 26. US share world oil: 20.49% 27. FCEVs per fueling sta+on: Dispensers/sta+on: Capital cost/sta+on: $3.0 million 30. Cost of capital (interest & deprecia+on): 15%

20 NB: Overall gain is fairly small. $3.5B/year by 2040 Project Social PV: $22.4B Carbon gains are (slightly) nega+ve Slightly higher emissions because of emissions producing H and small elas+city of supply

21 Case 1 Unsubsidized Introduc+on P(FCEV)=$60,000 in 2020 Declines to $30,000 in 15 years (2035) P(H)=$10/kg in 2020 Declines to $4.50 in 15 years No subsidies for either vehicle or fuel => Quan++es will be very low at beginning because vehicle & fuel are very expensive FCEVs and H eventually require no subsidies to be as airac+ve as HEVs

22 NB: Overall gain is small. $3.25B/by 2040 Project Social NPV: $17.4B (compared to $22.4B in base case) Carbon gains are slightly nega+ve, as before

23 Case 2a & 2b Reduced CO2 emissions/kg 2a: (e/k)/(e/g) =.5 2b: (e/k)/(e/g) = 0

24 With lower emissions in produc+on, some carbon gain Project Social PV = $19.3B Case 2a: (e/k)/(e/g)=.5

25 With zero emissions in produc+on carbon gain reaches $0.75B by 2040 Project Social NPV = $21.4B

26 Things to Note About Cases 1 & 2 No subsidies => sales are very low for 15 years, then rise when assumed cost parity with HEV occurs. May be unrealis+c if scale & learning by doing required for cost decline Sales (2040) = 4.6% of new vehicle sales Elas+city of oil supply is realis+cally small (0.3) so carbon effect is +ny actually negamve due to emissions from producing H Largest component of LR gain (>80%) is consumer surplus Reducing the CO2 footprint produces material carbon gains and increases the social PV of the project

27 Net Social Gains with Subsidies Assume: Manufacturer and government subsidies reduce net vehicle cost to equal HEVs Subsidies for fuel also make PV of fuel cost equal to HEVs Note: Subsidies are costs. Must be subtracted from other benefits to obtain the net social gain

28 Case 3: Vehicle & fuel subsides make FCEVs cost-equivalent with HEVs for consumers Project NPV=$3B Project NPV = -$3.4B

29 Case 3: Components of Gross Gains

30 Things to Note About Case 3 Subsidies are large because costs of H and vehicle are ini+ally high rela+ve to HEV P(H) = $10/kg => PV fuel subsidy = $12,346/vehicle in 2020 P(FCEV)=$60K => vehicle subsidy = $30,000 in 2020 Largest component of LR gain (>80%) is consumer surplus CS (2040) = $4B/year Monopsony & Security (2040) = $0.88B Carbon (2020) = -$0.06B (nega+ve) Project NPV = $3B

31 Case 4: Gasoline Price in 2020: $4/g Project NPV = $42B

32 Very Op+mis+c Case P(gasoline) = $4 in 2020 (e/k)/(e/g) = 0.5 (down from 0.9) P(carbon)= $100/ton Elas+city of world oil supply = 0.5 FCEV market share = 20% in 2035

33 NPV = $225B

34 Project Social NPV = $225 Billion Substan+al CO2 gains because cost/ton is high ($100) and oil supply is more elas+c