Costs and Beneregts of Electric Vehicles

A 2010 Perspective

Fredrik Carlsson and Olof Johansson-Stenman

Address for Correspondence Fredrik Carlsson Department of Economics GoEgrave teborgUniversity Box 640 SE-40530 GoEgrave teborg Sweden Olof Johansson-Stenmanis at the same

Department The authors are most grateful for useful comments and suggestions from

David Bauner Bengt Johansson Urban KarlstroEgrave m the Journalrsquos Editor and in particulara very helpful anonymous referee Financial support from the Swedish Agency for Inno-

vation Systems (VINNOVA) is gratefully acknowledged

AbstractThis paper undertakes a social cost-beneregt analysis regarding an increase in the number of

electric vehicles in the Swedish transport sector by year 2010 Battery cars are generally

found to be socially unproregtable even though their private life-cycle costs and externalcosts are lower than those of petrol cars One important reason for this is that electric

vehicles are heavily `subsidisedrsquo by having in comparison with taxes on fossil fuel a very

low electricity tax `Hybridrsquo cars are more likely to be socially proregtable especially for city-based delivery trucks which may be both privately and socially proregtable without

subsidies

Date of receipt of regnal manuscript August 2002

Journal of Transport Economics and Policy Volume 37 Part 1 January 2003 pp 1plusmn28

1

Introduction

Despite their environmental advantages Electric Vehicles (EVs) are stillquite rare and various ways of supporting them regnancially are oftenproposed in public debate One argument is that even though EVs may notbe socially beneregcial today they probably will be tomorrow and that theintroduction of EVs is a long process that must start today The mainpurpose of this paper is to regnd out whether it is likely that EVs will beprivately and socially proregtable in the foreseeable future and henceindirectly to provide relevant information for policy decisions todayTherefore we undertake a social cost-beneregt analysis (CBA) of di erentkinds of electric vehicles including hybrid vehicles and fuel-cell vehicles inSweden for the year 2010 Such a project involves large uncertainties andmany simplireged assumptions Nevertheless we believe that order-of-magnitude estimates along with a detailed presentation of the assump-tions are valuable from a policy-making perspective

Considering the role of EVs in public debate about transport and theenvironment it is somewhat surprising that until recently so few sys-tematic CBAs of EVs have been undertaken Kazimi (1997ab) using amicro-simulation model estimated the environmental beneregts of intro-ducing battery-driven electric cars (BCs) in the US The cost side was notanalysed but based on her results she judged that large price-reductions ofalternative-fuel vehicles would not be socially beneregcial Funk and Rabl(1999) found that the life cycle costs for BCs in Greater Paris (France)were higher than the corresponding costs for petrol cars but were lowerthan the costs for diesel cars Hahn (1995) discussed the cost e ectivenessof various measures to improve environmental quality in the transportsector His main conclusion was that tighter air pollution standards andimproved fuel qualities were more cost-e cient than for example anintroduction of BCs Similarly Wang (1997) in a survey of American cost-e ectiveness studies concluded that zero-emission vehicles were among theleast cost-e ective alternatives More recently Lipman et al (2000) andOgden et al (2001) have presented detailed social cost estimates of fuel-celland hybrid vehicles whereas Delucchi and Lipman (2001) based ondamage costs from Delucchi (2000) provides detailed and comprehensivelife-cycle cost estimates for BCs and concludes that battery performancemust improve and costs be reduced substantially if BCs are to becomecompetitive1

1 Delucchi (2001) is another source of detailed life-cycle cost estimates

Journal of Transport Economics and Policy Volume 37 Part 1

2

This study presents less detailed information on the manufacturing andmaintenance costs trying instead to capture reasonable order-of-magni-tude estimates of all relevant cost elements while simultaneously high-lighting the general equilibrium e ects of changed tax revenues This isparticularly important in countries with high fuel taxes as is the case inmost European countries A fairly common view seems to be thatalthough BCs are not privately proregtable they are or will be sociallyproregtable when the external costs are taken into account However as willbe demonstrated it may actually be the opposite that is that BCs maybecome privately proregtable (in certain market segments) but sociallyunproregtable when tax-revenue e ects are taken into account despite thefact that BCs are environmentally superior

The remainder of this paper is organised as follows In Section 2 atheoretical CBA model is outlined in a stylised general equilibrium modelThis model deals with how to handle welfare e ects of changed externalcosts tax revenues from indirect general taxation as well as targetedexternality-correcting taxes and excess burden e ects associated withnegative revenue e ects of public subsidies In particular it is shown thatunder certain frequently made separability assumptions of the utilityfunction one can disregard excess burden e ects of the lost tax revenuesassociated with public subsidies of for example electric vehicles Thesubsequent CBA will consequently disregard such e ects and is hencequite ``traditionalrsquorsquo in this sense Section 3 presents and discusses the largenumber of assumptions made Section 4 presents the results of the CBAwhile Section 5 provides concluding remarks

The Cost-Beneregt Model

We use a standard cost-beneregt model where we are primarily interested inconsumer welfare or utility Although the standard CB model has beenused in numerous applications in the transport sector some features of themodel remain unclear (at least in the Swedish policy debate) (i) Howshould the correction for changed tax revenues be made and in particularshould one include value added taxes (VAT) that are based on other taxes(such as energy taxes or environmental taxes)2 (ii) Should one correct for

2 Energy taxes and environmentally related fuel taxes are currently part of the VAT tax

base in Sweden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

3

excess-burden e ects and if so how To deal with these issues we apply asimple representative-individual3model with the following utility function

U D u(l x c e D) (1)

where l is leisure x is a composite good or put simply money spent oneverything except road transport c and e are the consumption of con-ventional and electric-vehicle based road transport respectively and D isenvironmental damage and where D D D(C E) C D

PniD1 ci D nc and

E DPn

iD1 ei D ne We assume that ul gt 0 ux gt 0 uc gt 0 ue gt 0 uD lt 0where subscripts denote partial derivatives so that utility increases withincreasing consumption of goods and transport services and decreaseswith increasing environmental damage and that u is strictly quasi-concavein the choice variables to ensure a unique equilibrium Further we assumethat DC gt 0 DE gt 0 DC gt DE so that environmental damage increasesmore per unit of conventional car transport compared to EV transportFor simplicity we assume that total transport is constant so thatdC D iexcldE Using this and totally di erentiating (1) we can express theoverall utility change from marginal changes of l x and e

dU D ul d l C ux dx C ue C nuDDE iexcl uc iexcl nuDDC( ) de (2)

On the production side we assume a linear technology so thatw(t iexcl l) D x C qcc C q ee where w is individual productivity and (exogen-ous) wage level t is total available time and qc and q e are the productionprices of conventional and EV transport respectively Without loss ofgenerality the production price of good x is normalised to one By totallydi erentiating this function and rearranging we obtain

dx D iexcl(q e iexcl qc) de iexcl w dl (3)

The individual maximises utility subject to a budget constraint withconsumer prices including taxes implying the optimal conditions

ul D wm ux D (1 C t x)m uc D (qc C t c)m ue D (q e C t e)m (4)

where t x t c and t e are the consumption taxes on the respective good andwhere m is the marginal utility of (net) income Inserting (3) and (4) into

Journal of Transport Economics and Policy Volume 37 Part 1

3 We do not deal with distributional aspects explicitly We believe such e ects are relatively

minor in this context but we believe the overall e ects of subidising private EVs are

regressive since new cars are primarily bought by relatively wealthy people

4

(2) we can write the social Marginal Net Beneregt (MNB) of replacingconventional cars with EVs as

MNB acutedU=de

mD (MDc iexclMDe) iexcl ((t c iexcl qct x) iexcl (t eiexcl q etx)) iexcl w

dl

de(1C tx)

(5)

where MDc iexcl MDe acute iexcl(n=m)uD DC iexcl DE( ) is the di erence in marginalexternal costs between conventional and electric vehicles and where wecan interpret t x to be a general consumption tax or a VAT Hence thesocial net beneregt per unit of additional e that is electric-vehicle-basedroad transport equals the di erence in marginal external cost betweenconventional cars and electric vehicles minus the tax-di erence net ofVAT on the producer price minus a term related to possible labour-supplychanges

The last term of (5) relates to excess-burden e ects that is in additionto direct resource costs there is an additional cost of raising public rev-enues since tax increases in an imperfect second-best world cause distor-tions in the economy From individual utility maximisation we have

px u

lD w

u

x (6)

Totally di erentiating this for regxed prices implies

px ull dl C ulx dx C ulc dc C ule de C ulDdD( )

D w uxl d l C uxx dx C uxc dc C uxe de C uxDdD( ) (7)

This can be re-written applying the same conditions as previously toobtain

d l

deD iexcl

wuxx iexcl pxulx( )(q e iexcl qc) C pxule iexcl wuxe( ) C wuxc iexcl pxulc( )C pxulD iexcl wuxD( )n(D=E) C wuxD iexcl pxulD( )n(D=C)

pxull C pxw iexcl wuxl iexcl w2 (8)

If utility is separable in the externality so that the utility functionmay be written U D u ( f (l x c e) D) we have pxulD iexcl wuxD DwuxD iexcl pxulD D 0 Furthermore if leisure is weakly separable from theconsumption goods as follows U D u(l g(x c e) D) then we havewuxx iexcl pxulx D pxule iexcl wuxe D wuxc iexcl pxulc D 0 In the special case whenutility can be written as U D u( f (l g(x c e)) D) both kinds of separ-ability are fulreglled simultaneously Substituting both associated conditionsinto (8) implies that d l=de D 0 so that we can disregard excess-burden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

5

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Introduction

Despite their environmental advantages Electric Vehicles (EVs) are stillquite rare and various ways of supporting them regnancially are oftenproposed in public debate One argument is that even though EVs may notbe socially beneregcial today they probably will be tomorrow and that theintroduction of EVs is a long process that must start today The mainpurpose of this paper is to regnd out whether it is likely that EVs will beprivately and socially proregtable in the foreseeable future and henceindirectly to provide relevant information for policy decisions todayTherefore we undertake a social cost-beneregt analysis (CBA) of di erentkinds of electric vehicles including hybrid vehicles and fuel-cell vehicles inSweden for the year 2010 Such a project involves large uncertainties andmany simplireged assumptions Nevertheless we believe that order-of-magnitude estimates along with a detailed presentation of the assump-tions are valuable from a policy-making perspective

Considering the role of EVs in public debate about transport and theenvironment it is somewhat surprising that until recently so few sys-tematic CBAs of EVs have been undertaken Kazimi (1997ab) using amicro-simulation model estimated the environmental beneregts of intro-ducing battery-driven electric cars (BCs) in the US The cost side was notanalysed but based on her results she judged that large price-reductions ofalternative-fuel vehicles would not be socially beneregcial Funk and Rabl(1999) found that the life cycle costs for BCs in Greater Paris (France)were higher than the corresponding costs for petrol cars but were lowerthan the costs for diesel cars Hahn (1995) discussed the cost e ectivenessof various measures to improve environmental quality in the transportsector His main conclusion was that tighter air pollution standards andimproved fuel qualities were more cost-e cient than for example anintroduction of BCs Similarly Wang (1997) in a survey of American cost-e ectiveness studies concluded that zero-emission vehicles were among theleast cost-e ective alternatives More recently Lipman et al (2000) andOgden et al (2001) have presented detailed social cost estimates of fuel-celland hybrid vehicles whereas Delucchi and Lipman (2001) based ondamage costs from Delucchi (2000) provides detailed and comprehensivelife-cycle cost estimates for BCs and concludes that battery performancemust improve and costs be reduced substantially if BCs are to becomecompetitive1

1 Delucchi (2001) is another source of detailed life-cycle cost estimates

Journal of Transport Economics and Policy Volume 37 Part 1

2

This study presents less detailed information on the manufacturing andmaintenance costs trying instead to capture reasonable order-of-magni-tude estimates of all relevant cost elements while simultaneously high-lighting the general equilibrium e ects of changed tax revenues This isparticularly important in countries with high fuel taxes as is the case inmost European countries A fairly common view seems to be thatalthough BCs are not privately proregtable they are or will be sociallyproregtable when the external costs are taken into account However as willbe demonstrated it may actually be the opposite that is that BCs maybecome privately proregtable (in certain market segments) but sociallyunproregtable when tax-revenue e ects are taken into account despite thefact that BCs are environmentally superior

The remainder of this paper is organised as follows In Section 2 atheoretical CBA model is outlined in a stylised general equilibrium modelThis model deals with how to handle welfare e ects of changed externalcosts tax revenues from indirect general taxation as well as targetedexternality-correcting taxes and excess burden e ects associated withnegative revenue e ects of public subsidies In particular it is shown thatunder certain frequently made separability assumptions of the utilityfunction one can disregard excess burden e ects of the lost tax revenuesassociated with public subsidies of for example electric vehicles Thesubsequent CBA will consequently disregard such e ects and is hencequite ``traditionalrsquorsquo in this sense Section 3 presents and discusses the largenumber of assumptions made Section 4 presents the results of the CBAwhile Section 5 provides concluding remarks

The Cost-Beneregt Model

We use a standard cost-beneregt model where we are primarily interested inconsumer welfare or utility Although the standard CB model has beenused in numerous applications in the transport sector some features of themodel remain unclear (at least in the Swedish policy debate) (i) Howshould the correction for changed tax revenues be made and in particularshould one include value added taxes (VAT) that are based on other taxes(such as energy taxes or environmental taxes)2 (ii) Should one correct for

2 Energy taxes and environmentally related fuel taxes are currently part of the VAT tax

base in Sweden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

3

excess-burden e ects and if so how To deal with these issues we apply asimple representative-individual3model with the following utility function

U D u(l x c e D) (1)

where l is leisure x is a composite good or put simply money spent oneverything except road transport c and e are the consumption of con-ventional and electric-vehicle based road transport respectively and D isenvironmental damage and where D D D(C E) C D

PniD1 ci D nc and

E DPn

iD1 ei D ne We assume that ul gt 0 ux gt 0 uc gt 0 ue gt 0 uD lt 0where subscripts denote partial derivatives so that utility increases withincreasing consumption of goods and transport services and decreaseswith increasing environmental damage and that u is strictly quasi-concavein the choice variables to ensure a unique equilibrium Further we assumethat DC gt 0 DE gt 0 DC gt DE so that environmental damage increasesmore per unit of conventional car transport compared to EV transportFor simplicity we assume that total transport is constant so thatdC D iexcldE Using this and totally di erentiating (1) we can express theoverall utility change from marginal changes of l x and e

dU D ul d l C ux dx C ue C nuDDE iexcl uc iexcl nuDDC( ) de (2)

On the production side we assume a linear technology so thatw(t iexcl l) D x C qcc C q ee where w is individual productivity and (exogen-ous) wage level t is total available time and qc and q e are the productionprices of conventional and EV transport respectively Without loss ofgenerality the production price of good x is normalised to one By totallydi erentiating this function and rearranging we obtain

dx D iexcl(q e iexcl qc) de iexcl w dl (3)

The individual maximises utility subject to a budget constraint withconsumer prices including taxes implying the optimal conditions

ul D wm ux D (1 C t x)m uc D (qc C t c)m ue D (q e C t e)m (4)

where t x t c and t e are the consumption taxes on the respective good andwhere m is the marginal utility of (net) income Inserting (3) and (4) into

Journal of Transport Economics and Policy Volume 37 Part 1

3 We do not deal with distributional aspects explicitly We believe such e ects are relatively

minor in this context but we believe the overall e ects of subidising private EVs are

regressive since new cars are primarily bought by relatively wealthy people

4

(2) we can write the social Marginal Net Beneregt (MNB) of replacingconventional cars with EVs as

MNB acutedU=de

mD (MDc iexclMDe) iexcl ((t c iexcl qct x) iexcl (t eiexcl q etx)) iexcl w

dl

de(1C tx)

(5)

where MDc iexcl MDe acute iexcl(n=m)uD DC iexcl DE( ) is the di erence in marginalexternal costs between conventional and electric vehicles and where wecan interpret t x to be a general consumption tax or a VAT Hence thesocial net beneregt per unit of additional e that is electric-vehicle-basedroad transport equals the di erence in marginal external cost betweenconventional cars and electric vehicles minus the tax-di erence net ofVAT on the producer price minus a term related to possible labour-supplychanges

The last term of (5) relates to excess-burden e ects that is in additionto direct resource costs there is an additional cost of raising public rev-enues since tax increases in an imperfect second-best world cause distor-tions in the economy From individual utility maximisation we have

px u

lD w

u

x (6)

Totally di erentiating this for regxed prices implies

px ull dl C ulx dx C ulc dc C ule de C ulDdD( )

D w uxl d l C uxx dx C uxc dc C uxe de C uxDdD( ) (7)

This can be re-written applying the same conditions as previously toobtain

d l

deD iexcl

wuxx iexcl pxulx( )(q e iexcl qc) C pxule iexcl wuxe( ) C wuxc iexcl pxulc( )C pxulD iexcl wuxD( )n(D=E) C wuxD iexcl pxulD( )n(D=C)

pxull C pxw iexcl wuxl iexcl w2 (8)

If utility is separable in the externality so that the utility functionmay be written U D u ( f (l x c e) D) we have pxulD iexcl wuxD DwuxD iexcl pxulD D 0 Furthermore if leisure is weakly separable from theconsumption goods as follows U D u(l g(x c e) D) then we havewuxx iexcl pxulx D pxule iexcl wuxe D wuxc iexcl pxulc D 0 In the special case whenutility can be written as U D u( f (l g(x c e)) D) both kinds of separ-ability are fulreglled simultaneously Substituting both associated conditionsinto (8) implies that d l=de D 0 so that we can disregard excess-burden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

5

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

This study presents less detailed information on the manufacturing andmaintenance costs trying instead to capture reasonable order-of-magni-tude estimates of all relevant cost elements while simultaneously high-lighting the general equilibrium e ects of changed tax revenues This isparticularly important in countries with high fuel taxes as is the case inmost European countries A fairly common view seems to be thatalthough BCs are not privately proregtable they are or will be sociallyproregtable when the external costs are taken into account However as willbe demonstrated it may actually be the opposite that is that BCs maybecome privately proregtable (in certain market segments) but sociallyunproregtable when tax-revenue e ects are taken into account despite thefact that BCs are environmentally superior

The remainder of this paper is organised as follows In Section 2 atheoretical CBA model is outlined in a stylised general equilibrium modelThis model deals with how to handle welfare e ects of changed externalcosts tax revenues from indirect general taxation as well as targetedexternality-correcting taxes and excess burden e ects associated withnegative revenue e ects of public subsidies In particular it is shown thatunder certain frequently made separability assumptions of the utilityfunction one can disregard excess burden e ects of the lost tax revenuesassociated with public subsidies of for example electric vehicles Thesubsequent CBA will consequently disregard such e ects and is hencequite ``traditionalrsquorsquo in this sense Section 3 presents and discusses the largenumber of assumptions made Section 4 presents the results of the CBAwhile Section 5 provides concluding remarks

The Cost-Beneregt Model

We use a standard cost-beneregt model where we are primarily interested inconsumer welfare or utility Although the standard CB model has beenused in numerous applications in the transport sector some features of themodel remain unclear (at least in the Swedish policy debate) (i) Howshould the correction for changed tax revenues be made and in particularshould one include value added taxes (VAT) that are based on other taxes(such as energy taxes or environmental taxes)2 (ii) Should one correct for

2 Energy taxes and environmentally related fuel taxes are currently part of the VAT tax

base in Sweden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

3

excess-burden e ects and if so how To deal with these issues we apply asimple representative-individual3model with the following utility function

U D u(l x c e D) (1)

where l is leisure x is a composite good or put simply money spent oneverything except road transport c and e are the consumption of con-ventional and electric-vehicle based road transport respectively and D isenvironmental damage and where D D D(C E) C D

PniD1 ci D nc and

E DPn

iD1 ei D ne We assume that ul gt 0 ux gt 0 uc gt 0 ue gt 0 uD lt 0where subscripts denote partial derivatives so that utility increases withincreasing consumption of goods and transport services and decreaseswith increasing environmental damage and that u is strictly quasi-concavein the choice variables to ensure a unique equilibrium Further we assumethat DC gt 0 DE gt 0 DC gt DE so that environmental damage increasesmore per unit of conventional car transport compared to EV transportFor simplicity we assume that total transport is constant so thatdC D iexcldE Using this and totally di erentiating (1) we can express theoverall utility change from marginal changes of l x and e

dU D ul d l C ux dx C ue C nuDDE iexcl uc iexcl nuDDC( ) de (2)

On the production side we assume a linear technology so thatw(t iexcl l) D x C qcc C q ee where w is individual productivity and (exogen-ous) wage level t is total available time and qc and q e are the productionprices of conventional and EV transport respectively Without loss ofgenerality the production price of good x is normalised to one By totallydi erentiating this function and rearranging we obtain

dx D iexcl(q e iexcl qc) de iexcl w dl (3)

The individual maximises utility subject to a budget constraint withconsumer prices including taxes implying the optimal conditions

ul D wm ux D (1 C t x)m uc D (qc C t c)m ue D (q e C t e)m (4)

where t x t c and t e are the consumption taxes on the respective good andwhere m is the marginal utility of (net) income Inserting (3) and (4) into

Journal of Transport Economics and Policy Volume 37 Part 1

3 We do not deal with distributional aspects explicitly We believe such e ects are relatively

minor in this context but we believe the overall e ects of subidising private EVs are

regressive since new cars are primarily bought by relatively wealthy people

4

(2) we can write the social Marginal Net Beneregt (MNB) of replacingconventional cars with EVs as

MNB acutedU=de

mD (MDc iexclMDe) iexcl ((t c iexcl qct x) iexcl (t eiexcl q etx)) iexcl w

dl

de(1C tx)

(5)

where MDc iexcl MDe acute iexcl(n=m)uD DC iexcl DE( ) is the di erence in marginalexternal costs between conventional and electric vehicles and where wecan interpret t x to be a general consumption tax or a VAT Hence thesocial net beneregt per unit of additional e that is electric-vehicle-basedroad transport equals the di erence in marginal external cost betweenconventional cars and electric vehicles minus the tax-di erence net ofVAT on the producer price minus a term related to possible labour-supplychanges

The last term of (5) relates to excess-burden e ects that is in additionto direct resource costs there is an additional cost of raising public rev-enues since tax increases in an imperfect second-best world cause distor-tions in the economy From individual utility maximisation we have

px u

lD w

u

x (6)

Totally di erentiating this for regxed prices implies

px ull dl C ulx dx C ulc dc C ule de C ulDdD( )

D w uxl d l C uxx dx C uxc dc C uxe de C uxDdD( ) (7)

This can be re-written applying the same conditions as previously toobtain

d l

deD iexcl

wuxx iexcl pxulx( )(q e iexcl qc) C pxule iexcl wuxe( ) C wuxc iexcl pxulc( )C pxulD iexcl wuxD( )n(D=E) C wuxD iexcl pxulD( )n(D=C)

pxull C pxw iexcl wuxl iexcl w2 (8)

If utility is separable in the externality so that the utility functionmay be written U D u ( f (l x c e) D) we have pxulD iexcl wuxD DwuxD iexcl pxulD D 0 Furthermore if leisure is weakly separable from theconsumption goods as follows U D u(l g(x c e) D) then we havewuxx iexcl pxulx D pxule iexcl wuxe D wuxc iexcl pxulc D 0 In the special case whenutility can be written as U D u( f (l g(x c e)) D) both kinds of separ-ability are fulreglled simultaneously Substituting both associated conditionsinto (8) implies that d l=de D 0 so that we can disregard excess-burden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

5

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

excess-burden e ects and if so how To deal with these issues we apply asimple representative-individual3model with the following utility function

U D u(l x c e D) (1)

where l is leisure x is a composite good or put simply money spent oneverything except road transport c and e are the consumption of con-ventional and electric-vehicle based road transport respectively and D isenvironmental damage and where D D D(C E) C D

PniD1 ci D nc and

E DPn

iD1 ei D ne We assume that ul gt 0 ux gt 0 uc gt 0 ue gt 0 uD lt 0where subscripts denote partial derivatives so that utility increases withincreasing consumption of goods and transport services and decreaseswith increasing environmental damage and that u is strictly quasi-concavein the choice variables to ensure a unique equilibrium Further we assumethat DC gt 0 DE gt 0 DC gt DE so that environmental damage increasesmore per unit of conventional car transport compared to EV transportFor simplicity we assume that total transport is constant so thatdC D iexcldE Using this and totally di erentiating (1) we can express theoverall utility change from marginal changes of l x and e

dU D ul d l C ux dx C ue C nuDDE iexcl uc iexcl nuDDC( ) de (2)

On the production side we assume a linear technology so thatw(t iexcl l) D x C qcc C q ee where w is individual productivity and (exogen-ous) wage level t is total available time and qc and q e are the productionprices of conventional and EV transport respectively Without loss ofgenerality the production price of good x is normalised to one By totallydi erentiating this function and rearranging we obtain

dx D iexcl(q e iexcl qc) de iexcl w dl (3)

The individual maximises utility subject to a budget constraint withconsumer prices including taxes implying the optimal conditions

ul D wm ux D (1 C t x)m uc D (qc C t c)m ue D (q e C t e)m (4)

where t x t c and t e are the consumption taxes on the respective good andwhere m is the marginal utility of (net) income Inserting (3) and (4) into

Journal of Transport Economics and Policy Volume 37 Part 1

3 We do not deal with distributional aspects explicitly We believe such e ects are relatively

minor in this context but we believe the overall e ects of subidising private EVs are

regressive since new cars are primarily bought by relatively wealthy people

4

(2) we can write the social Marginal Net Beneregt (MNB) of replacingconventional cars with EVs as

MNB acutedU=de

mD (MDc iexclMDe) iexcl ((t c iexcl qct x) iexcl (t eiexcl q etx)) iexcl w

dl

de(1C tx)

(5)

where MDc iexcl MDe acute iexcl(n=m)uD DC iexcl DE( ) is the di erence in marginalexternal costs between conventional and electric vehicles and where wecan interpret t x to be a general consumption tax or a VAT Hence thesocial net beneregt per unit of additional e that is electric-vehicle-basedroad transport equals the di erence in marginal external cost betweenconventional cars and electric vehicles minus the tax-di erence net ofVAT on the producer price minus a term related to possible labour-supplychanges

The last term of (5) relates to excess-burden e ects that is in additionto direct resource costs there is an additional cost of raising public rev-enues since tax increases in an imperfect second-best world cause distor-tions in the economy From individual utility maximisation we have

px u

lD w

u

x (6)

Totally di erentiating this for regxed prices implies

px ull dl C ulx dx C ulc dc C ule de C ulDdD( )

D w uxl d l C uxx dx C uxc dc C uxe de C uxDdD( ) (7)

This can be re-written applying the same conditions as previously toobtain

d l

deD iexcl

wuxx iexcl pxulx( )(q e iexcl qc) C pxule iexcl wuxe( ) C wuxc iexcl pxulc( )C pxulD iexcl wuxD( )n(D=E) C wuxD iexcl pxulD( )n(D=C)

pxull C pxw iexcl wuxl iexcl w2 (8)

If utility is separable in the externality so that the utility functionmay be written U D u ( f (l x c e) D) we have pxulD iexcl wuxD DwuxD iexcl pxulD D 0 Furthermore if leisure is weakly separable from theconsumption goods as follows U D u(l g(x c e) D) then we havewuxx iexcl pxulx D pxule iexcl wuxe D wuxc iexcl pxulc D 0 In the special case whenutility can be written as U D u( f (l g(x c e)) D) both kinds of separ-ability are fulreglled simultaneously Substituting both associated conditionsinto (8) implies that d l=de D 0 so that we can disregard excess-burden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

5

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

(2) we can write the social Marginal Net Beneregt (MNB) of replacingconventional cars with EVs as

MNB acutedU=de

mD (MDc iexclMDe) iexcl ((t c iexcl qct x) iexcl (t eiexcl q etx)) iexcl w

dl

de(1C tx)

(5)

where MDc iexcl MDe acute iexcl(n=m)uD DC iexcl DE( ) is the di erence in marginalexternal costs between conventional and electric vehicles and where wecan interpret t x to be a general consumption tax or a VAT Hence thesocial net beneregt per unit of additional e that is electric-vehicle-basedroad transport equals the di erence in marginal external cost betweenconventional cars and electric vehicles minus the tax-di erence net ofVAT on the producer price minus a term related to possible labour-supplychanges

The last term of (5) relates to excess-burden e ects that is in additionto direct resource costs there is an additional cost of raising public rev-enues since tax increases in an imperfect second-best world cause distor-tions in the economy From individual utility maximisation we have

px u

lD w

u

x (6)

Totally di erentiating this for regxed prices implies

px ull dl C ulx dx C ulc dc C ule de C ulDdD( )

D w uxl d l C uxx dx C uxc dc C uxe de C uxDdD( ) (7)

This can be re-written applying the same conditions as previously toobtain

d l

deD iexcl

wuxx iexcl pxulx( )(q e iexcl qc) C pxule iexcl wuxe( ) C wuxc iexcl pxulc( )C pxulD iexcl wuxD( )n(D=E) C wuxD iexcl pxulD( )n(D=C)

pxull C pxw iexcl wuxl iexcl w2 (8)

If utility is separable in the externality so that the utility functionmay be written U D u ( f (l x c e) D) we have pxulD iexcl wuxD DwuxD iexcl pxulD D 0 Furthermore if leisure is weakly separable from theconsumption goods as follows U D u(l g(x c e) D) then we havewuxx iexcl pxulx D pxule iexcl wuxe D wuxc iexcl pxulc D 0 In the special case whenutility can be written as U D u( f (l g(x c e)) D) both kinds of separ-ability are fulreglled simultaneously Substituting both associated conditionsinto (8) implies that d l=de D 0 so that we can disregard excess-burden

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

5

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

e ects altogether The separability assumptions imply that environmentaldamage does not a ect consumer choice and an exogenous change inleisure does not a ect the choice between transport goods and othergoods This is of course restrictive but it provides a natural benchmarkcase when dealing with public subsidies Similar assumptions are oftenmade in the literature (for example Bovenberg and De Mooij 1994) Itshould also be emphasised that the separability assumption does not implythat excess-burden considerations can be neglected generally Given theassumption it is straightforward to show that an optimal externality-cor-recting tax (net of indirect taxes) is smaller (larger) than a standardPigouvian tax if the uncompensated labour-supply elasticity with respectto a proportional income tax is smaller (larger) than zero see Bovenbergand De Moiij (1994) Similarly an optimal investment to decrease envir-onmental damage D directly is smaller (or larger) than the one corre-sponding to the Samuelson (1954) e ciency rule if the uncompensatedlabour-supply elasticity with respect to a proportional income tax issmaller (larger) than zero see Atkinson and Stern (1974)

To understand the intuition behind the result consider the (unrealistic)case where a tax increase is used to subsidize all goods by a certain per-centage Assuming no transaction costs this would clearly imply the samerelative costs for all goods and leisure as was the case before the changehence there should be no correction for excess burden If only one good issubsidised one would of course not expect to regnd the same results sincethat good may for example be more complementary with leisure thanother goods However it still appears reasonable to take the case of noexcess burden as a natural benchmark if one has no other speciregc infor-mation4 Consequently in this study we assume that the public subsidy ofelectric vehicles does not increase the excess burden Furthermore werecommend generally as a benchmark case that public subsidies shouldnot be corrected for excess burden This is contrary to the current practicein Sweden where a factor equal to 13 (denoted ``tax-factor 2rsquorsquo in Swedishtransport planning) is used to remacrect the marginal cost of raising publicfunds as well as the marginal cost of public subsidies

The social marginal net beneregt of an increase in the number of electricvehicles can then be written as

MNB D MDc iexcl MDe( ) iexcl (t c iexcl qctx) iexcl (t e iexcl q et x)( ) acute centMD iexcl centR (9)

4 Furthermore in a setting with many non-identical consumers and an optimal non-linearincome tax the works by Christiansen (1981) Boadway and Keen (1993) and Kaplow

(1996) imply that correcting for marginal excess burden in public investments can be

questioned more generally

Journal of Transport Economics and Policy Volume 37 Part 1

6

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

where centMD is the decrease in marginal damage or external costs fromreplacing one conventional vehicle with an EV and centR is the associateddecrease in tax revenues In other words the social net beneregt of replacingone conventional vehicle with an EV is the di erence between the beneregtin terms of decreased external costs and the cost in terms of reducedtaxation The reason di erences in private (for example vehicle) costs donot enter this expression is that these costs are already taken into accountby a rational utility-maximising consumer

In the special case where we are in a regrst-best world with optimalPigouvian taxes equal to the marginal damage for both types of vehicleswe regnd that the social value of such a shift is zero It follows that changesin tax revenues should include VAT based on speciregc taxes but not the``generalrsquorsquo VAT based on the production price (or the value added) Onceagain this is contrary to the current practice in Sweden As we will seechanges in tax revenues will constitute a substantial element in the regnalCBA

Assumptions

General assumptions and vehicles analysedThe time perspective of this study is the year 2010 for which we analysethe introduction of various EVs We assume that petrol and diesel pricesas well as fuel and vehicle taxes are similar to those of today Theseassumptions provide a natural benchmark case since it is not clear whe-ther it is more or less likely that prices and taxes will increase rather thandecrease

This paper will contrary to a number of recent studies in this sub-regeldrely on conventional CBA methodology and not use life-cycle analysisthus we do not for example take environmental e ects in vehicle pro-duction into account This is for three reasons First our analysis focuseson the di erences between various types of vehicle and since they are quitesimilar in most respects it is reasonable to expect that the associatedenvironmental di erences are often fairly limited However this need notalways be the case for example Lave et al (1995) argue that BCs withlead batteries cause large emissions of lead Second it is still not clear thatlife-cycle costs should be taken into account theoretically If theseexternalities are already internalised through the authorities in the pro-ducing country then taking the same external costs into account again willlead to ine ciencies this parallels the conclusions from the ``small-num-

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

7

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

berrsquorsquo externality case within a country (see Coase 1960)5 Admittedlyhowever close to optimal externality-correcting taxes are in most coun-tries clearly the exception rather than the rule But even if welfare-improving internalisation is not implemented the responsibility for thiscan hardly be given to another country Third there are huge practicaldi culties when obtaining vehicle-type speciregc emission-estimates sincethese vary largely between di erent countries and regions Furthermore itappears likely that emission di erences arising from di erent legal rulesbetween countries are quite large compared to the ones resulting frommanufacturing di erent types of vehicles within a country

For battery cars (BCs) we focus exclusively on small-sized cars such asthe Renault ClioVW Polo since the speciregc features of BCs make largercars less realistic and attractive (Sperling and Lipman 2000) In order tolimit the infrastructure investments needed for these cars and since thehealth costs of local emissions are much larger in cities we assume that allthese BCs are sold in the three major cities in Sweden StockholmGoEgrave teborg and MalmoEgrave

There are many types of hybrid vehicles but they can be divided intotwo broad categories (i) vehicles using petrol as their ``primaryrsquorsquo source ofenergy and which are hence independent of central electricity productionand (ii) vehicles operating as battery vehicles most of the time andtherefore using centrally produced electricity as their primary source ofenergy but also having a combustion engine as an auxiliary engine inorder to increase performance and driving distance capacity In this paperwe focus on the former alternative for hybrid passenger cars (HCs) since itappears more likely that they will become socially beneregcial Cars such asthe Toyota Prius and the Honda Insight that are already in the market areindications of this but there is other evidence as well see for exampleHarrop and Harrop (1999) and Sperling and Lipman (2000) We discusstwo types of HCs one ``mildrsquorsquo and one ``advancedrsquorsquo hybrid where the mildhybrid is a less developed type that is usually driven by a petrol engineand the advanced hybrid (such as the Toyota Prius or the Honda Insight)that has a somewhat larger battery and can be driven as a pure BC forshorter distances at a low speed for example in tra c queues We assumethat half of the HCs are sold within the three major cities in Sweden andthe other half are sold throughout the rest of the country In addition weanalyse an introduction of methanol-driven fuel-cell cars (FCCs)

5 In this case the agents would negotiate directly with each other to obtain an e cient

allocation An additional tax from the society would hence be distortionaryand move the

outcome away from what is e cient

Journal of Transport Economics and Policy Volume 37 Part 1

8

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

For heavy vehicles we focus on city-based delivery trucks which areused in the cities but not in the countryside We analyse two types ofhybrid diesel trucks (HTs) (i) trucks using diesel as the ``primaryrsquorsquo sourceof energy and (ii) trucks that can be charged with electricity from thenational grid and can be used as a pure battery truck for shorter distancesWe assume that both FCCs and HTs are only sold in the three main citiesin Sweden

For each EV introduced the sale of conventional vehicles is reduced byone unit implying that the overall size of the car stock does not changewith the number of EVs introduced Further we assume the same averagedriving distance for the EVs as the conventional cars they replace Boththese assumptions simplify the analysis drastically since we do not have toestimate all external cost elements associated with road tra c only thosethat di er between vehicle types For example we do not have to estimatethe external costs of congestion road wear and tear6 barrier e ects oraccidents nor the consumer surplus change associated with a changeddriving distance

Nevertheless it is possible that overall car sales would increase Forexample some previous one-car households might diversify their carholding due to the di erent characteristics of EVs and BCs in particularthat make them more attractive for some kinds of trip Nevertheless sincethe costs of such diversiregcation are substantial and the characteristics ofthe EVs are not that di erent from conventional vehicles of a comparablesize we suspect that the overall e ect on total sale would be limited

It is also possible that the average driving distance would be lower thanthat of the replaced conventional cars If so this would tend to bias theenvironmental beneregts upwards It is not possible to say in which directionthis assumption would bias the overall CBA however since this dependson factors such as whether the taxes on car tra c today are higher orlower than the socially optimal level To determine this we need to valueall external costs as well as estimate the spatial distribution of the changein tra c volume

6 However as pointed out by a referee road wear and tear is a function of the axle weight

which may be higher for EVs Still the associated costs are small for passenger cars and

for trucks the increased weight may be partly o set by changed design and constructionthrough more axles or a more even weight distribution between the axles (since the

relationship between axle weight and damage is highly non-linear it is mainly the highest

axle weight of the vehicle that matters)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

9

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Production and maintenance costsIn a background report to Carlsson and Johansson-Stenman (2000a)Duleep (2000) estimated the costs of the di erent EVs for 2010 For BCscharacteristics and incremental price are presented in Table 1 where thebaseline is a small petrol car such as a Renault Clio The comparison isdone for two di erent battery capacities both of the Nickel-MetalHydride (NiMH) type Most analysts seem to believe in this type of bat-tery considering both performance and manufacturing costs although theuncertainties are large and possible alternatives include lithium polymerand lithium-ion batteries see Lipman (1999) and Sperling and Lipman(2000) We assume that the batteries will last for the lifetime of the carsince the battery technology is expected to improve greatly by 2010(Duleep 2001 personal communication) Still that may be somewhatoptimistic in particular for BCs and the cost of replacing a battery ifneeded is very large (Delucchi and Lipman 2001) We see that both thecarrsquos weight and its price increase drastically with battery capacity andhence with driving range and performance We therefore concentrate onthe low-range vehicle in our economic analysis For BCs and FCCsDuleep (2000) assumes that worldwide sales for each model are relativelysmall and production is set at 20000 units per year per model

For HCs and FCCs the baseline for the comparisons presented in Table2 is a Volvo V70 Controlling for di erences in fuel economy and emissioncharacteristics (and possibly reliability and service costs) the market valuecan be assumed to be about the same since the cars are assumed to have

Table 1Characteristics and Incremental Price for a Small BC

Relative to a Comparable Size Petrol Car by year 2010

Low range High range

Weight 830 1720Drag coe cient 028 028

Energy e ciency (kWhkm) 011 023

Motor power (kW) 415 81Battery wt (kg) 225 775

Battery size (kWh) 169 58

Range to 20 per cent dod 120 200Units produced per year 20000 20000

Expected life length (years) 17 17

Annual driving distance (km) 15000 15000Incremental price (over small car) USD 6406 22686

Own assumptions

(Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

10

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

largely similar characteristics Duleep (2000) assumes that HCs will be soldworldwide in a signiregcant volume by 2010 and that each hybrid makemodel is produced at about 100000 units per year Hydrogen-based FCCsare shown only for comparison we do no believe that they will be arealistic option by 2010

Repair and maintenance costs for EVs are of course di cult to esti-mate but careful and detailed estimates are provided by Delucchi andLipman (2001) They also report cost estimates to enable charging thevehicles at home From their results one can conclude that the sum of thepresent value of the estimated repair and maintenance costs insurancecosts and the costs associated with home charging7 are quite similar for

Table 2Characteristics and Incremental Price (2000 USD) for HCs and FCCs

Relative to a Comparable Volvo V70 Petrol Car

2010 20102000 2010 2010 mild 2010 adv hydrogen methanol

base conventional hybrid hybrid fuel-cell fuel-cell

Weight 1430 1290 1320 1340 1376 1340

Engine type 25L I-5 25L I-4 18L I-4 16L I-4 Fuel cell Fuel cell

Combustion Homog GDI Homog Atkinson Hydrogen MeOH

reformer

Motor power (kw) 121 112 90 70 70 75

Elec Motor None None 12 kW 30 kW peak 70 kW 75 kW

City FC

(L100 km) 1060 80 65 534 289 393

Highway FC 644 53 52 480 270 359

Composite FC 872 68 595 510 280 378

Battery plusmn plusmn 1 kW-HR 25 kW-HR plusmn plusmn

12 kW 30 kW

Units produced

per year 100000 100000 20000 20000

Expected life

length (years) 17 17 17 17 17 17

Annual driving

distance (km) 15000 15000 15000 15000 15000 15000

Additional costs

USD Base case 805 2197 4225 4845 6568

Own assumptions

Ignoring energy losses when producing the hydrogen see table 5

Ignoring energy losses when producing the methanol see table 5

Source Duleep (2000)

7 How large these costs are naturally depends on the speciregc equipment that is considered

necessary BraEgrave nnstroEgrave m (2000) for example presents somewhat higher cost estimates

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

11

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

BCs as for conventional vehicles Although these estimates are for anAmerican context it is hard to see why it would be very di erent inwestern European countries It is also di cult to see why there would beany big di erences for other kinds of EV see Lave and MacLean (2002)For these reasons and in order to keep the analysis as simple as possiblewe will refrain from explicitly including these cost elements in the CBA

Table 3 presents the characteristics and prices of hybrid diesel trucks(HTs) and compares them to conventional diesel trucks in 2010 A mildHT refers to a type of truck that is never charged with electricity from thenational grid whereas an advanced HT could be charged from thenational grid and be used as a pure battery truck for shorter distances

Central infrastructureThe need for investments in infrastructure for BCs has been analysed in abackground report to Carlsson and Johansson-Stenman (2000a) byBraEgrave nnstroEgrave m (2000) According to this report the introduction of BCs will

Table 3Characteristics and Incremental Price (2000 USD) for Conventional and

Hybrid Trucks

Mild hybrid truck Advanced hybrid truck

Conventional parallel hybrid series hybriddiesel not grid chargeable grid chargeable

Gross weight (tons) 12 12 12

Payload (tons) 65 65 62

Engine power 165 kW 2200 125 kW 2800 125 kW 2800Engine type 6 L I-6 diesel 4 L I-4 diesel 4 L I-4 diesel

Elec Motor None 40 kW peak 125 kW continuous

Generator None None 90 kWBattery None 6 kWh 40 kW 12 kWh 80 kW

Pure BC range None None 12 to 15 km city

Fuel cons (L100km) 28 185 202Units produced per year 2000 2000

Expected life length

(years) 17 17 17Annual driving

distance (km) 30000 30000 30000

Additional costs USD Base case 7575 28270

Own assumptions

The batteries and other components are assumed to be produced at a large scale

Source Duleep (2000)

Journal of Transport Economics and Policy Volume 37 Part 1

12

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

most likely not require investments in the general electricity net (althoughinvestments are needed to enable charging of the vehicles as mentioned)In addition to conventional charging at home centrally distributed rapidcharging stations would largely increase the macrexibility and service levelsince the risks associated with a sudden electricity shortage woulddecrease Despite the name rapid charging is still rather slow compared tofuelling a conventional petrol car however and may require almost halfan hour (assuming that there is no need to wait in a queue) BraEgrave nnstroEgrave m(2000) assumed one station per 40 cars distributed in the centre of the cityand found the annual cost (including city land rental) to be as high as 855USD per vehicle per year8

OEgrave stman (2000) in another background report analysed the additionalinfrastructure costs for methanol to be used by the FCCs relative to thecorresponding costs for petrol The additional distribution costs are lar-gely due to the lower scale compared to petrol but also due to the lowerenergy density of methanol (almost a factor 2 compared to petrol) Heconcluded that an additional cost of about 012 USDlitre of petrol-equivalent methanol (or about 006 USDlitre of methanol) is a reasonableestimate for the additional costs

Emission factorsThe environmental costs associated with di erent types of vehicle con-stitute an important part of our analysis There are two important com-ponents (i) the emissions associated with di erent vehicles and (ii) thevaluation of these emissions For petrol and diesel vehicles we use theestimates of emission factors in Ahlvik et al (1996) These are estimatedaverage emission factors throughout the lifetime of a car of a certain agebased on many sources including decided and planned future emissionstandards within the EU

In comparison with conventional vehicles there is little systematicempirical evidence on present or future emission factors for di erent kindsof hybrid vehicle Furthermore what evidence there is is often based oninformation from car manufacturers for speciregc car models (see Lave andMacLean 2002) Even so most seem to agree that hybrid vehicles do havea potential for non-negligible emission reductions Based on order-of-magnitude discussions with various experts we make the admittedly crudeassumptions that the emission factors (except for CO2) of mild andadvanced HCs to be 50 and 25 per cent respectively compared to theemission factors for petrol cars Similarly we assume the emission factors

8 The exchange rate of 1 USD D 10 SEK is used throughout the paper

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

13

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

from mild and advanced HTs to be 50 and 25 per cent9 respectivelycompared to the emission factors for diesel trucks CO2 emissions from allvehicles are directly proportional to the fossil content of the fuel used10

and the emission factors used in the calculations are presented in Table 4

Table 4Estimated Emission Factors for Vehicles of

Di erent Vintages

Year VOC gkm NOx gkm Pm mgkm

Passenger cars petrol1996 089 026 13

2010 008 004 12

Passenger cars petrol (city)

1996 187 034 7

2010 017 005 12

Passenger cars mild petrol hybrid

1996 045 013 652010 004 002 06

Passengers cars advanced petrol hybrid1996 0225 007 325

2010 002 001 03

Diesel trucks

1996 072 97 200

2010 030 49 100

Mild hybrid trucks

1996 036 485 1002010 015 245 50

Advanced hybrid trucks1996 015 204 21

2010 006 103 105

Sources Ahlvik et al (1996) and own calculations

assumptions in text

9 The lower emission factors for advanced HCs and HTs should be seen as ``e ectiversquorsquo

reductions given that vehicles can be used as pure battery vehicles for shorter distances

where the emissions are particularly damaging The advanced HT can in addition begrid-charged for which we assume no local emissions

10 The fossil fuel content per energy unit di ers however among the fuels as follows

Petrol 260 gkWh Diesel (or oil) 271 gkWh Coal 326 gkWh and Natural gas 202 g

kWh

Journal of Transport Economics and Policy Volume 37 Part 1

14

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Emissions from electricity productionIn terms of emissions what consequences will arise from an additionalkWh of electricity Within the kind of international deregulated electricitymarket that has developed in the northern part of Europe marginalelectricity production comes from fossil sources whereas production fromhydropower and nuclear power plants is regxed11 Furthermore environ-mental e ects are independent of where the CO2 emissions occur If thegovernmental objective is to reduce the impact of CO2 emissions globallyrather than to reduce CO2 emissions from Sweden it follows that theexternal costs of CO2 emissions should be based on electricity productionfrom fossil fuels12 We assume perhaps somewhat optimistically thatmarginal electricity production by 2010 will be based on natural gasrather than coal or oil

However one could also argue that the main objective of the Swedishpolicy on greenhouse gases is not to reduce CO2 emissions worldwide butrather to reduce such emissions in Sweden since Sweden along with othercountries has signed international reduction agreements that it would liketo comply with for the lowest possible cost If this is the social objectiveno CO2 costs from imported electricity production should be includedSince the ``truersquorsquo underlying objective is di cult to obtain we present theresults for both cases Perhaps the most realistic assumption however isthat the Swedish government does care about global emissions but careseven more about emissions in Sweden implying an intermediate casebetween the two cases presented We ignore emissions from importedelectricity for the same reasons that we do not perform life-cycle analysisof vehicle production

Energy e ciency prices and taxesTable 5 reports our assumptions regarding energy use for di erent types ofvehicle again largely in accordance with Duleep (2000) In order tofacilitate comparisons we report fuel consumption in di erent units Theelectricity consumption per km for a BC is highly uncertain but we rely onthe assumptions made in Duleep (2000) For HCs petrol is always used astheir primary source of energy Since the petrol engine can be utilised moree ciently in a hybrid vehicle since it can be optimised for a smaller rangein terms of RPM energy consumption is lower than in conventional petrol

11 They are regxed annually hence they are independent of whether electricity demand

increases or decreases by 5 per cent for example However they will vary greatly duringthe year mainly due to the seasonal characteristics of hydropower

12 To base the costs on some kind of average electricity production as is often proposed

would clearly be incorrect

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

15

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

cars For conventional diesel and hybrid trucks we use the estimates inDuleep (2000) and we assume that the advanced HT uses diesel as itsprimary source of energy for 85 per cent of the driving distance andelectricity for the remaining 15 per cent

We assume that regxed real fuel prices per litre are equal to 09 USD(including VAT) for petrol 096 USD (including VAT) for methanol and05 USD (excluding VAT) for diesel13 The fuel-tax levels including VAT

Table 5Estimated Energy use per Distance Unit Expressed in Di erent UnitsComposite Fuel Driving Cycle Assumed Except for Small Vehicles and

Trucks where a City-based Driving Cycle is Assumed

Petrol Dieseleq l10km eq l10km kWhkm

Petrol passenger car 067 061 059

Battery car (low range) electricity from grid 013 012 011Battery car (low range) in terms of fossil

natural gas energy 025 023 022

Battery car (high range) electricity from grid 027 024 023Battery car (high range) in terms of fossil

natural gas energy 052 047 046Mild hybrid petrol car 059 053 052

Advanced hybrid petrol car 051 046 044

Fuel cell car Methanol 038 034 033Fuel cell car

Methanol including methanol production 054 049 047

Fuel cell car Hydro energy 028 025 024Fuel cell car Hydro energy including

hydrogen production 033 029 028

Diesel truck 310 280 270Mild hybrid truck 205 185 179

Advanced hybrid truck pure diesel mode 224 202 195

Advanced hybrid truck15 electricity (05 kWhkm) 199 179 173

Advanced hybrid truck 15 electricity in

terms of fossil natural gas 207 187 180

The same values are used for smaller cars which are compared with battery cars as for bigger

cars that are compared with hybrid and fuel cell cars The reason is that the smaller cars are only

driven in bigger cities implying higher fuel consumption per se

Based on 50 energy e ciency from natural gas power production and 7 distribution

lossesBased on 70 energy e ciency from natural gas (Katofsky 1993)Based on 85 energy e ciency from natural gas (Johansson and AEcirc man 1999)

13 This is lower than the regular price at fuel stations since all users of heavy vehicles have

large rebates

Journal of Transport Economics and Policy Volume 37 Part 1

16

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

for the year 2000 are 055375 USDlitre of petrol and 03325 USD perlitre of diesel The Electricity tax including VAT is 0019 USDkWh

Environmental shadow pricesWe use the o cial values for transport infrastructure investments inSweden as recommended by SIKA (2000) to evaluate local and regionalenvironmental e ects This is presented in Table 614

The values recommended by SIKA are like virtually all environmentalshadow values highly uncertain However many international estimatesof local environmental costs are of the same order of magnitude seeCarlsson and Johansson-Stenman (2000b) Greene et al (1997) Maddisonet al (1996) Mayeres et al (1996) and Small and Kazimi (1993) A recentcomprehensive study on health e ects by McCubbin and Delucchi (1999)estimates local environmental costs of di erent types of emission TheSIKA values for VOC and particles are larger than these values but theUS values for NOX are generally higher

The damage cost of CO2 emissions is even more uncertain and scien-tiregcally reported cost estimates include values from about 10 USDton ofC (Nordhaus 1994) up to 600 USDton of C (Azar and Sterner 1996)where the latter study used the same underlying scientiregc assumption butdi erent assumptions on distributional consequences and (in particular)discount rates In addition there is great uncertainty about the underlyingmeteorological mechanisms and the associated impacts on human wel-fare both now and in the future At the same time Sweden along withother countries is spending its limited resources on combating CO2

emissions in order to fulregl international agreements and possibly for otherreasons too A good proxy for the associated shadow price is the house-hold tax level which is about 200 USDton of C (053 SEKkg CO2)

Table 6Local Environmental Costs Per Kilogram Emissions in Sweden for

Di erent Regions

Regional costs (USDkg) Local health costs (USDkg)

ETHETHETHETHETHETHETHETHETHETHETH ETHETHETHETHETHETHETHETHETHETHETHETHETHETHETH

VOC NOx VOC NOx Particles

Average Stockholm 3 6 41 49 690Average rest of Sweden 3 6 15 49 260

Source SIKA (2000)

14 There is no reported shadow value for SOx presumably due to the fact that almost all

diesel sold in Sweden is of a high quality with a very low sulfur content (lt 10 ppm)

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

17

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Given that the tax level is set optimally based on the governmentalobjectives (whatever they are) then this number remacrects the social valueper unit of CO2 reduction15 Therefore we use the tax level as an estimateof the external costs of CO2 emissions in this study It is also often used inother applications as a shadow price of CO2 emissions in Sweden despitethe fact that it is not directly based on any CO2 damage-cost estimates Itshould be noted though that the value used is higher than most damage-cost estimates and it is also higher than the 100 USDton of C estimateused by Ogden et al (2001) which was based on the cost of achieving``deep reductionsrsquorsquo in CO2 emissions at coal power plants

We follow the standard assumption of the literature regarding lineardose-response functions This assumption may imply an overestimation ofmarginal environmental costs at concentration levels lower than todayrsquoslevels if the ``truersquorsquo marginal environmental costs are increasing slightlywith the concentration levels or if threshold levels exist On the otherhand it is possible that other associated environmental and health riskswill be identireged in the future which will correspondingly imply higherfuture environmental shadow values

There are few available estimates of the marginal external cost of noisefor di erent vehicles but we know that most EVs are less noisy than petroland diesel vehicles In order to have a simple measure of the marginalexternal cost of noise per passenger per kilometre we take the results forSweden given by SIKA (2000) as a point of departure They present crudemarginal cost estimates of 007 USD10 km for conventional cars and 047USD10 km for heavy trucks in cities These estimates are of the sameorder of magnitude as the ``high-cost casersquorsquo presented in Delucchi and Hsu(1998) but are substantially higher than their ``base casersquorsquo and ``low-costcasersquorsquo For obvious reasons even less is known about the external cost ofnoise from non-conventional vehicles However based on the abovenumbers we make the assumptions presented in Table 7 In addition wetake into account the fact that tyre-noise dominates engine-noise at higherspeeds implying that the percentage reductions are larger in city tra cFurther the values remacrect that we assume that BCs are driven only in thelarger cities whereas we assume that HTs are driven in di erent kinds ofcity but not in the countryside

Occasionally it has been argued that there is a negative side e ect ofquieter vehicles in terms of decreased safety since silent vehicles are more

15 However as noted by a referee the political cultures vary largely and in some countriesit appears unrealistic to assume that the taxes are set optimally whatever the underlying

social objectives are Hence the assumption made may not be realistic for another

country

Journal of Transport Economics and Policy Volume 37 Part 1

18

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

di cult to detect On the other hand one can argue that noise makes itmore di cult to concentrate and to communicate with other people suchas children and hence that safety could improve by lowering the generalnoise levels In the lack of clear evidence regarding this point we assumethat the net e ect is zero that is we do not include any possible indirecte ects on safety

Cost-Beneregt Analysis

External costs per distance unitGiven our assumptions we calculate the environmental cost per distanceunit for each vehicle type and the results are reported in Table 7 Wepresent the results only for the biggest city (Stockholm) and for theaverage of the rest of Sweden although di erent values for a number ofcities are used in the regnal CBA In the case where marginal electricityproduction is assumed not to produce any external costs the environ-mental costs for EVs are zero BCs generally have the lowest environ-mental costs as expected We see that the external environmental costs forpetrol passenger cars are quite small for those made in 2010 which is aremacrection of the expected rapid reduction of emissions see Table 4 Eventhough BCs HCs and FCCs produce even lower local and regionalemissions the cost di erence in absolute terms is relatively small

Diesel truck emissions do not decrease as drastically by 2010 as petrolpassenger car emissions The local environmental costs particularly inlarger cities are substantial mainly due to particulate emissions but alsodue to noise pollution

CBA resultsOne of the most di cult components in producing a CBA on EVs is theroad-usersrsquo valuation of the EV per se For example apart from thee ects on their personal budget do most car buyers prefer hybridvehicles or petrol vehicles Theoretically an individual will choose an EVrather than a petrol car when WTP C centPV gt 0 where WTP is theadditional willingness to pay for having an EV instead of a conventionalcar given an equal regnancial budget between the two centPV is the presentvalue of the additional cost of having an EV that is the present value of

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

19

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Table 7Estimated External Environmental Costs for Di erent EVs in USD10 km

Local Env Regional CO2 NoiseCosts Env Costs Costs Costs Total

Gasoline passenger carsStockholm 0018 0008 0082 007 0178

Avg Sweden small cities 0008 0008 0082 002 0119

Battery cars CO2 from electricity production excluded

Stockholm 0 0 0 003 0030

Battery cars CO2 from electricity production included

Stockholm 0 0 0024 003 0054

Fuel cell cars

Stockholm 0 0 0051 004 0091

Mild hybrid cars

Stockholm 0007 0002 0073 006 0142

Avg Sweden small cities 0003 0002 0073 0018 0096

Advanced hybrid cars

Stockholm 0003 0001 0062 005 0117Avg Sweden small cities 0002 0001 0062 0016 0067

Diesel trucksStockholm 0528 0137 0394 05 1558

Avg Sweden small cities 0267 0137 0394 03 1098

Mild hybrid trucks

Stockholm 0264 0068 0260 030 0892

Avg Sweden small cities 0134 0068 0260 018 0642

Advanced hybrid trucks CO2 from elect production excluded

Stockholm 0111 0030 0242 025 0633Avg Sweden small cities 0056 0030 0242 015 0478

Advanced hybrid trucks CO2 from elect production included

Stockholm 0111 0030 0257 025 0649

Avg Sweden small cities 0056 0030 0257 015 0494

Journal of Transport Economics and Policy Volume 37 Part 1

20

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

the EV price di erential (including private infrastructure costs) minusfuel cost savings To avoid having to calculate WTP we will calculatebackwards the critical level of WTP at which the EV becomes privatelyproregtable From a social perspective the most important issue is whetheror not EVs should be subsidised and if so by how much Below wecalculate the socially optimal EV subsidy for each vehicle type By doingthis we do not intend to argue in favour of subsidies as a suitable policyinstrument Rather we believe that there are other more suitable policyinstruments such as di erentiated road pricing (Johansson-Stenman1999) especially in a long-term perspective Even so the values give anindication of the degree to which it is socially proregtable to subsidise suchvehicles

Table 8 shows that BCs are actually regnancially proregtable in the sensethat the expected lifetime cost of a BC is lower than that of a corre-sponding petrol car This does not imply that many passenger car buyerswill regnd BCs attractive however since many crucial characteristics di ersuch as acceleration and engine power driving range and battery chargingtime Indeed the recent literature on the choice among di erent types ofvehicle indicates that BCs must be substantially cheaper than petrol cars inorder to obtain a non-negligible market share (Brownstone et al 2000Brownstone and Train 1999 and Ewing and SarigoEgrave lluEgrave 1998)16 Howeverthese are American studies and it is di cult to use them in a Swedishsetting since for example the Swedish and the American fuel prices di erdramatically Nevertheless it appears reasonable that most buyers of asmall new car in Sweden would also prefer a conventional car despite theadditional regnancial life-cycle costs Still the regnancial analysis indicatesthat speciregc niches might exist where BCs will be considered attractive Itshould be noted that the potential regnancial proregtability of BCs hingeslargely upon the relatively high fuel price in Sweden compared to forexample the US Hence the results here are not incompatible withDelucchi and Lipman (2001) which found that BCs are unlikely to beregnancially proregtable in the US

In either case private proregtability may di er substantially from socialproregtability Since the external costs associated with BCs are lower thanthose of petrol vehicles one would perhaps expect that their social prof-itability would be better than their private proregtability However it turnsout that this is not the case This is because governmental tax revenuesover the life cycle of a vehicle decrease greatly when replacing a conven-tional car with a BC This remacrects the fact that compared with conven-

16 These studies are based on stated preference data except Brownstone et al (2000) which

uses both stated and revealed preference data

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

21

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Table 8Present Value of Costs and Beneregts in 1000 USD Per Car

Real Discount Rate 4 Per Year

Battery car Battery car Mild Advanced Mild Advanced

CO2 from electricity CO2 from electricity hybrid hybrid Fuel-cell hybrid hybrid

production excluded production included car car car truck truck

Price calculationIncremental price iexcl640 iexcl640 iexcl220 iexcl423 iexcl657 iexcl758 iexcl2827

Cost saving fuel 1024 1024 128 268 352 1734 1974

Required WTP forprivate proregtability iexcl384 iexcl384 092 154 305 iexcl976 850

Social calculationEnviron beneregt 295 273 053 102 156 2362 3212

Tax revenues iexcl641 iexcl641 iexcl079 iexcl165 iexcl298 iexcl1153 iexcl1262

Motivated electric-vehiclesubsidy iexcl346 iexcl368 iexcl026 iexcl063 iexcl142 1209 1905

Journ

alof

Tran

sport

Eco

nom

icsan

dP

olicyV

olu

me

37

Part

1

22

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

tional vehicles electric vehicles are already heavily ``subsidisedrsquorsquo throughthe relatively low electricity tax (as compared with fuel taxes)17 This is asocial cost element that is often overlooked in the current discussions onpolicy In addition the di erence in external costs found here is typicallymuch smaller than the di erence in taxes Indeed according to the resultsevery switch from a conventional car to a BC will imply a net present valuecost to the rest of society equal to almost 4000 USD This is despite thefact that we have used the questionable assumption that electricity pro-duction causes no external costs The possibility of publicly provided rapidcharging would further amplify the social costs of BCs and would imply asocial cost per car of about 10000 USD18 Consequently we can concludethat it appears very unlikely that an introduction of BCs through publicsubsidies would be socially proregtable

Mild and advanced HCs are more promising from a social point ofview although the calculations show that there is no motivation forsubsidising here either Whether or not these vehicles are privatelyproregtable depends on the willingness to pay for driving alternative fuelvehicles which would be at least 900 and 1500 USD for the mild andadvanced HCs respectively Since these vehicles have characteristicssimilar to those of conventional petrol cars the additional WTP for anHC remacrects a WTP for a more environmentally friendly car It is pos-sible that a non-negligible fraction of new car buyers including com-panies would be willing to pay an environmental premium of this size inorder to signal environmental awareness or social responsibility forexample

Neither FCC seems to be privately proregtable In addition sincemethanol cannot be expected to be supplied outside large cities at leastnot in a consumer friendly way the private WTP in favour of a conven-tional vehicle may be large implying that a sizeable subsidy may be neededto sell FCCs in signiregcant quantities Furthermore there is no reason tosubsidise FCCs either However in a longer time perspective hydro-powered FCCs with greater energy e ciency will be possible The vehiclecost will then also be lower since no transformer will be needed

Despite worse environmental performance with respect to local andregional pollutants the less advanced HT is socially more proregtable than

17 However we do not make any assumption or any calculations regarding whether theabsolute tax level of BCs is too low from a social point of view or not Nor do we say

anything about the appropriate tax level of petrol cars and their use But the result

indicates that the tax di erence between the use of petrol cars and BCs is socially toolarge and that is what drives the CB results

18 Presumably the WTP for a BC would increase due to such possibilities but hardly in this

order of magnitude

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

23

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

the advanced HT19 Interestingly this type of truck is more proregtable thanconventional diesel trucks The present value of the social net cost ofreplacing one diesel truck is 21850 USD It should be noted that animportant part of social proregtability is the reduction in external noise costTo a large extent this is due to our assumption that HTs are mainly drivenin city centres The less advanced type of truck is also proregtable for theconsumer while the more advanced HT is only privately proregtable if theassociated WTP is su ciently high

Discussion and Conclusions

One purpose of this paper was to shed light on the question of whethergovernments should regnancially promote the introduction of EVs by forexample subsidising them The conclusion is negative for most types ofpassenger car at least on a large scale One reason is the often substantialloss in tax revenue that the government would face if a consumer switchesto an EV

The perspective is di erent for city-based hybrid delivery trucks whichappear socially proregtable by 2010 even though this conclusion is sensitiveto the assumptions made Since these trucks use diesel as their primaryfuel there are no large social ``transition costsrsquorsquo of adapting HTs andthere is no particular infrastructure needed

Further these HTs are also likely to be privately proregtable due toimproved fuel economy implying that no (or only small) social subsidieswould be needed We focused our analysis of heavy vehicles on city-baseddelivery trucks rather than on city buses since the latter may needadditional climate control equipment20Nevertheless much of the analysisis valid for city buses as well

The results rest generally on a large number of crucial assumptionssome of which are rather uncertain For example future emission levels ofpassenger cars as well as of trucks are of course uncertain as is techno-logical development in general If there was to be an unforeseen breakthrough

19 We only report the results for marginal electricity production from fossil fuels since thedi erence between the two cases is very small

20 This was the suggestion by Duleep (2000 personal communication) for why hybrid

delivery trucks may be more promising than city buses However Sperling and Lipman(2000) argue for a high potential for city buses Even so they do not seem to compare

these with city-based delivery trucks but rather with trucks in general See Schimek

(2001) for social cost estimates of hybrid buses

Journal of Transport Economics and Policy Volume 37 Part 1

24

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

in battery technology which would largely improve performance then wecould certainly not rule out BCs Similarly if the development of dieseltrucks was to prove better than we assume with respect to noise fueleconomy and particle emissions then the proregtability of HTs wouldcorrespondingly decrease or possibly disappear The same applies to theassumed valuations of noise and CO2 which are in comparison with otherstudies on the high side On the other hand the epidemiological knowl-edge of the health e ects is still quite limited and it is possible that newrelations implying higher damages than assumed today will be obtained inthe future Equally the reverse outcome from increased knowledge cannotbe excluded either

The proregtability of FCCs is largely reduced by the fact that a dis-tribution system for hydrogen was not considered realistic in a 10-yearperspective The use of methanol leads both to more costly cars since areformer is needed and to lower energy e ciency Even so it is possible(and perhaps likely) that a widespread hydrogen distribution system willnever be designed However there are other alternatives besides methanoland there are even discussions about using petrol in the fuel cells but areformer is still needed to make hydrogen The major advantage of usingpetrol as a substitute for methanol is the existing widespread distributionsystem In addition the petrol itself may in the future be cheaper than thequality of petrol that we currently use because no octane-increasingadditions would be needed since there would be no combustion Onedisadvantage is that the achieved energy e ciency will not be as good asthe technology itself permits but it could still be better than conventionalpetrol or diesel engines It could also be a virtually zero emission vehiclefrom a local and regional point of view Furthermore the reformer neednot be in the vehicles it would be possible to have large-scale reformers atfuel stations so that the cars will still be driven on hydrogen Never-theless there remain non-negligible technological di culties and unsolvedquestions

Even though the uncertainties involved are large a number of con-clusions and insights appear fairly solid First compared to conventionalpetrol passenger cars BCs simply do not seem to be socially proregtableunless an unanticipated major breakthrough in battery technology takesplace Second there are a number of other EVs that may be sociallyproregtable including various kinds of HCs HTs and FCCs Howeversince the private proregtability in most cases seems to be of a similar order ofmagnitude as the social proregtability there is no clear case for publicsubsidies here either

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

25

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

References

Ahlvik P A Laveskog R Westerholm and K-E EgebaEgrave ck (1996) Emissionsfaktorer foEgraver

fordon drivna med biodrivmedel Motortestcenter (MTC) vid AB Svensk Bilprovning

Stockholms Universitet samt Tekniska HoEgrave gskolan i LuleaEcirc Atkinson A B and N H Stern (1974) ``Pigou Taxation and Public Goodsrsquorsquo Review of

Economic Studies 41 19plusmn28

Azar C and T Sterner (1996) ``Discounting and Distributional Considerations in theContext of Global Warmingrsquorsquo Ecological Economics 19 169plusmn85

Boadway R and M Keen (1993) ``Public Goods Self-selection and Optimal Income

Taxationrsquorsquo International Economic Review 34 463plusmn78Bovenberg A L and R A de Mooij (1994) ``Environmental Levies and Distortionary

Taxationrsquorsquo American Economic Review 94 1085plusmn89Brownstone D D Bunch and K Train (2000) ``Joint Mixed Logit Models of Stated and

Revealed Preferences for Alternative-fuel Vehiclesrsquorsquo Transportation Research B 34

315plusmn38Brownstone D and K Train (1999) ``Forecasting New Product Penetration with Flexible

Substitution Patternsrsquorsquo Journal of Econometrics 89 109plusmn29

BraEgrave nnstroEgrave m P (2000) Background Report to KFB on Electric Power Infrastructure SyconEnergy Consult

BaEgrave ckstroEgrave m S (1998) Environmental Assessment of Energy Supply Chains for Electric

Propelled Road Tra c Report to KFBCarlsson F and O Johansson-Stenman (2000a) ``The Costs and Beneregts of Electric

VehiclesETHShould Battery Hybrid and Fuel-cell Vehicles be Publicly Supported in

Swedenrsquorsquo KFB-rapport 200046Carlsson F and O Johansson-Stenman (2000b) ``Willingness to Pay for Improved Air

Quality in Swedenrsquorsquo Applied Economics 32 661plusmn70

Christiansen V (1981) ``Evaluation of Public Projects Under Optimal Taxationrsquorsquo Reviewof Economic Studies 48 447plusmn57

Coase R (1960) ``The Problem of Social Costrsquorsquo Journal of Law and Economics 111 1plusmn44

Delucchi M (1993) ``Greenhouse-gas Emissions from the use of New Fuels forTransportation and Electricityrsquorsquo Transportation Research A 27 187plusmn91

Delucchi M (2000) ``Environmental Externalities of Motor-Vehicle use in the USrsquorsquo

Journal of Transport Economics and Policy 34 135plusmn68Delucchi M (2001) Electric and Gasoline Vehicle Lifecycle Cost and Energy-Use Model

UCD-ITS-RR-99-4Report to the California Air Research Board Revised regnal report

Institute of Transportation Studies University of California DaviesDelucchi M and S-L Hsu (1998) ``The External Damage Cost of Noise from Motor

Vehiclesrsquorsquo Journal of Transportation and Statistics 1 1plusmn24

Delucchi M and T E Lipman (2001) ``An Analysis of the Retail and Lifecycle ofBattery-powered Electric Vehiclesrsquorsquo Transportation Research Part D 6 371plusmn404

Duleep (2000) Background Memo to KFB on Electric Vehicle Cost Estimates

Ewing G and E SarigoEgrave lluEgrave (1998) ``Car Fuel-type Choice under Travel DemandManagement and Economic Incentivesrsquorsquo Transportation Research D 3 429plusmn44

Funk K and A Rabl (1999) ``Electric versus Conventional Vehicles Social Costs and

Beneregts in Francersquorsquo Transportation Research D 4 397plusmn411Greene D L D W Jones and M A Delucchi (1997) The Full Costs and Beneregts of

Transportation Springer Verlag Heidelberg

Journal of Transport Economics and Policy Volume 37 Part 1

26

Hahn RW (1995) ``Choosing among Fuels and Technologies for Cleaning Up the AirrsquorsquoJournal of Policy Analysis and Management 14 532plusmn54

Harrop P and G Harrop (1999) Electric Vehicles are Proregtable Where Why WhatNext Footnote publications Hampshire UK

Johansson B and M AEcirc hman (1999) Koldioxidneutrala transporsystem En studie av

energie ektiva fordon och foEgravernybar energi Institutionen foEgrave r teknik och samhaEgrave lle Lundtekniska hoEgrave gskola

Johansson-StenmanO (1999) ``Regulating Road Transport Externalities Pricing Versus

Command-and-Controlrsquorsquo in Sterner T (ed) The Market and the Environment EdwardElgar

Katofsky R E (1993) The Production of Fluid Fuels from Biomass PUCESS Report No

279 The centre for energy and environmental studies Princeton UniversityKaplow A (1996) ``The Optimal Supply of Public Goods and the Distortionary Cost of

Taxationrsquorsquo National Tax Journal 49 513plusmn33

Kazimi C (1997a) ``Valuing Alternative-Fuel Vehicles in Southern Californiarsquorsquo AmericanEconomic Review 87 265plusmn71

Kazimi C (1997b) ``Evaluating the Environmental Impact of Alternative-Fuel Vehiclesrsquorsquo

Journal of Environmental Economics and Management 33 163plusmn85Lave L B C T Hendrickson and F C McMichael (1995) ``Environmental Implications

of Electric Carsrsquorsquo Science 268 993plusmn995

Lave L B and H L MacLean (2002) ``An Environmental-Economic Evaluation ofHybrid Electric Vehicles Toyotarsquos Prius vs its Conventional Internal Combustion

Engine Corollarsquorsquo Transportation Research Part D 7 155plusmn62

Lipman T (1999) The Cost of Manufacturing Electric Vehicle Batteries Report forCalifornia Air Research Board UCD-ITS-RR-99-5 Institute of Transportation

Studies University of California Davis

Lipman T M Delucchi and D J Friedman (2000) A Vision of Zero-emission VehiclesScenario Cost Analysis from 2003 to 2030 Final Report prepared for The Union of

Concerned Scientists Energy and Resources Group University of California

BerkeleyMaddison D D Pearce O Johansson E Calthrop T Litman and E Verhoef (1996)

The True Cost of Road Transport Blueprint 5 Earthscan CSERGE London

Mayeres I S Ochelen and S Proost (1996) ``The Marginal External Costs of UrbanTransportrsquorsquo Transportation Research Part D 1 111plusmn30

McCubbin D and M Delucchi (1999) ``The Health Costs of Motor-vehicle Related Air

Pollutionrsquorsquo Journal of Transport Economics and Policy 33 253plusmn86Nordhaus W D (1994) Managing the Global Commons The MIT Press Cambridge and

London

Ogden J M R H Williams and E D Larson (2001) Toward a Hydrogen-BasedTransportation System Final draft 8 May 2001 Center for Energy and Environmental

Studies Princeton University

OEgrave stman A (2000) Infrastructure for Distribution of Methanol for Fuel Cell Vehicles CostEvaluation May 2000 Report to KFB AtraxKemiinformation AB

Samuelson P A (1954) ``The Pure Theory of Public Expenditurersquorsquo Review of Economics

and Statistics 36 387plusmn89Schimek P (2001) ``Reducing Emissions from Transit Busesrsquorsquo Regional Science and

Urban Economics 31 433plusmn51

SIKA (2000) OEgrave versyn av foEgrave rutsaEgrave ttningarna foEgrave r marginalkostnadsbaserade avgifter itransportsystemet SIKA Rapport 200010

Costs and Beneregts of Electric Vehicles Carlsson and Johansson-Stenman

27

Small K A and Kazimi C (1995) ``On the Costs of Air Pollution from Motor VehiclesrsquorsquoJournal of Transport Economics and Policy 29 7plusmn32

Sperling D (1994) ``Prospects for Neighbourhood Electric Vehiclesrsquorsquo TransportationResearch Record 1444 16plusmn22

Sperling D and T Lipman (2000) ``International Assessment of Electric-Drive Vehicles

Policies Markets and Technologiesrsquorsquo Institute of Transportation Studies Universityof California Davis

Wang M Q (1997) ``Mobile Source Emission Control Cost-e ectiveness Issues

Uncertainties and Resultsrsquorsquo Transportation Research D 2 43plusmn56

Journal of Transport Economics and Policy Volume 37 Part 1

28