Consumer Preferences for Alternative Fuel Vehicles:
A Discrete Choice Analysis
André Hackbarth
*
and Reinhard Madlener
Institute for Future Energy Consumer Needs and Behavior (FCN), School of Business and
Economics / E.ON Energy Research Center, RWTH Aachen University, Mathieustrasse 10,
52074 Aachen, Germany
December 2011, revised December 2012
Abstract
In this paper we analyze the potential demand for privately used alternative fuel vehicles (AFVs), based on a
nationwide survey in Germany among (potential) car buyers. For this purpose, we applied a stated preference
discrete choice experiment, using a wide range of vehicle alternatives (gasoline/diesel, natural gas, hybrid, plug-
in hybrid, electric, hydrogen) and vehicle attributes. By applying both a multinomial logit model and a mixed
(error components) logit model, we estimate the attributes’ influence on vehicle choice and calculate consumers’
willingness-to-pay for the improvement of these attributes. Furthermore, in a scenario analysis, we simulate the
impact of monetary and non-monetary policy measures on vehicle choice probabilities. We find that the most
promising target group for the adoption of all kinds of AFVs is that of younger, well-educated, and
environmentally aware car buyers, who, in the case of electric vehicles, also have the possibility to plug-in their
car at home, and who have a high share of city trips and thus need a small car. Moreover, we find that,
depending on the vehicle alternative, environmental awareness, and budget constraints for the next vehicle
purchase, households are willing to pay substantial amounts for the improvement of fuel cost, driving range,
charging infrastructure, CO
2
emissions, vehicle tax exemptions, and free parking or bus lane access.
Furthermore, the scenario results suggest that conventional vehicles will maintain their dominance in the market,
whereas electric and hydrogen vehicles will remain unpopular. The market share of the latter is only expected to
rise markedly if massive and multiple policy interventions are implemented. Finally, we find evidence that an
increase in the fully electric vehicle’s driving range to a level comparable with all other vehicle alternatives has
the same impact on its choice probability as would a market-based, multiple measures policy intervention
package.
Keywords: Discrete choice model, Alternative fuel vehicles, Willingness-to-pay, Stated preferences, Mixed
logit, Error components model
JEL Classification Nos.: C25, D12, M38, Q58, R41
*
Corresponding author. Tel: +49 241 80 49834, Fax: +49 241 80 49829, E-mail: ahackbarth@eonerc.rwth-aachen.de.
2
1 Introduction
Due to its almost exclusive dependence on fossil fuels, the transportation sector accounts for
32% of the final energy demand in the European Union, and is responsible for about one fifth
of the total European Union greenhouse gas (GHG) emissions (EC, 2011a). Consequently, it
is one of the focal points of the European sustainability strategies, which aim at the mitigation
of substantial amounts of GHG emissions in several sectors of the economy. For instance, the
European Commission has set the ambitious goal of a 60% reduction of GHG emissions in the
transportation sector by 2050 compared to 1990 levels (EC, 2011b). However, the
achievement of this objective requires considerable efforts, amongst others the enhancement
of the vehicles’ fuel efficiency and the substitution of alternative fuels or electricity for
gasoline and diesel. Moreover, today’s transport system, which is largely based on individual
means of transportation, such as passenger cars, has to be changed fundamentally towards a
broad utilization of public modes of transport or shared vehicles.
Acknowledging the fact that mobility patterns are difficult to change rapidly, the European
Commission adopted several regulations to improve vehicles’ specific GHG emissions in the
short term. For instance, emission performance standards for new passenger cars were set to
95 gCO
2
/km on average by 2020 (EC, 2009b), with gradually stiffened interim targets.
Additionally, the European Commission determined that the share of renewable energy
should at least amount to 10% of the final energy consumption in transport by 2020 (EC,
2009a). Beyond that, most European governments have decided to implement further-
reaching programs and regulations to accelerate the diffusion of alternative fuel vehicles
(AFVs) in general and electric cars in particular
1
. For example, purchase and tax incentives
for (partially) electric or other ‘environmentally-friendly’ vehicles are granted in Spain,
France, the UK, Ireland, Sweden, Belgium, and the Netherlands, to name but a few. These
inducements to buy amount up to €9,510, as in Belgium, or even up to 70% of the investment,
as in Andalucia (for a useful review of electric vehicle promotion strategies, see e.g. ACEA,
2012).
1
AFVs comprise vehicles that run on liquid or gaseous fuels other than gasoline and diesel, or at least partly on
electricity, e.g. biofuels, natural gas (liquefied petroleum gas (LPG) or compressed natural gas (CNG)),
hydrogen (e.g. fuel cell vehicles), (plug-in) hybrid electric, and fully electric vehicles.
3
In Germany, the largest European economy, with its pronounced automotive
manufacturing sector, the government has set the goal to get one million electric vehicles on
the road by 2020 and to become a leading market for and provider of electric mobility
(Bundesregierung, 2009). To reach these targets, research on various technical, economic, and
behavioral aspects is coordinated centrally and will be funded by more than €1.5 billion in
total up to 2013 (Bundesregierung, 2011). In addition, a ten-year motor vehicle tax exemption
for electric vehicles was introduced in a first step. Further monetary incentives, such as
advantageous taxation rules for commercially used electric vehicles, have been initiated, and
several non-monetary buying inducements are under consideration, such as the permission for
bus lane usage or special parking areas (BMF, 2011; Bundesregierung, 2011). However, in
contrast to the European countries mentioned before, and to the loudly voiced annoyance of
German vehicle manufacturers, purchase premiums for vehicles with electrified drivetrains
are not granted.
Even though electric mobility is currently the primary topic of interest of German policy-
makers, other alternative fuels are being supported as well. For example, tax reductions for
natural gas fuels have been recently prolonged until 2018. Moreover, a minimum quota of
6.25% for biofuels to replace gasoline and diesel has been introduced (BImSchG, 2011), and a
public-private partnership, which runs until 2016 and provides €1.4 billion of funding, has
been established to boost research on hydrogen and fuel cells (BMVBS/BMWi/BMBF, 2006).
Despite these diversified endeavors of the government and administration, the assortment
of AFVs is still limited. Thus, it is hardly surprising that AFVs have not penetrated the market
yet to a large extent and amount to only about 1% of the overall vehicle stock in Germany
(KBA, 2011). However, the diffusion of AFVs might rise sharply in the next couple of years
for at least two reasons. First, all major vehicle manufacturers have started to bring mass-
produced, and thus affordable, plug-in hybrid or pure electric vehicles to market and have
announced that they will do so in the next couple of years with hydrogen-fueled vehicles.
Second, consumer prices for gasoline and diesel in 2012 (Q1-Q3) were at an all-time high and
expected to increase further (ADAC, 2012), which makes non-conventional fuels even more
attractive. For a fast market penetration of AFVs, however, it is necessary that also the
remaining features of AFVs match consumer preferences sufficiently.
The purpose of this paper is to assess the relative impact of the most important vehicle
attributes, such as purchase price, fuel cost, driving range, fuel availability, CO
2
emissions,
4
refueling time, and governmental incentives, on the potential demand for AFVs. Additionally,
we tackle the question of how much vehicle buyers are willing to pay for an improvement of
principal vehicle characteristics, such as a reduction of the purchase price, an extension of the
driving range or the acceleration of the battery recharging process for electric vehicles. On
this basis, we simulate how such beneficial changes affect the potential market shares of the
different propulsion technologies in a scenario-based analysis, whereby we predominantly
focus on the effects that assorted governmental incentive schemes wield on vehicle choice.
Moreover, we examine the acceptance of alternative fuels compared to gasoline and diesel for
distinct consumer groups, distinguished by socio-demographic characteristics. Taken together,
this information could be particularly helpful for policy-makers and industrial decision-
makers aiming to increase the adoption rate of AFVs in the future by focusing on the
improvement or subsidization of the most influential vehicle features and by specifically
adjusting their incentive schemes, marketing campaigns, and products to the preference
differences between consumer segments.
Our analysis is based on a thorough, Germany-wide, web-based stated preferences discrete
choice experiment, carried out among 711 potential car buyers in July and August of 2011.
Our study builds on the rich body of literature on the demand for AFVs, which has been
primarily carried out in the US (Beggs et al., 1981; Calfee, 1985; Bunch et al., 1993; Golob et
al., 1993; Brownstone and Train, 1999; Brownstone et al., 2000; Axsen et al., 2009; Hidrue et
al., 2011; Musti and Kockelman, 2011) and Canada (Ewing and Sarigöllü, 2000; Horne et al.,
2005; Potoglou and Kanaroglou, 2007; Mau et al., 2008; Axsen et al., 2009), but also in
Europe (Dagsvik et al., 2002; Batley et al., 2004; Caulfield et al., 2010; Mabit and Fosgerau,
2011; Lebeau et al., 2012; Achtnicht, 2012; Achtnicht et al., 2012; Ziegler, 2012), South
Korea (Ahn et al., 2008), and Japan (Ito et al., 2013). The works of Achtnicht (2012), Ziegler
(2012), and Achtnicht et al. (2012), which are all based on the same data set, have to be
pointed out, as they are, to the best of our knowledge, the only ones considering the German
market, and, hence, more closely related to our research. Achtnicht (2012) analyzed the
relevance of CO
2
emissions in vehicle choice decisions and, by applying a mixed (random
parameters) logit model, found that, subject to gender, age, and education, potential car
buyers are willing to pay substantial amounts for the abatement of vehicle emissions.
Achtnicht et al. (2012) examined the influence of fuel availability on vehicle choice. Based on
a standard logit model they found that the density of the respective refueling infrastructure
5
positively influences the demand for AFVs and, hence, is a prerequisite for significant vehicle
adoption. They also show a significant impact of several socio-demographic characteristics,
such as age and environmental awareness, on the potential market shares of AFVs. Their
results further reveal that German car buyers are ceteris paribus more reluctant towards some
types of propulsion technology, namely biofuel and electricity, than towards conventional
ones. In his comprehensive study, Ziegler (2012) deepens the understanding of the influence
that individual characteristics exercise on vehicle choice. Implementing flexible multinomial
probit models, numerous socio-demographic variables are found to positively affect the
demand for otherwise disfavored AFVs, such as biofuel, hydrogen, and electric cars. For
example, the results suggest that younger respondents prefer natural gas, biofuel, hydrogen,
and electric vehicles, that males choose natural gas and hydrogen vehicles more often, and
that environmental awareness increases the stated preference for biofuel, hydrogen, and
electrically-driven vehicles.
Similarly to this literature, our analysis is based on a broad variety of drivetrain
technologies and vehicle characteristics, i.e. we also consider conventional, natural gas,
hybrid, biofuel, electric, and hydrogen vehicles vis-a-vis purchase price, fuel cost, CO
2
emissions, and service station availability. However, we essentially expand the limitations of
these studies in two ways. First, we introduce plug-in hybrid electric vehicles (PHEVs) and
their particularities – two different refueling options with varying refueling times – as choice
alternatives in a discrete choice experiment.
2
Second, we characterize some of the vehicle
alternatives with additional attributes, i.e. the driving range on a full tank and/or battery, the
refueling and/or recharging time, and potential governmental actions to incentivize the
respective vehicle choice, an approach which has not been taken for Germany before. From
our point of view, the inclusion of the driving range and the recharging time is essential in
order to more realistically analyze consumer preferences regarding electric mobility. We are
therefore able to contribute to the current research and debate about the best strategy for a
fostering of electric vehicles by estimating willingness-to-pay (WTP) measures for driving
2
To be precise, PHEVs have already been introduced in a choice experiment by Musti and Kockelman (2011).
However, in their survey, the different vehicle alternatives were only described by purchase price and fuel cost at
fixed values and, thus, not varied by design.
