Energy Research & Social Science 55 (2019) 46–61
Contents lists available at ScienceDirect
Energy Research & Social Science
journal homepage: www.elsevier.com/locate/erss
Original research article
Can we hope for a collective shift in electric vehicle adoption? Testing
salience and norm-based interventions in South Tyrol, Italy
T
Nives DellaValle , Alyona Zubaryeva
⁎
Eurac Research – Institute for Renewable Energy, Via A. Volta 13/A, 39100 Bolzano, Italy
A R TICL E INFO
A BSTR A CT
Keywords:
Electric vehicles
Mobility
Behavioral economics
Survey experiment
Electric-drive vehicles (EVs) are a promising technology for containing environmental problems of the
transport sector. However, to be effective on the environment, these need to be purchased by consumers in
large quantities. Not only material barriers, but also cognitive biases prevent their diffusion. Particularly, EVs
fail to capture a significant passenger car market share, even in regions that can overcome most material
barriers. We report the findings from an online survey experiment administered to respondents of a region
identified as a potential EV lead market, but which still fails to reach a significant EV market share. We test
the effect of two behaviorally informed strategies on EV preferences: a norm-based and a salience intervention. To better identify treatment effect, we control for pro-environment self-identity, heterogeneity in
key economic preferences, and size. Results show that making future cost savings salient significantly increases the likelihood of choosing EVs. However, the effect is limited to individuals displaying preferences for
big vehicles and for valuing future benefits, and high values in the pro-environment self-identity measure. In
addition, results show that EV choices are unaffected by the descriptive norm embedded in the norm-based
intervention.
1. Introduction
In recent decades, road transport has largely relied on oil-based
fuels, reaching the level of 96% of energy consumed by vehicles in 2015
[1]. This and other factors, such as a growing trend of green-house gas
emissions and urban air pollution, have led to steady and continuous
efforts to introduce alternative fuelled vehicles, in particular electricdrive vehicles (EV – comprising plug-in hybrid electric vehicles (PHEV),
battery electric vehicles (BEV) and fuel cell vehicles (FCV)) in the
European market. These vehicles are seen as one of the most relevant
innovations that can make the transport sector more sustainable and
address environmental problems caused by conventionally fuelled vehicles [2,3]. However, to be fully beneficial to the environment, these
alternative vehicles need to be purchased by consumers in large
quantities.
In Europe, the transport sector is still the second largest source of
greenhouse gas (GHG) emissions, accounting for about a quarter of total
emissions [4]. The European Union set the target of reducing CO2
emissions in all sectors by 80% by 2050. Also, the 2011 EC White Paper
on Transport set a target of reducing road transport emissions by 60%
of 1990 levels by 2050, and within this to “halve the use of ‘conventionally fueled’ cars in urban transport by 2030 and phase them out
in cities by 2050” [5]. Both short- and long-term estimates of EV market
penetration vary greatly [6]. However, despite the rapidly growing
sales,1 EVs have failed to capture a significant passenger car market
share and continue to be dependent on support measures, such as financial incentives [7].
There are various factors that may promote the market uptake of
EVs in Europe. Among these, a high GDP/capita, fuel cost savings from
gasoline and petrol, a high share of renewable energy share (RES) in the
energy mix and contribution to energy security are some crucial features of potential lead markets for EVs [8].
Italy is one of the main countries displaying such a potential (Fig. 1)
[9], and South Tyrol (Fig. 2) is a particularly interesting case study.
Located in the Italian Alps, this region exhibits most of the favorable
conditions for the diffusion of EV and hydrogen fuel cell vehicles [8].
Corresponding author.
E-mail addresses: nives.dellavalle@eurac.edu (N. DellaValle), alyona.zybaryeva@eurac.edu (A. Zubaryeva).
1
Source: European Automobiles Manufacturares Association Statistics.
⁎
https://doi.org/10.1016/j.erss.2019.05.005
Received 9 November 2018; Received in revised form 28 April 2019; Accepted 2 May 2019
2214-6296/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Fig. 1. Lead market forecast for BEV potential of NUTS2.
Fig. 2. Study area showing the province examined in this study.
Moreover, it has a population with high environmental2 and autonomy
concerns.3 Firstly, these high concerns for the environment [10] and
autonomy [11] may substantially contribute to promotion and adoption
of renewable energy systems.
Secondly, the high share of electricity produced in the region from
hydro-power (91.6%) might unlock environmental benefits of EVs.4
Moreover, many economic barriers might be overcome thanks to the
mix of high GDP/capita5 and the regional government's commitment to
reduce CO2 emissions largely from the transport sector down to 1.5 tons
per year per capita by 2050, this being the major contributor to CO2
emissions in the area [12]. In particular, to reach the goal of 60% of
vehicle-km travelled represented by zero-emission vehicles by 2050
[13], the regional government has incentivized EV purchases through
direct incentives; moreover, it has committed to improving the charging
infrastructure,6 and to offering test drives and tax reductions to the
citizens and companies.7
2
In 2016 the population voted against the opening of an airport via referendum – Source: Comune di Bolzano, Alto Adige.
3
In South Tyrol, there are three different language subgroups (Italian,
German and Ladin speakers) with special cultural differences. Therefore, the
Italian government guarantees South Tyrolean legislative and administrative
autonomy greater than the competencies of a region governed under normal
statute – Source: Provincia autonoma di Bolzano, Alto Adige.
4
Source: Istituto provinciale di statistica - Provincia autonoma di Bolzano,
Alto Adige.
5
Source: Istituto provinciale di statistica - Provincia autonoma di Bolzano,
Alto Adige.
6
Source: h2-suedtirol, Alperia Group.
7
Source: Green Mobility Alto Adige.
47
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Fig. 3. New BEV/PHEV registered in South Tyrol. Source: Ministero delle Infrastrutture e dei Trasporti.
However, despite this fertile ground, since 2010 only 0.1% of total
new vehicle registrations are represented by EVs in South Tyrol
(Fig. 3).8 Why is EV uptake still low in South Tyrol?
The reason might lie in the fact that material barriers are only some
of many factors preventing EV adoption [2]. As an example, individuals
might fail to adopt EVs because they tend to overvalue the high purchase price and undervalue the low lifecycle operating costs [14]. Such
a pattern of behavior can be well described by the field of behavioral
economics. This field has been influential in disrupting policy-making
that has traditionally been based on a rational choice model of human
behavior. By providing evidence that individuals exhibit systematic and
predictable patterns of decision-making that depart from this model's
assumptions (i.e. cognitive biases), behavioral economics has provided
policy-makers not only with a richer model of human behavior, but also
with a richer policy toolbox, which usually comprises taxes, incentives
and regulations. One concrete illustration is choice architecture: by intervening in the decision problem, governments can redirect behavior
without forbidding any option or changing economic incentives [15].
As an example, governments might implement strategies that simplify
the decision problem and rely on the power of social influence or
cognitive biases [16]. These strategies might help governments design
interventions that are more effective at tackling several policy challenges [4].
In this study, we investigate how the challenge of low EV uptake can
benefit from insights from behavioral economics. In particular, we test
the effect of two behaviorally informed strategies to promote preferences for EVs: one that relies on social influence (i.e., norm-based
intervention [17]), and another that simplifies the problem of deciding
to purchase a vehicle (i.e., salience intervention [18]). Firstly, informing individuals that similar peers have already chosen to purchase
EVs might be effective at shifting preferences for EVs. Secondly, making
EV future cost savings salient might prove to be an effective strategy for
helping individuals to appreciate future benefits of an EV. At the same
time, heterogeneity in behaviour might also affect several individual
choices [19]. As an example, those who display higher capacity to value
8
future consumption, i.e., higher willingness to delay benefits, are more
likely to invest in energy-efficient appliances [20]. Therefore, to isolate
the effect of our treatments on preferences for EVs, we control for major
sources of heterogeneity that might correlate with preferences for EVs.
The paper is divided into five parts. Section 2 explores the conceptual background and introduces the research questions. Section 3
presents the method and the sample. In Section 4, we perform the statistical analysis, of which the results are discussed in Section 5. Finally,
Section 6 concludes.
2. Conceptual background and research questions
To encourage diffusion of EVs, current governments, including
South Tyrol, have promoted several incentivizing policies by providing
financial subsidies, driving privileges, or tax rebates [21]. These policies have been proven to positively influence consumers’ EV adoption
intentions [22], however the magnitude of their efficacy is often lower
than expected [23]. One reason might be that these policies do not take
into account consumer heterogeneity [24]. As these policies may turn
out to be costly and inefficient, governments might look for alternative
cost-effective strategies aimed at promoting large-scale EV adoption. As
an example, they might implement strategies that act on the problem of
deciding to redirect behavior without forbidding any option or changing economic incentives [15]. In particular, they might implement
strategies that simplify the decision-making problem and rely on the
power of social influence [16].
2.1. Norm-based intervention
EVs need to be adopted by a large mass of individuals to release
their full potential on the environment. This implies that a collective
behavioral change is required. A norm-based intervention, i.e., an intervention relying on social influence, can be a candidate tool to shift
preferences in favor of EVs. However, compared to other interventions,
it is crucial to take a more cautious approach before implementing a
norm-based one, given that individuals might draw different inferences
on the type of information provided. Therefore, we first need to diagnose the nature of behavior to change [25].
Source: Ministero delle Infrastrutture e dei Trasporti.
48
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
information provision making salient future cost savings, can be a
candidate tool to circumvent discounting of future EV cost savings and
promote preferences for EVs.
Several studies have experimentally validated the positive effect of
making future cost savings salient on uptake of energy efficient technologies, such as efficient TVs [41], energy efficiency measures [42],
and energy efficiency appliances [43]. On the other hand, for what
concerns eco-friendly vehicles, it is still unclear whether making future
cost savings salient positively affects individuals’ choices in favor of
eco-friendly vehicles. While [44] find that providing information in the
framework of the European Energy Label rating scheme increases
purchase decisions for energy-efficient vehicles, [45] note that this effect is smaller than that found in a study using a similar design on
household appliances [46]. Similarly, [47] find that making running
costs salient positively affects eco-friendly car choices, while [48] find
that providing information about five-year cost savings does not change
stated preferences for purchasing a gasoline, hybrid, plug-in hybrid or
battery electric vehicle. There might be different reasons why there is
still no agreement on the efficacy of this salience intervention on preferences for eco-vehicles. One reason may be that previous studies
testing for the effect of salience of future cost savings did not control for
major sources of unobserved heterogeneity that might affect the inference of preferences for EVs. In particular, previous studies did not
account for factors that might contribute to the decision to adopt EVs.
In this study, we add to previous studies testing for the effect of salience
of future cost savings by examining sources of heterogeneity.
Following [26]'s definition of collective behaviors, individuals’
choice over a particular type of vehicle might look like a custom – i.e.
independently motivated choices of many individuals that happen to be
similar to each other. Instead, it is a type of interdependent collective
behavior depending on empirical expectations (i.e. descriptive norm)
[26], since individuals prefer a type of vehicle depending, among other
factors, on what they think (or see) most individuals in their reference
group do.
Therefore, to shift collective preferences towards EVs, it is crucial to
change the empirical expectations about others’ preferences for EVs. A
norm-based intervention providing information about what most individuals do can be a useful strategy for changing empirical expectations [17]. In particular, this intervention might be effective at inducing
individuals to believe that there exists a descriptive norm about the
target behavior (i.e. collective preferences for EVs). However, to be
effective, the norm-based intervention must specify the reference group
[17]. In particular, mere membership in a social group (such as gender,
ethnicity, nationality) is one of the strongest sources for influencing
behavior [27]. Therefore, when it is salient, it motivates individuals to
conform to the behavior of others relevant to them [27].
Several studies have assessed the positive effect of norm-based interventions on pro-environment intentions or actual behavior. As an
example, informing individuals about what others do has been proven
to significantly improve household electricity use [28,29], intentions
[30] and actual recycling behavior [31], towel reuse [32], and household water use [33]. On the other hand, no previous study has yet
tested the role of a norm-based intervention in shifting preferences for
EVs. Only [34] surveyed individuals on what individuals think others
do about EVs, but it did not test for the effect of providing information
about what relevant others do about EVs on preferences to adopt EVs.
Norm-based interventions can be candidate tools for changing empirical expectations required to shift collective behaviors, such as preferences for EVs; however, their efficacy is always a matter of testing.
As an example, [35] showed that simply providing information about
others’ energy consumption might have a differential effect on individuals, depending on whether their behavior is below or above the
suggested norm (i.e., boomerang effect).
Therefore, in this study, we add to previous research on EVs by
testing whether providing normative information is effective at shifting
preferences for EVs. As suggested by [17], to increase the efficacy of the
intervention, we decrease ambiguity of the reference group, by providing information about EV registrations made by members of an existing natural group (namely, South Tyroleans).
2.3. Some sources of heterogeneity
To isolate treatment effects, it is crucial to isolate the role of different sources of hereogeneity that might interact with the choice to
adopt EVs. In particular, the choice to adopt EVs can be influenced by
certain factors underlying the decision making process, which might
differ across individuals even after random assignment to different
groups.
As an example, the fact that EVs contribute to protecting the environment might contribute to drive intentions to adopt eco-friendly
vehicles [49]. As anticipated in the introduction, South Tyroleans might
appear to be willing to protect the environment, however, their underlying motivations might differ. For some individuals, protecting the
environment can be part of their self-identity [50], while for others,
protecting the environment underlines an intrinsic motivation to contribute to any public good [51]. Therefore, to better disentangle the
effect our treatments on the likelihood to choose EVs, we control not
only for pro-environment self-identity, but also for trust, reciprocity
and altruism, these being primary predictors of decisions to contribute
to a public good [52]. The decision to purchase an EV involves a degree
of risk tolerance, since adoption consequences are uncertain [53,54];
moreover, it is an investment decision in which individuals trade-off
utility at different points in time (i.e. high purchase price versus high
future cost savings) [20,55]. In this study, we thus control for individual time discounting and risk aversion preferences.
Finally, to better isolate the effect of our interventions, we also
control for size, which has been proven to affect preferences for EVs. As
an example, those who are willing to purchase large EVs (i.e., high-end
EVs) have been found to be different from those who are not willing to
purchase large EVs. While they have been found to be less sensitive to
purchase price, their higher interest in cash incentives suggest that this
interest is not related to the monetary value per se, but more in a form
of acknowledgement of their greener behavior [56,57]. For potential
high-end EV buyers, one may also hypothesize a conspicuous consumption effect, i.e. undertaking costly actions to signal their type as
green [58], which cannot be disregarded when evaluating the treatment effect.
2.2. Salience intervention
When purchasing a vehicle, individuals are usually poorly informed
about and cognitively constrained in assessing fuel economy [36].
Moreover, EVs being more expensive than conventional vehicles, individuals might not adopt EVs, because they anchor to the high purchase price. This is due to the fact that individuals tend myopically to
overvalue things that are closer in time, and undervalue those that are
distant in future [14,37]. At the same time, due to a lack of the basic
information needed to make optimal decisions as assumed by rational
decision-making [38], individuals might also not adopt EVs because
they make errors in estimating future cost savings. As argued by [39],
this undervaluation of future consequences might help explain the
general individual underinvestment in energy-efficient technologies
(i.e. “energy paradox” [40]).
This implies that governments might provide individuals with more
information about fuel economy to help them make more informed
purchasing decisions. However, information alone does not necessarily
result in a change in behavior and, thus, in shifting preferences. Instead,
information might be more effective at changing behavior when it is
framed in a way that acknowledges individuals’ imperfect informationprocessing capacities [16]. A salience intervention [18], i.e., an
49
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
recent years.13 In the salience treatment, compared to the baseline, individuals were also provided information about the vehicles’ life-cycle
operating costs (i.e. energy costs, variable costs, fixed costs and charging infrastructure) to make the EVs’ future cost savings salient.14 Before presenting the two hypothetical scenarios, a cheap talk message
was sent to reduce the hypothetical bias, based on that of [60].15
In the third stage, respondents were asked to state a ranking over
seven incentives (car-sharing, subsidies, free parking, leasing, use of bus
lines, free highway, and better charging infrastructure) that might increase their willingness to adopt EVs.16 Then, they were presented a
reduced version of the Collective Self-Esteem scale (CSE; [61]) and an
item based on [62], to assess normative beliefs on Southern Tyroleans’
green behaviour. For the Norm treatment, these scales allow us to
isolate the moderating role that identification with the group (i.e. South
Tyroleans) and perceptions that others’ green behaviour is normal (or
not) have on the intention to conform to the suggested descriptive
norm.
In the fourth section, we included eight items from the scale developed by [50], to measure pro-environment self-identity. Then, we
adopted the reduced-form module validated by [63] as an instrument
able to capture economic preferences (time and risk preferences, altruism, reciprocity and trust) in incentivized experiments.17
In the last section, we added a survey about transport habits, car
statistics and a set of socio-demographic factors.
2.4. Research questions
The aim of this study is to test the effect of two behaviorally informed interventions on preferences for EVs. In particular, we test
whether making EV's future cost savings salient and whether informing about EV choices made by members of the reference group
are two effective interventions for promoting preferences for EVs. To
improve identification of treatment effect, we control for key factors
that might confound preferences for EVs. In particular, we control for
key economic preferences, pro-environment self-identity and size.
Following [59]'s principles for more rigorous energy social science
research, we clearly state the research questions that we aim to address in our study:
1. Is informing about what similar others do an effective intervention
for promoting preferences for EVs?
2. Is making future cost savings salient an effective intervention for
promoting preferences for EVs?
3. Does controlling for economic preferences, pro-environment selfidentity, and size improve identification of treatment effects on
preferences for EVs?
3. Experimental design
An online survey experiment was designed and 591 participants
from South Tyrol completed the survey. The survey was distributed
using newsletters and social media to reach a wider audience and was
designed using a web-based survey system.9 Treatments were randomized in the survey redirect page. In particular, participants were assigned in a between-subject fashion to either the baseline, the Salience
or the Norm treatments. The three surveys were the same except from
the additional treatment-information provided in the hypothetical
scenarios stage.10
3.2. Description of the subject pool
The sample consists of 591 participants (see Table 1). 47% of the
participants were female and 47.55% were between 26 and 40 years
old. Most participants had a high school diploma (49 %), were German
speakers (55%) and came from towns (29%) and small cities (29%). 64
% were employed and 51 % had an income between €12.000–35.000.
The majority of respondents had on average 1.79 cars in the
household. And these cars are either small (35%) or medium (35%)
sized. Among these, only 1 % were EVs. For what concerns transport
habits, the average number of flights per year is 2.6, and normally (78
%) green transports (train, foot, bus or bike) are used for the daily
activities.18
When asked to evaluate the technical factors when purchasing a
vehicle, on average respondents reported price (4.73) as the most important factor, followed by fuel type (3.79) and operating costs (3.72).
Finally, when asked to consider which incentive would help increase
their willingness to purchase an EV, respondents ranked subsidies
(5.51) as the most effective, followed by free parking (4.99) and better
charging infrastructures (4.4).
3.1. The survey
To start with, respondents were asked to state a ranking over six
technical factors (fuel type, operating costs, space, price, emissions, and
autonomy) that they might take into account when they purchase a
vehicle.11 Then, they were shown two hypothetical purchase scenarios
in which information about price and technical characteristics was
provided (i.e. speed, fuel type, autonomy, space, and emissions). In
particular, they were asked to choose first between a small-sized EV and
a small-sized conventional vehicle, and then between a large sport EV
and a large sport conventional vehicle, to account for the influence of
size.12 In this stage, we evaluate the effect of our treatments on EV
choices. In the norm treatment, compared to the baseline, individuals
were also provided with a descriptive norm about EV choices made by
the members of the reference group (South Tyroleans) in the most
13
The graph showing the registrations made by South Tyroleans is available
in the Appendix, and was preceded by the question: “Did you know that South
Tyroleans increased in new Electric Vehicle registrations by 178 % in the period
2013–2017?”.
14
The life-cycle operating costs are available in the Appendix and were based
on the results of the Total Cost of Ownership (TCO) calculator developed by the
Oeko-Institut e.V.
15
“The experience from previous similar surveys is that people often state a
higher willingness to pay than what one is actually willing to pay for the good.
This difference in percentage is due to the fact that the purchasing questions are
hypothetical. To avoid this so-called hypothetical bias, it is important that you
make each of the following choices as if you would actually be facing these
exact choices in a store.”
16
Ranking was collected on a scale 1–7, with 1 meaning “Not at all important”, and 7 meaning “Very important”.
17
Answers in the third and fourth stages were collected on a Likert scale 1–7,
with 1 meaning “I totally disagree”, and 7 meaning “I totally agree”.
18
The extensive use of green transport for daily activities is representative of
Southern Tyroleans’ high environmental concerns, as anticipated in the
Introduction.
9
We used the web-based survey system Opinio.
The survey is available in the Appendix.
11
Ranking was collected on a scale 1–6, with 1 meaning “Not at all important”, and 6 meaning “Very important”.
12
To make the two scenarios plausible, we focused on conventional and
electric-fuelled cars that are in the top ten models sold in Italy and are comparable in size and technical features. Information about price, autonomy,
space, emissions, etc., was taken from the most known vehicle guide in Italy
(alVolante), so as to be accessible to all respondents. Since the aim was to select
vehicles comparable in size and technical features, it was not possible to find
vehicles with the same price and total costs. In particular, the price differential
between the small EV and the conventionally fuelled vehicle is 11.000, and the
total cost differential between the small EV and the conventionally fuelled
vehicle −7.266. In contrast, the price differential between the big EV and the
conventionally fuelled vehicle is 20.000, and the total cost differential between
the big EV and the conventionally fuelled vehicle −7.721.
10
50
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Table 1
Description of the subject pool.
Table 2
EVs choices.
N
%
Demographics
Female
Age [18–25]
Age [26–40]
Age [41–65]
Age [ > 65]
Language [Italian]
Language [German]
Language [Ladin]
Language [Other]
City size [ < 2.000]
City size [2.000–10.000]
City size [10.001–100.000]
City size [ > 100.000]
Job [Unemployed]
Job [Employed]
Job [Self-employed]
Job [Student]
Education [Primary school]
Education [High school]
Education [Graduate]
Education [Post graduate]
Income [ < 12.000]
Income [12.001–35.000]
Income [35.001–60.000]
Has children
N house inhabitants
584
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
591
47%(0.5)
21.66%(0.41)
47.55%(0.5)
29.44%(0.46)
1.35%(0.12)
41%(0.41)
55%(0.5)
2%(0.15)
2%(0.14)
15%(0.35)
29%(0.46)
29%(0.46)
27%(0.44)
2.5%(0.16)
64%(0.48)
14%(0.34)
15%(0.36)
0.6%(0.082)
49%(0.5)
11%(0.31)
34%(0.48)
13.6%(0.34)
51%(0.5)
23%(0.42)
36%(0.48)
2.86(1.25)
Car and transport statistics
N cars
Has small car
Has medium car
Has big car
Already uses EV
Green transport habits
N flights last year
591
591
591
591
591
591
591
1.79(0.98)
35%(0.48)
35%(0.47)
15.6%(0.364)
1%(0.1)
78%(0.41)
2.6(1.93)
Ranking: important factors when purchasing a car
Fuel type
591
Operating costs
591
Space
591
Price
591
CO2
591
Autonomy
591
3.79(1.63)
3.72(1.47)
3.02(1.53)
4.73(1.56)
2.81(1.62)
2.94(1.62)
Ranking: incentives to purchase EV
Car-sharing
Subsidy
Free parking
Leasing
Bus line
Free highway
Better charging infrastructure
3(2.08)
5.51(1.72)
4.99(1.59)
3.37(1.59)
2.94(1.76)
3.78(1.68)
4.4(2.01)
591
591
591
591
591
591
591
Baseline
Salience
Norm
% EV choices
Treatment-baseline comparison
51.2% [46.4%–56%]
59.4% [54.5%–64.4%]
53% [48.2%–58%]
0.0223*
0.5729
p-Values of Wilcoxon rank-sum test:
***p < 0.001.
**p < 0.01.
o
p < 0.1.
95% Confidence Intervals in parentheses [64].
* p < 0.05.
Table 3
Logit regression models: EV choices.
(1)
(2)
(3)
Variables
Norm
Salience
Big size
Willingness to risk
Willingness to delay
Altruism
Reciprocity
Trust
Social ID
Pro-environment ID
Green beliefs
Individual
characteristics
Constant
EV choice
0.0784(0.141)
0.334(0.143)*
No
EV choice
0.0909(0.163)
0.382(0.163) *
1.596(0.125)***
−0.0156(0.0485)
0.185(0.0682)**
0.0220(0.0633)
−0.132(0.0900)
0.00475(0.0487)
0.00636(0.0662)
0.428(0.0752) ***
−0.0721(0.0573)
No
EV choice
0.166(0.173)
0.413(0.167) *
1.739(0.140) ***
−0.0299(0.0539)
0.223(0.0722) **
0.00640 (0.0708)
−0.210(0.0954) *
0.00201(0.0504)
0.0454(0.0754)
0.329(0.0805) ***
−0.0979(0.0601)
Yes
0.0486 (0.0986)
−0.769(0.563)
−4.807(1.744)
ID
Observations
Pseudo R2
591
1182
0.0035
591
1182
0.1378
584
1168
0.1953
**
Robust standard errors in parentheses.
*** p < 0.001.
** p < 0.01.
* p < 0.05.
o
p < 0.1.
This insight is confirmed by the logit regression Model 1 (Table 3),
in which we estimate the probability that the EV is chosen using as
explanatory variables the two treatments (Norm is equal to 1 when
respondents are assigned to the norm-based treatment and zero otherwise; Salience is equal to 1 when respondents are assigned to the salience treatment and zero otherwise). To measure the positive effect of
Salience, we need to convert the coefficients βi of the logistic regression
from log-odds units into odds-ratios Exp(βi). In particular, the value of
odds ratios for Salience is 1.396 in Model (1), meaning that, compared
to the Baseline, the odds ratio of choosing EVs increases by 1.396 times
in Model (1).
Therefore, results:
4. Results
In this section, we first answer the three research questions. Then,
we examine results with a more exploratory approach.
4.1. Confirmatory analysis
1. do not confirm that informing about what similar others do is an
effective intervention for promoting preferences for EVs;
2. confirm that making future cost savings salient is an effective intervention for promoting preferences for EVs.
To answer the first two questions, we assess whether EV choices
differ across treatments. Table 2 shows that EV choices are affected only
by the salience treatment. We find that, compared to the Baseline
(51.2%) the percentage of EV choices is significantly higher for the
salience intervention (59.4%) (Table 2). On the other hand, while the
percentage of EV choices in the Norm treatment(53%) is higher than
that in the Baseline, this is not statistically significant.
To answer the third question, we eliminate potential sources of
unobserved heterogeneity that might correlate with EV choices. In
particular, we estimate the probability that the EV is chosen using as
51
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Table 4
Conditional marginal effects.
Delta-method
1. Salience
Below median WT
delay
No
Yes
Observations
0.0895(0.041)
0.037(0.0453)
591
*
Below median proenvironment ID
Big size
0.105(0.0394)**
0.037(0.0439)
591
0.095(0.044) *
0.04996(0.0392)
591
Robust standard errors in parentheses.
***p < 0.001.
** p < 0.01.
* p < 0.05.
o
p < 0.1.
explanatory variables not only the two treatment dummies, but also the
measures of pro-environment identity, key economic preferences
(willingness to risk, willingness to delay benefits, altruism, reciprocity
and trust), and size. Moreover, we add measures of beliefs about other
southern Tyroleans’ green behaviour and social identification as these
might confound the effect of the norm-based treatment.
First, the logit regression (Model (2) – Table 3) confirms that there
are positive significant differences only between EV choices made in the
baseline and in the salience treatment also when we control for sources
of heterogeneity (the value of odds ratios for Salience is 1.465 in Model
(2) meaning that, compared to the Baseline, the odds ratio of choosing
EVs increases by 1.465 times when controlling for sources of heterogeneity). Second, it suggests treatment effect heterogeneity. In particular, those who display higher capacity to value future events, i.e.
higher willingness to delay benefits, and those who identify themselves
as pro-environment individuals, are more likely to choose EVs. Moreover, when the choice is between a big conventional fuelled vehicle and
a big EV, the likelihood to choose EVs increases, compared with when
the choice is between a small conventionally fuelled vehicle and a small
EV.
To confirm treatment effect heterogeneity, we look at heterogeneous responses to Salience with respect to the measures of preferences to delay benefits, pro-environment self-identity and size. In
particular, to see whether the salience treatment also has a positive
impact on individuals with low time-preferences and pro-environment
self-identity values, we categorize individuals depending on whether
their level of willingness to delay benefits and pro-environment selfidentity is below the median level of willingness to delay and pro-environment self-identity in the sample. Moreover, to see whether the
salience treatment also has a positive impact on individuals who are not
motivated by a potential conspicuous consumption effect, we also see
the marginal effect of Salience for different vehicle sizes.
The marginal effects of Salience (Table 4 and Fig. 4) show that the
difference between responses of those assigned to the baseline and
those assigned to the salience treatment is significantly different only
for high values of pro-environment identity and willingness to delay,
and when the vehicle size is big. Hence, compared to the baseline, the
positive difference in the probability to choose EVs between those who
are assigned to the salience treatment and to the baseline increases
significantly conditionally on size, individual pro-environment selfidentity and time preferences.
Therefore, results confirm that:
Fig. 4. Conditional marginal effects.
4.2. Exploratory analysis
The choice to adopt EVs can be influenced by further factors that
might differ across individuals, even after the random assignment to
different treatment groups. Therefore, in Model 3 (Table 3), we estimate the probability that the EV is chosen using as explanatory variables other individual characteristics, in addition to the variables used
in Model 2 (see Table A2 in the Appendix for more detailed results). We
find that the positive effect of Salience is still significant (the value of
odds ratios for Salience is 1.511 in Model (3) meaning that, compared to
the Baseline, the odds ratio of choosing EVs increases by 1.511 times
when controlling also for individual characteristics).
Interestingly, we find that when we account for individual characteristics, the negative coefficient associated with preferences for
Reciprocity becomes significant: without adjusting for individual characteristics, unexplained variability is large and the effect from higher
preferences for reciprocity is too small to emerge. As suggested in
Section 2.3, displaying preferences for reciprocity can provide the basis
1. Controlling for economic preferences, pro-environment self-identity,
and size improves identification of (salience) treatment effects on
preferences for EVs.
52
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
for individuals to decide to contribute to a public good. This implies
that reciprocal individuals are more willing to contribute to a public
good when others are also willing to contribute [65], but when they
expect others to free ride from contributing, they “punish” others by
free riding too [66]. Looking at individual characteristics, we find that,
compared to German speakers, those who speak languages other than
Ladin and Italian are more likely to choose EVs. Similarly, we find that
compared to those who are employed, those who are not currently
working and – only slightly – those who are self-employed are more
likely to choose EVs. Not surprisingly, those who are already using EVs
are more likely to choose EVs. Finally, we find that for those whose CO2
emissions and fuel type are very important when purchasing a vehicle,
the likelihood to choose EVs increase substantially, and slightly for
those whose operating costs, purchasing price and autonomy are very
important when purchasing a vehicle. On the other hand, compared to
those who come from a big city, respondents coming from small towns,
are less likely to choose EVs.
buy EVs rather than actual purchasing choices. However, while there is
some evidence suggesting that purchasing intentions do not predict
well actual purchasing choices [69], we believe that this study provides
useful methodological insights to policy makers on how to design
complementary interventions aimed at promoting EV adoption. Future
research could attempt to validate the findings from this study in the
field. Secondly, to isolate the effect of our treatments, we controlled for
sources of heterogeneity drawing from the behavioural economic literature. Although the focus of the study was to identify treatment effects, accounting for other factors that might correlate the willingness
to choose EVs, such as choice overload, perceived behavioral control,
pioneering spirit and personality traits such as openness [23], would
improve the validity of the findings. Of course, this would make the
model less parsimonious.
6. Conclusions
To be effective on the environment, EVs have to be adopted by a
large mass of individuals. However, the market, even in regions (such as
South Tyrol) that are ready to tackle some material barriers related to
EV adoption, is still dominated by conventional vehicles. Several policy
mechanisms are being implemented to incentivize a market for EVs.
However, these have been proven to be costly and not fully effective.
One reason lies in the fact that these policies tackle most of the material
barriers related to EV diffusion, but often disregard not only the role of
cognitive biases associated with the decision problem of adopting EVs,
but also potential heterogeneous responses to interventions.
In this study, we examined to what extent two interventions enriched with behavioral economic insights are effective at simplifying
the decision problem related to EV adoption. In particular, we tested
the effect of a norm-based intervention, i.e., providing information
about EV choices made by the reference group, and that of a salience
intervention, i.e., making future cost savings salient, on EV choices.
Moreover, to better isolate treatment effects, we controlled for major
sources of individual heterogeneity that might affect EV choices,
namely economic preferences, pro-environment self-identity and size.
We find that while providing information about the choices of the
reference group has a positive coefficient on the likelihood to choose
EVs, it does not reach a significant level. On the other hand, we find
that making future cost savings salient is an effective strategy to increase the likelihood to choose EVs, and that heterogeneity in behavior
affects responses to the treatment. While this result confirms previous
studies already testing the role of salience on EV choices, it provides
new insights by proving that it can be effective on individuals who are
already motivated to choose EVs.
Overall, results suggest the need to always test the type of information embedded in norm-based interventions, since individuals’
willingness to change depends on the information provided. At the
same time, they provide evidence of the need to measure heterogeneity
in behavior before implementing information strategies aimed at
shifting preferences for EVs. Therefore, they highlight the importance
of always scanning the target context before implementing the intervention in the field [70].
Practically, in the context of large-scale surveys, governments might
promote the inclusion of items, such as those developed by [63], to
capture predictors of heterogeneity in behavior. This would allow a
better grasp of behavioral levers (barriers) that might enhance (inhibit)
the efficacy of a particular information strategy. In particular, governments can start targeting the subgroups displaying a higher probability
to positively react to the salience intervention. Thereafter, they might
implement a large-scale norm-based intervention, as soon as individuals
will start reacting to the salience intervention (the norm-based
5. Discussion
In this study, we experimentally tested how a salience and a normbased intervention affect individual willingness to choose EVs in South
Tyrol. Using an online survey study, we randomly assigned residents in
South Tyrol to hypothetical scenarios to buy either a conventionally
fuelled vehicle or an EV, which were informing either about EV choices
made by South Tyroleans in the last year, or which were making salient
future cost savings. In addition, we isolated treatment effects by controlling for major sources of heterogeneity that might impact EV
choices. While we found a positive effect of the salience intervention on
EV choices, there was no effect for the norm-based intervention. While
there might be several reasons why the norm-based intervention was
ineffective, we speculate that it is related to asymmetry in what individuals inferred from the normative information we provided [17]. In
particular, different individuals have not only different empirical expectations about the number of people following the prescribed descriptive norm, but also different thresholds for how many people are
necessary to conform to the norm [67,68]. Therefore, we believe that
the descriptive norm we provided was ineffective, because it did not
reach the threshold of people required to change empirical expectations. In particular, individuals might have perceived EV registrations
made by South Tyroleans in the last years still as too low. Future research testing norm-based interventions should investigate whether
individuals perceive that the normative information provided represents a sufficient threshold to change their empirical expectations.
For what concerns the salience intervention, we confirmed the
findings of previous studies supporting that it is effective at shifting
preferences for EVs [44,47]. In addition, by controlling for major
sources of heterogeneity, we showed that the effect is limited to those
who display high values in pro-environment self-identity, and preferences for big size and for delaying benefits.
Finally, when looking at the effect of individual characteristics, we
found reciprocal individuals are less likely to choose EVs. This suggests
that while they might share the belief that EVs are good for the environment, they might believe that too few individuals are adopting
them: the beliefs that there are non-contributors to the public good
induce reciprocal types to make selfish choices, i.e. not to contribute to
protect the environment by choosing EVs.
Overall, our results contribute to the research on EV adoption, by
experimentally testing the impact of two cost-effective interventions on
EV choices, and by isolating the effect through the account of major
sources of heterogeneity in behavior. Nevertheless, our study presents
some limitations. Firstly, this study measured hypothetical choices to
53
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Acknowledgements
intervention might be more effective when it provides information
about a high percentage of individuals already engaging in the target
behavior).
Hence, while we conclude that norm-based interventions should be
implemented at a later stage, future research should examine how other
behaviorally informed interventions affect EV adoption.
Overall, our study shows that information strategies acknowledging
individuals’ imperfect information-processing capacities can be costeffective interventions for promoting EV adoption. Therefore, it is of
practical relevance to local governments that face an increasing need to
develop complementary interventions to promote alternative-fuelled
vehicles’ diffusion.
The research leading to these results is part of the project Culture
Building and territorial Development. Come preparare un territorio alla
rivoluzione disruptive dell’e-mobility, funded by Jaguar Land Rover Italia,
in collaboration with Giulia Isetti, Philipp Corradini, Mirjam Gruber
and Gerhard Vanzi (Eurac Research). We are especially grateful to the
editors and the anonymous referees for their encouragement and constructive advice. We would also like to acknowledge the colleagues of
the Urban and Regional Energy System at the Institute for Renewable
Energy - Eurac Research, who supported this research by providing
their feedback. The authors thank the Department of Innovation,
Research and University of the Autonomous Province of Bozen/Bolzano
for covering the Open Access publication costs.
Conflict of interest
The author declare that there are no conflict of interests.
Appendix A
Figs. A1–A3
Fig. A1. Baseline. treatment.
54
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Fig. A2. Salience treatment.
55
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Fig. A3. Norm treatment.
Table A1
Table A1
Description of costs in the Salience treatment.
Infrastructural costs
Fuel/electricity costs
Variable costs
Fixed costs
Total costs
Price
Car type: small
EV
Conventionally fuelled
Difference
686
0
686
2776
8409
−5633
1812
3062
−1250
6339
7408
−1069
11,613
18,879
−7266
39,150
28,150
11,000
Car type: big
EV
Conventionally fuelled
Difference
686
0
686
3409
9074
−5665
2593
4059
−1466
7685
8961
−1276
14,373
22,094
−7721
88,000
68,000
20,000
56
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
Table A2
Logit regression model (3) – individual characteristics.
(3)
Variables
Norm
Salience
Big size
Willingness to risk
Willingness to delay
Altruism
Reciprocity
Trust
Social ID
Pro-environment ID
Green beliefs
Female
Age [18–25]
Age [26–40]
Age [41–65]
Language [Italian]
Language [Ladin]
Language [Other]
City size [ < 2.000]
City size [2.000–10.000]
City size [10.001–100.000]
Job [Unemployed]
Job [Self-employed]
Student
Education [High school]
Education [Graduate]
Education [Post graduate]
Income [12.001–35.000]
Income [35.001–60.000]
N house inhabitants
Has children
N cars
Already uses EV
Has small car
Has big car
N flights last year
Ranking [Fuel type]
Ranking [CO2]
Ranking [Operating costs]
Ranking [Price]
Ranking [Autonomy]
Ranking [Car-sharing]
Ranking [Subsidy]
Ranking [Free parking]
Ranking [Leasing]
Ranking [Bus line]
Ranking [Better charging infrastructure]
Constant
ID
Observations
Pseudo R2
EV choice
0.166(0.173)
0.413(0.167) *
1.739(0.140)***
−0.0299(0.0539)
0.223(0.0722) **
0.00640(0.0708)
−0.210(0.0954) *
0.00201(0.0504)
0.0454(0.0754)
0.329(0.0805) ***
−0.0979(0.0601)
0.0600(0.152)
−0.107(0.704)
0.00900(0.654)
−0.137(0.647)
0.247(0.181)
0.612(0.520)
0.930(0.435) *
−0.412 (0.244) *
−0.118(0.209)
−0.252(0.193)
1.044(0.422) *
0.398(0.220)o
0.210(0.319)
−0.237(0.326)
−0.143(0.357)
−0.0742(0.382)
0.239(0.207)
0.134(0.253)
0.0705(0.0670)
0.151(0.212)
−0.00352(0.0846)
2.098(1.043) *
−0.139(0.171)
−0.321(0.208)
0.00821(0.0427)
0.296(0.0617) ***
0.345(0.0613) ***
0.130(0.0645) *
0.153(0.0651) *
0.125(0.0572) *
−0.0394(0.0510)
0.0488(0.0620)
0.0954(0.0659)
0.0352(0.0626)
−0.0453(0.0737)
−0.00921(0.0582)
−4.807(1.744) **
584
1,168
0.1953
Robust standard errors in parentheses
*** p < 0.001.
** p < 0.01.
* p < 0.05.
o
p < 0.1.
57
N. DellaValle and A. Zubaryeva
WTD
How willing are you to give up something
that is beneficial for you today in order to benefit more from that in
future?
WTR
How willing are you to take risks?
Altruism
How willing are you to donate to causes without expecting anything in
return?
Reciprocity
When someone does me a favour I am willing to return it
Trust
I assume that people have only the best intentions
58
Environmental ID
I think of myself as someone who is very concerned with environmental
issues
Respecting the environment is an important part of my identity
I turn off lights when I do not use them
I buy energy- efficient appliances
I drive economically (e.g. braking or accelerating gently)
I walk, cycle or take public transport for short journeys (i.e. trips of < 3
miles)
I share a car journey with someone else
I cut down on the amount you fly
Normative beliefs
Many people in South Tyrol engage in environmentally friendly behaviour
Disagree
Disagree Somewhat
Neutral
Agree somewhat
Agree
Strongly agree
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
4
5
5
5
5
5
6
6
6
6
6
7
7
7
7
7
1
1
2
2
3
3
4
4
5
5
6
6
7
7
1
2
3
4
5
6
7
1
1
2
2
3
3
4
4
5
5
6
6
7
7
1
1
2
2
3
3
4
4
5
5
6
6
7
7
Energy Research & Social Science 55 (2019) 46–61
CSR
Overall, my social groups are considered well by others
Overall, my group memberships have very little to do with how I feel
about myself
In general, I’m glad to be a member of the social groups I belong to
In general, belonging to social groups is an important part of my self
image
Strongly disagree
N. DellaValle and A. Zubaryeva
Transport questions
How many cars do you have in your households?
Do you have a driving license?
Have you recently bought a new/ nearly new car or a used car in the last
24 months?
Which car type do you use?
Which fuel do you use?
59
1
Yes
Yes
2
Not
Not
3
4
5
6
7
Small
Petrol
Medium
Diesel
Large
gas
Sport
Hybrid
SUV
Electric (not plugin)
VAN
electric
Other
other
Do you own or lease the car that you use?
Own
Lease
For how many km is this car driven?
Less than 2000 km
2.000–10.000 km
Someone in the household
owns it
10.001–20,000
Someone in the household
leases it
20.0001–30,000
30,001–40,000
40.001–50.000
For which activity do you most use this car?
Commute to work/
school
1
Car
Car
Car
Car
Daily activities
Shopping
Business trips
Leisure
2
Bicycle
Bicycle
Bicycle
Bicycle
3
Scooter
Scooter
Scooter
Scooter
4
Train
Train
Train
Train
5
Bus/tram/metro
Bus/tram/metro
Bus/tram/metro
Bus/tram/metro
Going on
holiday
6
Airplane
Airplane
Airplane
Airplane
More than
50.0001
other
26–40
Male
2001–10.000
2
Not
Not
Not
Employed
41–65
More than 65
10,001–100.000
3
More than 100.000
4
More than 4
Student
Voluntary worker
Retired
Family career
High school
24.001–35.000
Bachelor
35.001–60.000
Master
More than 60.000
€
PhD
How many flights have you made last year?
Which transport vehicle do you use for going on holidays?
Which transport vehicle do you use for commuting to work/school?
Which transport vehicle do you use for shopping?
Which transport vehicle do you use for leisure (sport, social visit, day
trip)?
Sociodemographics
Which is your age?
Which is your gender?
How big (in terms of inhabitants) is the place where you live?
How many people live in your house?
Are you married or do you live with your partner?
Do you have disabilities?
Do you have children?
What describes you best?
Which is your highest educational qualification?
Which is approximately your gross household annual income?
18–25
Female
Up to 2000
1
Yes
Yes
Yes
Unemployed/Job
seeker
Elementary
Less than 12,000 €
Middle
12.001–24.000
7
Other
Other
Other
Other
Energy Research & Social Science 55 (2019) 46–61
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
References
experiment on curbside recycling, Basic Appl. Soc. Psychol. 21 (1999) 25–36.
[32] W.P. Schultz, A.M. Khazian, A.C. Zaleski, Using normative social influence to promote conservation among hotel guests, Soc. Infl. 3 (2008) 4–23.
[33] P.J. Ferraro, J.J. Miranda, M.K. Price, The persistence of treatment effects with
norm-based policy instruments: evidence from a randomized environmental policy
experiment, Am. Econ. Rev. 101 (2011) 318–322.
[34] J. Jansson, A. Nordlund, K. Westin, Examining drivers of sustainable consumption:
the influence of norms and opinion leadership on electric vehicle adoption in
Sweden, J. Clean. Prod. 154 (2017) 176–187.
[35] P.W. Schultz, J.M. Nolan, R.B. Cialdini, N.J. Goldstein, V. Griskevicius, The constructive, destructive, and reconstructive power of social norms, Psychol. Sci. 18
(2007) 429–434.
[36] H. Allcott, C. Knittel, Are Consumers Poorly Informed about Fuel Economy?
Evidence from Two Experiments, Technical Report, Natl. Bureau Econ. Res. (2017).
[37] S. Frederick, G. Loewenstein, T. O’donoghue, Time discounting and time preference: a critical review, J. Econ. Liter. 40 (2002) 351–401.
[38] T.S. Turrentine, K.S. Kurani, Car buyers and fuel economy? Energy Policy 35 (2007)
1213–1223.
[39] H. Allcott, N. Wozny, Gasoline prices, fuel economy, and the energy paradox, Rev.
Econ. Stat. 96 (2014) 779–795.
[40] A.B. Jaffe, R.N. Stavins, The energy-efficiency gap what does it mean? Energy
Policy 22 (1994) 804–810.
[41] S.L. Heinzle, Disclosure of energy operating cost information: a silver bullet for
overcoming the energy-efficiency gap? J. Consum. Policy 35 (2012) 43–64.
[42] R.G. Newell, J. Siikamäki, Nudging energy efficiency behavior: the role of information labels, J. Assoc. Environ. Resour. Econ. 1 (2014) 555–598.
[43] M. Andor, A. Gerster, S. Sommer, Consumer inattention, heuristic thinking and the
role of energy labels, Ruhr Economic Papers (2017).
[44] R. Wüstenhagen, K. Sammer, Wirksamkeit umweltpolitischer anreize zum kauf
energieeffizienter fahrzeuge: eine empirische analyse schweizer automobilkunden,
Zeitschrift für angewandte Umweltforschung (J. Environ. Res.) 18 (2007) 61–78.
[45] J. Kaenzig, R. Wüstenhagen, The effect of life cycle cost information on consumer
investment decisions regarding eco-innovation, J. Ind. Ecol. 14 (2010) 121–136.
[46] K. Sammer, R. Wüstenhagen, The influence of eco-labelling on consumer behaviourresults of a discrete choice analysis for washing machines, Bus. Strat. Environ. 15
(2006) 185–199.
[47] C. Codagnone, G.A. Veltri, F. Bogliacino, F. Lupiáñez-Villanueva, G. Gaskell,
A. Ivchenko, P. Ortoleva, F. Mureddu, Labels as nudges? An experimental study of
car eco-labels, Econ. Polit. 33 (2016) 403–432.
[48] J. Dumortier, S. Siddiki, S. Carley, J. Cisney, R.M. Krause, B.W. Lane, J.A. Rupp,
J.D. Graham, Effects of providing total cost of ownership information on consumers’
intent to purchase a hybrid or plug-in electric vehicle, Transport. Res. A: Policy
Pract. 72 (2015) 71–86.
[49] J. Jansson, A. Marell, A. Nordlund, Green consumer behavior: determinants of
curtailment and eco-innovation adoption, J. Consum. Market. 27 (2010) 358–370.
[50] L. Whitmarsh, S. O’Neill, Green identity, green living? The role of pro-environmental self-identity in determining consistency across diverse pro-environmental
behaviours, J. Environ. Psychol. 30 (2010) 305–314.
[51] R. Bénabou, J. Tirole, Incentives and prosocial behavior, Am. Econ. Rev. 96 (2006)
1652–1678.
[52] P. Kollock, Social dilemmas: the anatomy of cooperation, Annu. Rev. Sociol. 24
(1998) 183–214.
[53] M. Farsi, Risk aversion and willingness to pay for energy efficient systems in rental
apartments, Energy Policy 38 (2010) 3078–3088.
[54] Y. Qiu, G. Colson, C. Grebitus, Risk preferences and purchase of energy-efficient
technologies in the residential sector, Ecol. Econ. 107 (2014) 216–229.
[55] R.G. Newell, J. Siikamäki, Individual time preferences and energy efficiency, Am.
Econ. Rev. 105 (2015) 196–200.
[56] C.D. Higgins, M. Mohamed, M.R. Ferguson, Size matters: how vehicle body type
affects consumer preferences for electric vehicles, Transport. Res. A: Policy Pract.
100 (2017) 182–201.
[57] S. Hardman, E. Shiu, R. Steinberger-Wilckens, Comparing high-end and low-end
early adopters of battery electric vehicles, Transport. Res. A: Policy Pract. 88 (2016)
40–57.
[58] S.E. Sexton, A.L. Sexton, Conspicuous conservation: The Prius halo and willingness
to pay for environmental bona fides, J. Environ. Econ. Manage. 67 (2014) 303–317.
[59] B.K. Sovacool, J. Axsen, S. Sorrell, Promoting novelty, rigor, and style in energy
social science: towards codes of practice for appropriate methods and research
design, Energy Res. Soc. Sci. (2018).
[60] R.G. Cummings, L.O. Taylor, Unbiased value estimates for environmental goods: a
cheap talk design for the contingent valuation method, Am. Econ. Rev. 89 (1999)
649–665.
[61] R. Luhtanen, J. Crocker, A collective self-esteem scale: self-evaluation of one's social
identity, Pers. Soc. Psychol. Bull. 18 (1992) 302–318.
[62] K. Ando, S. Ohnuma, E.C. Chang, Comparing normative influences as determinants
of environmentally conscious behaviours between the USA and Japan, Asian J. Soc.
Psychol. 10 (2007) 171–178.
[63] A. Falk, A. Becker, T. Dohmen, D. Huffman, U. Sunde, The preference survey
[1] International Energy Agency, Global EV Outlook 2016, (2016) Online (accessed
14.06.18) https://www.iea.org/publications/freepublications/publication/Global_
EV_Outlook_2016.pdf.
[2] Z. Rezvani, J. Jansson, J. Bodin, Advances in consumer electric vehicle adoption
research: a review and research agenda, Transport. Res. D: Transp. Environ. 34
(2015) 122–136.
[3] N. Bergman, Stories of the future: personal mobility innovation in the United
Kingdom, Energy Res. Soc. Sci. 31 (2017) 184–193.
[4] J. Sousa Lourenco, E. Ciriolo, S. Rafael Rodrigues Viera De Almeida, X. Troussard,
Behavioural Insights Applied to Policy – European Report 2016, Publications Office
of the European Union, 2016.
[5] European Commission, A Roadmap for moving to a competitive low carbon
economy in 2050, (2011) Online (accessed 31.05.18) https://eur-lex.europa.eu/
legal-content/EN/TXT/PDF/?uri=CELEX:52011DC0112&from=EN.
[6] A. Zubaryeva, C. Thiel, E. Barbone, A. Mercier, Assessing factors for the identification of potential lead markets for electrified vehicles in Europe: expert opinion
elicitation, Technol. Forecast. Soc. Change 79 (2012) 1622–1637.
[7] C. Thiel, J. Krause, P. Dilara, Electric vehicles in the EU from 2010 to 2014 – is full
scale commercialisation near, JRC Sci. Policy Rep. (2015).
[8] A. Zubaryeva, C. Thiel, Analyzing potential lead markets for hydrogen fuel cell
vehicles in Europe: expert views and spatial perspective, Int. J. Hydrog. Energy 38
(2013) 15878–15886.
[9] A. Zubaryeva, C. Thiel, N. Zaccarelli, E. Barbone, A. Mercier, Spatial multi-criteria
assessment of potential lead markets for electrified vehicles in Europe, Transport.
Res. A: Policy Pract. 46 (2012) 1477–1489.
[10] E.H. Noppers, K. Keizer, J.W. Bolderdijk, L. Steg, The adoption of sustainable innovations: driven by symbolic and environmental motives, Glob. Environ. Change
25 (2014) 52–62.
[11] F. Ecker, U.J. Hahnel, H. Spada, Promoting decentralized sustainable energy systems in different supply scenarios: the role of autarky aspiration, Front. Energy Res.
5 (2017) 14.
[12] Klimaland, PIANO CLIMA Energia-Alto Adige-2050, (2011) Online (accessed
14.06.18) http://www.provincia.bz.it/agenzia-ambiente/download/PianoClima_
Energia_AA2050_Ansicht.pdf.
[13] M.G. Prina, M. Cozzini, G. Garegnani, G. Manzolini, D. Moser, U.F. Oberegger,
R. Pernetti, R. Vaccaro, W. Sparber, Multi-objective optimization algorithm coupled
to EnergyPLAN software: the EPLANopt model, Energy 149 (2018) 213–221.
[14] G. Loewenstein, D. Prelec, Anomalies in intertemporal choice: evidence and an
interpretation, Q. J. Econ. 107 (1992) 573–597.
[15] R.H. Thaler, C.R. Sunstein, Nudge: improving decisions about health, Wealth
Happiness 6 (2008).
[16] C.R. Sunstein, Misconceptions about nudges, J. Behav. Econ. Policy 2 (1) (2018)
61–67.
[17] C. Bicchieri, E. Dimant, Nudging with care: the risks and benefits of social information, Public Choice (2019) Forthcoming.
[18] M. Lehner, O. Mont, E. Heiskanen, Nudging – a promising tool for sustainable
consumption behaviour? J. Clean. Prod. 134 (2016) 166–177.
[19] A. Falk, A. Becker, T.J. Dohmen, B. Enke, D. Huffman, The nature and predictive
power of preferences: global evidence, CEPR Discussion Papers (2015).
[20] H. Allcott, D. Taubinsky, Evaluating behaviorally motivated policy: experimental
evidence from the lightbulb market, Am. Econ. Rev. 105 (2015) 2501–2538.
[21] W. Li, R. Long, H. Chen, Consumers’ evaluation of national new energy vehicle
policy in china: an analysis based on a four paradigm model, Energy Policy 99
(2016) 33–41.
[22] Y. Zhang, Y. Yu, B. Zou, Analyzing public awareness and acceptance of alternative
fuel vehicles in china: the case of EV, Energy Policy 39 (2011) 7015–7024.
[23] W. Li, R. Long, H. Chen, J. Geng, A review of factors influencing consumer intentions to adopt battery electric vehicles, Renew. Sustain. Energy Rev. 78 (2017)
318–328.
[24] E.H. Green, S.J. Skerlos, J.J. Winebrake, Increasing electric vehicle policy efficiency
and effectiveness by reducing mainstream market bias, Energy Policy 65 (2014)
562–566.
[25] C. Bicchieri, Diagnosing Norms, Oxford University Press, 2016.
[26] C. Bicchieri, The Grammar of Society, Technical Report, Cambridge University
Press, 2006.
[27] H. Tajfel, M.G. Billig, R.P. Bundy, C. Flament, Social categorization and intergroup
behaviour, Eur. J. Soc. Psychol. 1 (1971) 149–178.
[28] J.M. Nolan, P.W. Schultz, R.B. Cialdini, N.J. Goldstein, V. Griskevicius, Normative
social influence is underdetected, Pers. Soc. Psychol. Bull. 34 (2008) 913–923.
[29] H. Allcott, Social norms and energy conservation, J. Publ. Econ. 95 (2011)
1082–1095.
[30] F. Fornara, G. Carrus, P. Passafaro, M. Bonnes, Distinguishing the sources of normative influence on proenvironmental behaviors: the role of local norms in
household waste recycling, Group Process. Interg. Relat. 14 (2011) 623–635.
[31] P.W. Schultz, Changing behavior with normative feedback interventions: a field
60
Energy Research & Social Science 55 (2019) 46–61
N. DellaValle and A. Zubaryeva
[64]
[65]
[66]
[67]
schemes, Am. Econ. Rev. 97 (2007) 999–1012.
[68] M. Granovetter, Threshold models of collective behavior, Am. J. Sociol. 83 (1978)
1420–1443.
[69] V.G. Morwitz, J.H. Steckel, A. Gupta, When do purchase intentions predict sales?
Int. J. Forecast. 23 (2007) 347–364.
[70] N. DellaValle, A. Bisello, J. Balest, In search of behavioural and social levers for
effective social housing retrofit programs, Energy Build. 172 (2018) 517–524.
module: a validated instrument for measuring risk, time, and social preferences,
Netspar discussion paper (2016).
G. Cumming, The new statistics: why and how, Psychol. Sci. 25 (2014) 7–29.
R. Sugden, Reciprocity: the supply of public goods through voluntary contributions,
Econ. J. 94 (1984) 772–787.
E. Fehr, S. Gächter, Fairness and retaliation: the economics of reciprocity, J. Econ.
Perspect. 14 (2000) 159–181.
D. Sliwka, Trust as a signal of a social norm and the hidden costs of incentive
61