Economic Analysis and Policy 76 (2022) 373–384

Contents lists available at ScienceDirect

Economic Analysis and Policy

journal homepage: www.elsevier.com/locate/eap

Analyses of topical policy issues

Effects of policy instruments on electric scooter adoption in

Jakarta, Indonesia: A discrete choice experiment approach✩
∗
Siwon Choi a ,1 , Kyuil Kwak a ,1 , Soyoung Yang a , Sesil Lim b , JongRoul Woo a,c ,
a
Energy Environment Policy and Technology, Graduate School of Energy and Environment (KU-KIST Green School), Korea
University, Seoul, 02841, South Korea
b
Department of Industrial Engineering, Inha University, Incheon, 22212, South Korea
c
Policy Department, Korea Institute of Science and Technology, 5, Hwarang-ro 14-gil, Seongbuk-gu, Seoul, 02792, Republic of Korea

article info a b s t r a c t

Article history: Electrification of the transport sector is expanding in both developing and developed
Received 6 May 2022 countries in response to air pollution and climate change. Indonesia, the world’s 10th
Received in revised form 11 August 2022 biggest CO2 emitter having the world’s third largest two-wheeler market, aims to deploy
Accepted 20 August 2022
2.1 million electric two-wheelers by 2025 and is developing electric vehicle policies. This
Available online 28 August 2022
study designs a consumers’ choice model for buying a scooter in Jakarta, Indonesia, and
JEL classification: simulates the effects of two supportive policies to expand electric two-wheelers (provid-
D12 ing subsidies and building charging infrastructure) using a discrete choice experiment.
O33 We calculate the budget required for these subsidies. The results show that people in
Q42 Jakarta highly prefer electric scooters and consider their environmental impacts. A 20%
R48 subsidy increases the estimated annual market share of new electric scooter sales to
Keywords: 10%. In addition to a 20% subsidy, building charging infrastructure for electric scooters
Policy instrument to the same level as conventional gas stations leads to estimated new sales of electric
Electric vehicle scooters reaching 21%. This study suggests the need to develop comprehensive policy
Electric scooter packages to successfully deploy electric two-wheelers in Jakarta and provides reference
Technology adoption data to improve the efficiency and effectiveness of related policies.
Choice experiment © 2022 Economic Society of Australia, Queensland. Published by Elsevier B.V. All rights
Discrete choice model reserved.

1. Introduction

With rising average global temperature and increasing greenhouse gas emissions, many countries have adopted the
Paris Agreement and implemented related policies at the national level (UNEP, 2020; WMO, 2020). The transportation
sector accounts for about a quarter of the world’s energy-related direct CO2 emissions. Therefore, as one way to respond
to climate change and decrease CO2 emissions, many governments are emphasizing the importance of the transport
sector and pursuing policies to deploy electric vehicles (EVs), such as battery electric vehicles (BEVs), as a substitute
for internal combustion engine vehicles (ICEVs). Technologies related to EVs, including battery and charging technologies,

✩ Acknowledgments: This work was supported by the KIST Institutional Program (Project No. 2V09130-21-P036). This work was supported by
Korea Environment Industry & Technology Institute (KEITI) through ‘‘Climate Change R&D Project for New Climate Regime’’, funded by Korea Ministry
of Environment (MOE) (2022003560009).
∗ Corresponding author at: Energy Environment Policy and Technology, Graduate School of Energy and Environment (KU-KIST Green School),
Korea University, Seoul, 02841, South Korea.
E-mail address: [email protected] (J. Woo).
1 These authors contributed equally to this work.

https://doi.org/10.1016/j.eap.2022.08.015
0313-5926/© 2022 Economic Society of Australia, Queensland. Published by Elsevier B.V. All rights reserved.
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

have developed along with the global EV trend, however, the distribution of BEVs remains in the initial phase and is
mainly occurring in the United States, Europe, and China (IEA, 2020).
There are policy instruments for extending BEVs, such as incentives for purchasing or using BEVs and disincentives
for ICEVs. They play a significant role in encouraging consumers who are not familiar with new technologies that
are considered unproven or who are reluctant to adopt BEVs (Egbue and Long, 2012; Hardman, 2019; Brückmann
and Bernauer, 2020; IEA, 2020). Therefore, several studies have been conducted to evaluate and develop policies for
accelerating BEV deployment. Research, for example, analyzing factors affecting BEV purchase decisions by potential
consumers or actual drivers, or designing policies that can contribute to EV market share at local and national levels,
has mainly been conducted in developed economies such as China, the United States, Norway, Sweden, and Switzerland,
which are actively pursuing BEV expansion (Bjerkan et al., 2016; Langbroek et al., 2016; Mersky et al., 2016; Zhang et al.,
2016; Egnér and Trosvik, 2018; Brückmann and Bernauer, 2020; Sheldon and Dua, 2020).
Sheldon and Dua (2020) analyze the impacts and cost-effectiveness of a BEV subsidy program in China. They predict
the BEV market share, assuming that the beneficiaries would be limited to low-income consumers and the budget would
be reduced. Zhang et al. (2016) and Bjerkan et al. (2016) examine the factors affecting BEV purchases and the effectiveness
of related policies in Norway. Zhang et al. (2016) analyze the impacts of BEV price and other incentives (bus lane access,
toll waivers, and expansion of charging stations) on BEV expansion and emphasize the importance of extending charging
infrastructure. Bjerkan et al. (2016) examine seven incentives for BEV use, including exemptions from purchase tax and
value-added tax, vehicle license fee reduction, and free parking, and show the importance of lowering the up-front price
of BEVs. In addition, Mersky et al. (2016) and Egnér and Trosvik (2018) analyze EV-related policies at the regional or
municipal levels in Norway and Sweden, respectively.
Developing countries are also considering introducing policies related to expanding EVs and have begun to respond
to climate change. As Indonesia is the 10th biggest CO2 emitter globally, it is essential for the country to reduce CO2
emissions in the transport sector, which accounts for approximately 28.4% of the total emissions (IEA Data and Statistics,
2021; World Bank Data, 2021). Indonesia has set targets for EVs and is preparing related policies, to cope with the increase
in oil consumption, soaring economic growth, vehicle ownership, and climate change. Also, Indonesia has the world’s
fourth largest population, with an annual GDP per capita growth rate of approximately 3.9% in 2019. Their final energy
consumption is increasing along with rapid economic growth (IEA Data and Statistics, 2021; World Bank Data, 2021).
In particular, the transport sector accounts for 45% of final energy consumption, and approximately 95% of the energy
consumed is from oil products. Increasing oil consumption causes excess domestic energy supply and trade deficits (EIA,
2015; IESR, 2020; IEA Data and Statistics, 2021). In terms of transportation, Indonesia has the third largest two-wheeler
market after China and India and is much bigger than the four-wheelers (Yuniarto et al., 2022). The number of motorcycles
in Indonesia increased at an average annual growth rate of 12.14% between 2000 and 2018 and reached 106 million in
2018 (Badan Pusat Statistik, 2021). In 2018, domestic sales of motorcycles were approximately 7 million units, and their
demand remains high. In Jakarta alone, approximately 0.86 million units of motorcycle were sold in 2018 (CEIC data,
2022). Consequently, the Indonesian government is focusing on the penetration of electric two-wheelers and has set a
goal of deploying 2.1 million electric two-wheelers by 2025 (Asosiasi Industry Sepeda Motor Indonesia Anon, 2021a). The
expansion of electric two-wheelers in Indonesia can bring positive effects such as preventing dependence on fossil fuels
and slowing down climate change (Prasetio et al., 2019). However, detailed policy instruments, such as specific incentives
to achieve EV goals, have not yet been proposed, therefore, it is crucial to develop relevant policies (IESR, 2019, 2020).
Although developing policies to achieve the EV penetration goal is critical in Indonesia, there are still only a few
relevant studies. Some studies merely survey the preference level or recognition for EV purchase, and there has been
no quantitative analysis of policy effects (Syamnur et al., 2019; Utami et al., 2020). Specifically, a study on the preference
of consumers in Indonesia when purchasing an electric scooter (Guerra, 2019), and a study that analyzed factors affecting
consumers’ purchase intention for electric vehicles (Wicaksono and Aprianingsih, 2021) were conducted previously.
However, there are no studies that analyze the effects of subsidies and infrastructure expansion quantitatively. Therefore,
this study can narrow the gap in the existing literature.
Representative policy instruments that the Indonesian government can consider expanding electric two-wheelers
include (1) providing subsidies for buying electric two-wheelers and (2) building charging infrastructure. Therefore,
this study identifies necessary attributes that affect the purchase of two-wheelers in Jakarta, Indonesia, and derives
marginal willingness-to-pay (MWTP) using a choice experiment survey and discrete choice model. Next, through ex-ante
simulation, we predict the effects of major policy alternatives (providing subsidies and building charging infrastructure
for electric two-wheelers) on the expansion of electric two-wheelers. The findings can help the Indonesian government
develop policies to achieve EV penetration goals as reference data.
This paper consists of the following sections: Section 2 details the methodology, Section 3 presents the main results
and policy-based simulation, and Section 4 presents the conclusions and policy implications of the study.

2. Methodology and data

2.1. Choice experiment survey

To analyze the consumer preference for electric two-wheelers in Jakarta, we conducted a choice experiment survey
on 504 men and women aged 20 to 59 years in Jakarta, the capital of Indonesia. The survey was implemented online
374
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Table 1
Descriptive statistics of the sample.
Response Percentage
Total 504 100%
Gender Man 252 50%
Woman 252 50%
20–29 144 28.6%
Age 30–39 136 27.0%
40–49 128 25.4%
50–59 96 19.0%
Education level Before high school graduation 71 14.1%
University and above 433 85.9%
Less than 9 million IDR (617 USD) 126 25.0%
Monthly income 9 million–15 million IDR (617–1,028 USD) 110 21.8%
15 million–30 million IDR (1,028–2,057 USD) 119 23.6%
More than 3 million IDR (2,057 USD) 78 15.5%
Jakarta Pusat (Central Jakarta) 138 27.4%
Jakarta Barat (West Jakarta) 73 14.5%
Region
Jakarta Selatan (South Jakarta) 146 29.0%
Jakarta Timur (East Jakarta) 109 21.6%
Jakarta Utara (North Jakarta) 38 7.5%

Note: 1 USD = 14,582 IDR (2020 average, source: OECD Statistics; https://stats.oecd.org/).

from September 12th to 23rd, 2020, by MACROMILL EMBRAIN, a professional survey company. The sample group, similar
to the population composition, was extracted using a purposive quota-sampling method based on gender and age. The
descriptive statistics of the sample are shown in Table 1.
This study uses a choice experiment survey to analyze the key attributes that affect consumers’ decision to buy two-
wheelers in Jakarta. The attributes are set based on the 125cc scooter, which has the highest demand in Indonesia, to help
consumers understand each of them and make their choices easy. The survey questionnaire included content about fuels
and operating systems (three levels), accessibility of charging stations and charge time (four levels), and driving range,
fuel costs, and price (three levels) (Table 2).

2.1.1. Fuel types & operating systems

Currently, the conventional scooters that use fossil fuels are widely distributed throughout Indonesia. Against this
backdrop, the Indonesian government is considering the implementation of subsidies for electric vehicles and is especially
trying to expand electric scooters.
The international society as a whole is shifting to electric vehicles for carbon neutrality. Indonesia also joined in when
it announced the electrification strategy of the transportation sector. For instance, the General Plan for Energy (2017)
announced plans to expand the number of electric two-wheelers to 2.1 million by 2025. Also, the Ministry of Industry in
Indonesia aims to produce 7,700,000 electric two-wheelers by 2025 and 13,002,000 by 2030.
Regarding charging stations, there are 122 of them in Indonesia as of May 2021, but most of them are installed in Java
and Bali islands. Thus, the Indonesian government is setting the objective of expanding the number of charging stations
as well (Mahalana et al., 2021).
With EVs emerging in the vehicle market, types of fuels and operating systems have become one of the top attributes
that consumers consider when purchasing vehicles. Existing studies have shown that consumers’ awareness and their
evaluation of the environmental value of EVs affect purchase decisions (Egbue and Long, 2012; Choi et al., 2018; Egnér
and Trosvik, 2018). Egbue and Long (2012) show that oil consumption reductions are the most attractive feature of EVs
to potential early adopters in the United States, and environmental impacts – especially fuel sources for electricity – can
be one of the potential obstacles to EV expansion.
In addition, Choi et al. (2018) point out that switching to an eco-friendly electricity generation mix in Korea improves
the relative environmental impacts of EVs and plays an important role in promoting their purchase. Also, there are research
results on the prospect of expansion or development of electricity generation methods from renewable energy sources
as renewable energy usage will expand in Indonesia (Rahman et al., 2021; Winarno et al., 2016). Thus, in this study,
electricity as fuel for electric two-wheelers is divided into electricity generated from fossil fuels and renewable energy to
determine how the environmental impacts of electric two-wheelers affect consumers’ choices for buying two-wheelers
in Jakarta.
Furthermore, many studies confirm that the density of charging stations, charging time, maximum driving distance,
fuel economy, and vehicle prices are valid attributes that consumers consider before buying vehicles. Thus, we also include
them in the choice experiment survey (Diamond, 2009; Axsen et al., 2010; Egbue and Long, 2012; Bakker and Trip, 2013;
Sierzchula et al., 2014; Zhang et al., 2016; Moon et al., 2018; Sheldon and Dua, 2020).
375
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Table 2
Attributes and attribute levels for the choice experiment.
Attributes Descriptions of the attributes and the attribute levels
1. Fuel types & operating systems Description 1) Gasoline scooter: Oil fuel (gasoline) is used in an internal
combustion engine. It requires filling up at a gas station to drive.
While driving, exhaust fumes (harmful substances) are emitted through
the exhaust pipe.
2) Electric scooter: Electricity as fuel is used to turn the electric motor.
The only way to drive is to charge with electricity. Compared to
gasoline scooters, they are relatively quiet, smooth, and easy to operate
and maintain.
Attribute levels (3) 1) Gasoline (fuel: gasoline)
2-1) Electric (fuel: electricity generated from fossil fuels)
2-2) Electric (fuel: electricity generated from renewable energy)
2. Accessibility of charging stations Description Given the number of conventional gas stations for gasoline scooters is
100%, the percentage of charging stations
Attribute levels (4) 1) 100%
2) 80%
3) 50%
4) 10%
3. Charging time Description Time required to charge a scooter
Attribute levels (4) 1) 5 min.
2) 20 min.
3) 150 min.
4) 240 min.
4. Drivable range Description Maximum driving distance (km) when charged once
Attribute levels (3) 1) 60 km
2) 100 km
3) 200 km
5. Fuel cost Description Fuel cost for 1 km driving
Attribute levels (3) 1) 60 IDR/km (0.0041 USD/km)
2) 95 IDR/km (0.0065 USD/km)
3) 130 IDR/km (0.0089 USD/km)
6. Price Description Purchasing price of a scooter (125 cc)
Attribute levels (3) 1) 17 million IDR (1,165 USD)
2) 20 million IDR (1,371 USD)
3) 23 million IDR (1,577 USD)

2.1.2. Accessibility of charging stations

Several studies show that the expansion of electric vehicle charging station infrastructure will accelerate the expansion
of electric vehicles (Brückmann and Bernauer, 2020; Hardman, 2019; Rietmann and Lieven, 2019; Coffman et al., 2017).
Thus, the factor, accessibility of charging station, was designed to be decreasing from 100%; to 80%, 50%, and 10%. Through
this attribute, it is possible to confirm the effect of expanding electric charging station infrastructure on consumers’
choices.
Many governments provide incentive policies for EV purchase and use, such as subsidies or tax exemptions to lower
EV prices, which play a significant role in promoting EV deployment and related market development (Hardman, 2019;
IEA, 2020). Building EV charging infrastructure, which is insufficient compared to conventional ICEVs, is recognized as a
necessary policy for consumers to decide on EV use (Bakker and Trip, 2013).

2.1.3. Charging time

According to previous studies, charging time and driving range are important when it comes to choosing electric
scooters. Thus, consumers have high marginal utility in purchasing electric vehicles (Liao et al., 2017; Jensen et al.,
2013; Dimitropoulos et al., 2013; Bockarjova et al., 2014). Electric scooters have different battery capacities and charging
technologies for each manufacturing brand. However, generally fast-charging takes around 2 h and regular(slow)-charging
takes around 4 h (Habibie and Sutopo, 2020) when charging an electric scooter with 5 kWh battery capacity. For the
gasoline charging, we set it as it takes around 5 min to charge. Therefore, in the case of charging time, the attributes were
designed in a wide range from 5 min, which is the time to charge the gasoline scooter at a conventional gasoline station
to 4 h (240 min) which is the time to charge the electric scooter at a fast-charging station. In this study, we included
the charging time from 5 min which is the gasoline charging time to 240 min which is a regular(slow)-charging time for
electric scooter in order to provide the scooter users the choices from a varied range.
376
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Fig. 1. Example of a choice experiment.

2.1.4. Drivable range

Likewise, the drivable range of electric scooter is set from 60 km which is the average driving range of electric scooters,
to 200 km which is the average driving range of gasoline scooters.

2.1.5. Fuel cost

In terms of fuel cost, Indonesia is the 40th lowest country in gasoline prices, with an average price of 10,510 IDR/per
liter (GlobalPatrolPrices, 2019). Calculating this into price per km, the average fuel cost of gasoline scooters is 175 IDR/km
(10,510 IDR/L÷60 km/L = 175 IDR/km). In case of electricity prices, Perusahaan Listrik Negara (PLN), an Indonesian power
company has suggested selling electricity for EV charging to third parties at 714 IDR (0.05 USD) per kWh, resulting in up
to 1300 IDR (0.08 USD) per kWh for consumers to purchase (Mahalana et al., 2021).
The fuel cost when using electric scooters with a battery capacity of 1.5kWh and a maximum drivable range of 85 km
is 91.8 IDR/km. The calculation is provided below.
1300 IDR/kWh × 1.5 kWh × 4 h = 7800 IDR
7800 IDR ÷ 85 km = 91.8 IDR/km
Thus, the levels of the fuel cost attribute were set as 60 IDR/km, 95 IDR/km, and 130 IDR/km.

2.1.6. Price
In the case of scooter prices, attributes were set based on the prices of the five scooters which are the most popular
in Indonesia. Among the most-sold scooters in Indonesia, the cheapest model was Honda Beat Street which costs 17.15
million IDR. The average price of the 5 most popular scooter models was 28 million IDR. We assumed that consumers
would be able to purchase electric scooters with subsidies, and set the price attribute as 17 million IDR, 20 million IDR,
and 23 million IDR.
Only 64 alternative cards that can be generated by combinations of attributes and their levels presented in Table 2
are derived using a fractional factorial design that ensures orthogonality between individual attributes. Thus, there are 16
alternative sets consisting of four alternative cards per set. All respondents are divided into four groups, and each group
is given four alternative sets. Respondents in each group choose the most preferred alternative card from each alternative
set (see the example in Fig. 1).

2.2. Model specifications

A mixed logit model is used to identify consumers’ preference for electric scooters in Jakarta. Basically, the mixed
logit model defines Unjt , the utility of individual n when choosing alternative j in a set of alternatives t, based on random
utility theory as shown in Eq. (1) (McFadden, 1986; Train, 2009). The utility Unjt consists of a deterministic part Vnjt
and a stochastic part εnjt , where Vnjt consists of vector xjt , representing the observable properties of alternative j in the
alternative set t, and estimated coefficient vector β .

Unjt = Vnjt + εnjt = β ′ xjt + εnjt (1)

As shown in Eq. (2), the probability that individual n chooses alternative j among an alternative set t is defined to be
equal to the probability that the utility of choosing alternative j is greater than the utility of choosing other alternatives.
377
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

The Eq. (3) can be derived from the assumption that the error term εnj follows an independent and identically distributed
type I extreme value distribution (Gumbel distribution) (McFadden, 1986).

Pnjt = Pr Unjt > Unmt , ∀m ̸ = j = Pr(εnmt − εnjt < Vnjt − Vnmt , ∀m ̸ = j)

( )
(2)
exp(Vnjt )
Pnjt = ∑ , j = 1, . . . , J (3)
m exp(Vnmt )
A basic multinomial logit model has the advantages of being easy to estimate and having a broader scope of application.
However, the ratio of choice probabilities between the two alternatives is constant, and the heterogeneity of individual
preferences is not considered based on the independence of irrelevant alternatives (IIA) assumption (Train, 2003). Latent
class logit and mixed logit models have been used to compensate for these constraints (Shin et al., 2014; Moon et al., 2017;
Woo et al., 2017). This study uses a mixed logit model that allows consideration of heterogeneity of individual preferences
and mitigates IIA constraints by assuming an estimated coefficient β as a random variable with a probability density
function f (β ). The estimated coefficient of the utility function represents individual preferences, and f (β ) is assumed to
follow a normal distribution with mean b and variance W . Thus, the choice probabilities of the mixed logit model can be
expressed like Eq. (4), identical to the integration of choice probabilities of multinomial logit model concerning β .
∫
exp(Vnjt )
Pnjt = ∑ f (β|b, W ) dβ (4)
m exp(Vnmt )
This study uses Bayesian estimation because the mixed logit model has no closed-form solution when estimating
parameters using the classical maximum likelihood estimation method (Allenby and Rossi, 1998; Train, 2001, 2003).
To compare preferences using estimates more easily, MWTP and relative importance (RI) are derived. MWTP indicates
the amount an individual is willing to pay for changing an attribute, which can be obtained by dividing the attribute
variable estimate by the price variable estimate, as shown in Eq. (5) (Train, 2009). In Eq. (5), xa and β̂a represent values
and estimated coefficients of any attribute a, not price, respectively, and xprice and β̂price represent values and estimated
coefficients of the price attribute, respectively.

∂ U /∂ xa β̂a
MWTPa = − =− (5)
∂ U /∂ xprice β̂price
RI indicates how much each attribute influences the individual’s choice. It can be derived using the part-worth of each
attribute, as shown in Eq. (6) (Train, 2009). The part-worth of any attribute a, including price, is calculated by multiplying
the estimated coefficient of the attribute a by the range of its level (maximum value minus minimum value).
part − w ortha
RIa = ∑ × 100 (6)
a part − w ortha

3. Analysis results

3.1. Estimation results and discussion

When asked whether respondents were willing to buy EVs in the future before entering the choice experiment survey,
87% of respondents gave positive responses. Utami et al. (2020) surveyed Indonesians about their intention to use electric
two-wheelers, confirming that approximately 82% of respondents were willing to buy them. In a survey in Surakarta,
Indonesia, Syamnur et al. (2019) found that 71% of respondents wanted to buy EVs. These results imply that people have
a high interest in EVs in Indonesia and introducing related policies can encourage them to use EVs.
The utility function for buying a scooter, as defined in this study, is shown in Eq. (7). The six attribute variables are
set to affect the utility. The fuel types and operating systems are imposed, as two dummy variables, and the baseline is
a gasoline scooter. The coefficient β , which represents the effects of each attribute variable on utility, is estimated using
Bayesian estimation. To be more specific, the last 5,000 out of 10,000 repetitive trials are used for estimation, and the
results are as shown in Table 3.
Unjt = βn,1 Djt ,electric_fossil_fuels + βn,2 Djt ,electric_renewable + βn,3 Xjt ,charging_facility (7)
+ βn,4 Djt ,charging_time + βn,5 Xjt ,range + βn,6 Xjt ,fuel_cost + βn,7 Xjt ,price + εnjt
Results reveal that fuel types and operating systems, accessibility of charging stations, charging time, drivable range,
fuel costs, and price significantly affect choices to buy a scooter for consumers in Jakarta. Among them, fuel types and
operating systems are relatively more important than other attributes when making decisions about purchasing a scooter,
with a RI of 37.7%. Regarding fuel types and operating systems, consumers in Jakarta generally prefer electric scooters,
which use electricity generated from fossil fuels or renewable energy to gasoline scooters. The median values of MWTP
for buying an electric scooter using electricity from fossil fuels and renewable energy instead of a gasoline scooter are
estimated at 1,527,148 IDR (104.73 USD) and 3,298,192 IDR (226.18 USD), respectively. In other words, consumers in
Jakarta generally prefer electric scooters more than gasoline scooters and electricity from renewable energy than from
378
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Table 3
Estimation results of the mixed logit model.
Attributes Marginal utility of attributes (β ) Median MWTP RI(%)
Mean (b) Variance(W)
Fuel types & Baseline(gasoline) – – – –
operating systems electric-fossil fuels (β1 ) 1.5862* 8.9532* 1,527,148 IDR 37.7
(0.1958) (1.3786) = 104 USD
electric-renewable energy (β2 ) 1.6287* 18.8767* 3,298,192 IDR
(0.2371) (2.9744) = 226 USD
Accessibility of charging stationsa (β3 ) 0.9703* 1.5840* 14,359 IDR/%p 12.3
(0.1486) (0.6684) = 0.98 USD/%p
Charging timeb (β4 ) −0.5184* 0.5815* −19,978 IDR/min 17.7
(0.0599) (0.0976) = −1.37 USD/min
Drivable rangec (β5 ) 0.6709* 0.5089* 21,672 IDR/km 11.9
(0.0723) (0.1049) = 1.48 USD/km
Fuel costd (β6 ) −0.7242* 1.0492* −39,150 IDR/(IDR/km) 7.8
(0.1924) (0.3294) = −2.68 USD/(IDR/km)
Pricee (β7 ) −1.1747* 4.0399* – 12.6
(0.1808) (1.2316)

Note: Posterior standard deviations are in parentheses.;* indicate significance at the 5% and 1% levels, respectively.
a
unit: %p.
b
unit: min.
c
unit: km.
d
unit: IDR/km.
e
million IDR.

fossil fuels as fuel for electric scooters and are willing to pay more. This implies that residents of Jakarta consider the
environmental impact of scooters to be a significant factor when purchasing them.
After fuel types and operating systems, charging time is the second most important attribute (RI 17.7%). In terms
of MWTP, respondents are willing to pay an additional 19,978 IDR (1.37 USD) to reduce one minute of charging time.
Additionally, price, accessibility of charging stations, and Drivable range are relatively important, in that order. It is
estimated that respondents are willing to pay an additional 14,359 IDR (0.98 USD) and 21,672 IDR (1.49 USD) to increase
accessibility of charging stations by 1%p and the maximum driving distance by 1 km, respectively. Considering the current
state of technologies, these attributes are considerably inferior to those of conventional gasoline scooters. Therefore, there
seems to be an urgent need to develop policies and technologies to make electric scooters a substitute for conventional
gasoline scooters to expand the Indonesian EV market.
Furthermore, the RI of fuel costs is the lowest at 7.8%. That may be because the cost of gasoline fuels is insignificant in
Jakarta, hence, consumers in Jakarta do not significantly consider it when buying a scooter. Unlike these results, however,
Utami et al. (2020) find that Indonesian people consider charging cost discounts more important than purchase incentives
and charging costs more necessary than purchase costs using a simple comparison of Likert scales about financial,
technological, and political attributes of electric two-wheelers such as maintenance costs, safety, and charging station
availability. This disagreement may have resulted from differences in the scope and methods of the research, specifically,
this study uses a choice experiment that considers trade-offs between attributes and a regional scope limited to the richest
city, Jakarta.
The result that fuel types and operating systems, charging time, price, accessibility of charging stations, drivable range,
and fuel costs significantly affect the purchase decision of a scooter by people in Jakarta shows that these attributes
should be considered when developing policies to deploy electric scooters. In particular, it is necessary to analyze how
the attributes can be designed so that the government can implement measures to adjust them using policy instruments,
that is, price, accessibility of charging stations, fuel costs, and power mix, and how these designs affect the expansion of
EVs.

3.2. Ex-ante simulation results and discussion

3.2.1. Simulation overview

Currently, the Indonesian government is implementing the electrification of the transportation sector as a part
of its carbon neutralization policy and is considering providing subsidies and supporting the expansion of charging
infrastructure. This study conducts simulations to quantify how the consumer acceptance of electric scooters in Jakarta
would change when implementing policies to expand electric scooters by providing subsidies for purchasing or building
charging infrastructure for electric scooters.
This study focuses on the simulation of the subsidy policy effect at the current point. Thus, it does not take into
consideration the decrease in prices due to the improved performance of EV technologies such as batteries over time. In
379
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Table 4
Specification of representative scooter for each scooter type.
Attributes Honda Beat Gesits Electric
Fuel types and operating systems Gasoline Electricity
Accessibility of charging stations 100% 10%
Charging time 10 min. 180 min.
Drivable range 400 km 70 km
Fuel costs 130 IDR/km 55 IDR/km
(0.0089 USD/km) (0.0038 USD/km)
Price 16.5 million IDR 25 million IDR
(1,131.53 USD) (1,174.44 USD)

Note: 1 USD = 14,582 IDR (2020 average, source: OECD Statistics; https://stats.oecd.org/).

Table 5
Price and charging infrastructure assumption for scenario analysis.
Policy instrument: Providing subsidies for buying electric scooters 0% 10% 20% 30% 40%
Price Gasoline scooter 16.5 16.5 16.5 16.5 16.5
*unit: million IDR (USD) (1,131.53) (1,131.53) (1,131.53) (1,131.53) (1,131.53)
Electric scooter 25 22.5 20 17.5 15
(1,714.44) (1,543) (1,371.55) (1,299.1) (1,028.67)
Policy instrument: Building charging stations for electric scooters
Subsidy Electric scooter 20% 20% 20% 20% 20%
Accessibility of charging Electric scooter 10% 30% 50% 70% 100%
stations

Note: 1 USD = 14,582 IDR (2020 average, source: OECD Statistics; https://stats.oecd.org/).

this study, the price of scooter is set at the current point, and the preference of consumers in Jakarta is analyzed when the
subsidy was provided from the minimum level to the maximum level. The results-driven from the simulation presented
in this study would also be able to anticipate how much effect can technology improvement bring on the expansion of
electric scooters as it shows the effect of the relative price drop on the expansion of electric scooters. Consumers’ choices
are simulated when each attribute level changes by policy enforcement, based on the consumers’ choice model estimated
in the previous section and the representatives of the scooter market in Jakarta, that is, the ‘‘Honda Beat’’ for gasoline
scooters and ‘‘Gesits Electric’’ for electric scooters (Table 4).
The first scenario assumes that the price of electric scooters is reduced to a level equal to that of gasoline scooters
by providing subsidies and that all other attributes except price remain unchanged (Table 5). Especially, in the case of
the charging station attribute, the level was set at 10% which is the current level of the infrastructure. The subsidy policy
can be inefficient when the price of EVs still remains too high for most consumers (Bakker and Trip, 2013). Therefore,
the scenario is designed to provide subsidies of up to 40% of the price of a standard electric scooter (Gesits Electric) in
Table 4. When providing a 40% subsidy, the price of an electric scooter drops to 15 million IDR, making it cheaper than a
gasoline scooter.
The second scenario assumes building the charging infrastructure for electric scooters to increase accessibility by 10%
to 100% (Table 5). Like scenario 1, other attributes were set unchanged except for the subsidy level. In scenario 2, the
subsidy level was set at 20% which can be considered medium in a range of 0% to 50%. Zhang et al. (2016) find that the
density of charging stations for BEVs has the greatest impact on BEV sales over toll waivers or bus lane access in Norway.
In addition, Brückmann and Bernauer (2020) show that non-EV drivers oppose keeping the current state of charging
infrastructure and want more charging infrastructure even though they do not benefit from it. Thus, the second scenario
aimed at predicting how the increase rate of choice probabilities change as the charging infrastructure is further expanded
in addition to the subsidy policy.

3.2.2. Simulation results

The results of the two simulations are shown in Figs. 2 and 3 to quantify the impacts of policies on Indonesian
consumers’ acceptability of electric scooters. Fig. 2 shows the result of the first scenario, a change in the choice
probabilities for consumers to choose an electric scooter (market share of new electric scooters sales) depending on
subsidy levels. The choice probability of electric scooters is only about 16% when they are cheaper than gasoline scooters,
which is an increase of only 10%p. This result shows that providing subsidies alone as an incentive policy to expand the
use of electric scooters has limited effects on promoting their purchase in Jakarta. Another possible explanation is that
people in Jakarta who responded to the survey in this study have a higher average salary than those in other regions.
The average salary in Indonesia in 2020 was 1,120 USD/month, whereas in Jakarta, it was 1,277 USD/month, which is 14%
higher (Destination Scanner Anon, 2021b). Some studies show that consumers with higher income have lower sensitivity
to BEV price, and those with lower income have a higher price elasticity of BEV demand, which can affect price reductions
380
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Table 6
Scenario 2: Simulation result.
Condition Subsidy 20% 20% 20% 20% 20%
Accessibility of 10% 30% 50% 70% 100%
charging stations
Result Choice probabilities (electricity 9.87% 11.78% 13.95% 16.45% 20.84%
from renewable resources)
Choice probabilities (electricity 9.57% 11.34% 13.45% 15.87% 20.15%
from fossil fuels)

Fig. 2. Simulation result of scenario 1. Note: ‘‘Subsidy’’ represents the ratio of subsidy provided in the proportion to the price of electric scooters.

considerably (Zhang et al., 2016; Sheldon and Dua, 2020). Therefore, the subsidies’ effects may be higher outside Jakarta,
even within Indonesia.
The second scenario, which is increasing the accessibility of charging stations for electric scooters from 10% to 100%
(the same level as conventional gas stations) along with subsidy level of 20% (given), leads to an increase in the choice
probability of electric scooters from about 9.87% (9.57% for electricity from fossil fuels) to 20.84% (20.15% for electricity
from fossil fuels) (Fig. 3). Consumers’ choice probability of electric scooters reaches only 9.87% (9.57% for electricity
from fossil fuel) when the accessibility of charging stations is at the 10% level with 20% subsidy. However, when the
accessibility of charging stations increases, the consumers’ choice probability also increases even with the same level of
subsidy (Table 6). Thus, it is shown that the infrastructure expansion policy plays a significant role in increasing consumers’
choice probabilities when it is applied along with the subsidy policy.
The result of the second simulation, similar to the first one, implies that both policies, subsidies and expansion
of charging infrastructure, have limitations in deploying electric scooters sufficiently for the complete substitution of
conventional gasoline scooters in Jakarta. According to a survey in Surakarta, Indonesia, Syamnur et al. (2019) show that
57%, 79%, and 83% of respondents reported that EVs should be comparable to ICEVs in price, performance, and charging
infrastructure, respectively. Therefore, it can be deduced that lower prices and expansion of charging infrastructure,
including a reduction in charging time, and longer drivable range, should be implemented through the development of
battery and charging technologies to achieve the complete transition to electric scooters in Indonesia.
In addition, changes in the choice probabilities when changing fuel types and operating systems are identified and
inferred as the most important attributes for respondents to buy a scooter according to the results of the choice
experiment survey. In Section 3.1, it is confirmed through mixed logit analysis that compared with conventional gasoline
scooters, the median value of MWTP for buying electric scooters powered by renewable energy is higher than that for
purchasing electric scooters powered by fossil fuels. However, the simulation results show that changing the power source
of electric scooters from fossil fuels to renewable energy does not significantly impact the choice probabilities of electric
scooters, on average. In conclusion, many consumers are willing to pay more for electric scooters powered by renewable
energy than by fossil fuels; however, concerning the overall two-wheeler market, they are unlikely to have a sufficiently
substantial impact to affect the market share of new electric scooter sales.
In addition to the market share estimation, the budget required for expanding electric scooters is estimated based on
the first simulation results, as shown in Table 7. According to the simulation, the Indonesian government should subsidize
approximately 40% of the price of electric scooters (10 million IDR per unit) to increase their new sales of electric scooters
381
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

Fig. 3. Simulation result of scenario 2. Note: Subsidy is set at the level of 20%, and ‘‘Subsidy’’ represents the ratio of subsidy provided in the
proportion to the price of electric scooters. ‘‘Accessibility’’ represents the ratio of electricity charging station in comparison to conventional gasoline
station, and the current number of conventional gasoline stations is given as 100% in this paper.

Table 7
Budget required for deployment of electric scooters through providing subsidies.
Level of subsidy for buying Subsidy 10% Subsidy 20% Subsidy 30% Subsidy 40%
electric scooters (2.5 million IDR per (5 million IDR per (7.5 million IDR per (10 million IDR per
unit) unit) unit) unit)
Annual sales of electric scooters* 62,866 82,302 106,898 137,600
(Market share of new sales (7.31%) (9.57%) (12.43%) (16.0%)
electric scooter)
Government expenditure 150 billion IDR 410 billion IDR 800 billion IDR 1,380 billion IDR
(10.28 million USD) (28.11 million USD) (54.86 million USD) (94.63 million USD)

Note: Based on the consumers’ preference in Jakarta determined from the estimation in this study.
1 USD = 14,582 IDR (2020 average, source: OECD Statistics; https://stats.oecd.org/)
* Annual motorcycle sales in Jakarta, 2018 are assumed at 0.86 million units (CEIC data, 2022).

to approximately 16%. Therefore, assuming that the annual sales volume of motorcycles is approximately 0.86 million units
(2019) in Jakarta, 16% of which – 137,600 units – are electric scooters. Therefore, the subsidy that the government must
provide is 1,380 billion IDR (1.04 million electric scooters × 10 million IDR per unit).
It is tedious to accurately calculate the cost of building charging stations because they vary depending on the region,
including the level of transmission and distribution infrastructure. Thus, this study does not determine the government
expenditures associated with the second scenario. Nevertheless, the result of the second simulation shows that allocating
the budget to build the charging infrastructure for electric scooters can significantly affect expanding electric scooters.

4. Conclusion and policy implications

This study identifies factors that significantly impact the purchase decision of scooters in Jakarta, Indonesia, and
analyzes how supportive policies (providing subsidies and building charging stations for electric scooters) can affect
consumers’ choices and the deployment of electric scooters. The Indonesian government plans to transform the transport
sector into an electric one in response to soaring economic growth, vehicle ownership, and climate change. Although
developing policies aim to achieve the transformation in the transport sector, detailed policy instruments have not been
presented, and related research is insufficient in aiding the development of policies. Therefore, this study suggests the
necessity and importance of comprehensive policy package development to achieve the goal of deploying electric two-
wheelers in Indonesia and is expected to contribute to developing related policies, providing reference data to improve
their efficiency and effectiveness.
Using a choice experiment survey and discrete choice models, the results of analyzing six associated attributes –
specifically, fuel types and operating systems, accessibility of charging stations, charging time, drivable range, fuel costs,
and price – reveal that all significantly affect consumers’ choice to buy scooters in Jakarta. Among them, fuel types and
operating systems are considered the most important attributes, with a high preference for electric scooters compared
to conventional gasoline scooters and a high MWTP. Although the impact on the choice probabilities of electric scooters
382
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

caused by changing their power sources (fossil fuels or renewable energy) is insignificant, the results of this study show
that consumers in Jakarta value the environmental impacts of vehicles, which can be expected to induce actual EV use
based on their preferences through introducing relevant policies.
Two simulations show that it is necessary to lower the price and expand charging infrastructure, reduce charging time,
and enable longer drivable range through the development of battery and charging technologies. When all attributes are
constant except for price, the subsidies, from 0% to 40% of the price of electric scooters, improve the choice probability of
electric scooters from 6% to 16%; the choice probability reaches 20.84% (20.15% for electricity from fossil fuels) when
providing a 20% subsidy and building charging infrastructure for electric scooters at the same level as gas stations.
Increasing accessibility of charging infrastructure can be a significant contribution to the expansion of electric scooters
since the consumers’ choice probability is only 9.87% (9.57% for electricity from the fossil fuels) at the 10% accessibility
with the same level of subsidy which is 20%. The subsidy and building charging infrastructure policy in Jakarta would
increase the new sales of electric scooter. However, it will take some time to replace the pre-existed stock of gasoline
scooters to electric scooters due to their usage life-expectancy. These results imply that supportive policies in which the
Indonesian government plans to expand EVs by providing subsidies and building charging infrastructure are insufficient to
induce consumers who have high preferences for electric scooters to recognize and use EVs as a substitute for conventional
gasoline vehicles.
Based on the first simulation providing 10% to 40% subsidies for the price of electric scooters, the budget required for
policy implementation is estimated to be in the range of 150 billion IDR (10.28 million USD) to 1380 billion IDR (94.63
million USD). Although it is difficult to estimate the budget needed to build charging infrastructure, as mentioned earlier,
this study suggests that policy packages are required to achieve the transformation of the transport sector in Jakarta,
Indonesia into an electric one, considering policy feasibility and various attributes of electric two-wheelers.
In addition, the development of comprehensive policy packages can contribute to the expansion of EVs, as per results
showing high preferences of consumers in Jakarta for electric scooters and significant impacts of electric scooters’
attributes on purchase choice. However, the two policies that the government plans are insufficient to increase accept-
ability for EVs. Other results showing the budget required for the government to provide subsidies are also beneficial
for considering the cost-effectiveness of policies. In conclusion, this study expects to improve policy effectiveness and
efficiency in Indonesia with an experiment targeted at Jakarta, which is about to develop related policies in an early stage
of the EV market by providing reference data.
Abbreviations
BEVs = battery electric vehicles
EVs = electric vehicles
ICEVs = internal combustion engine vehicles
IIA = independence from the irrelevant alternatives
MWTP = marginal willingness-to-pay
RI = relative importance

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have
appeared to influence the work reported in this paper.

References

Allenby, G.M., Rossi, P.E., 1998. Marketing models of consumer heterogeneity. J. Econometrics 89 (1–2), 57–78.
Anon, 2021a. Asosiasi industry sepeda motor Indonesia. https://www.aisi.or.id/statistic/, (Accessed 8 January 2021).
Anon, 2021b. Destination scanner. https://destinationscanner.com/average-salary-in-indonesia/, (Accessed 8 January 2021).
Axsen, J., Kurani, K.S., Burke, A., 2010. Are batteries ready for plug-in hybrid buyers? Transp. Policy 17 (3), 173–182.
Badan Pusat Statistik, 2021. https://www.bps.go.id/linkTableDinamis/view/id/1133, (Accessed 8 January 2021).
Bakker, S., Trip, J.J., 2013. Policy options to support the adoption of electric vehicles in the urban environment. Transp. Res. D 25, 18–23.
Bjerkan, K.Y., Nørbech, T.E., Nordtømme, M.E., 2016. Incentives for promoting battery electric vehicle (BEV) adoption in Norway. Transp. Res. D 43,
169–180.
Bockarjova, M., Rietveld, P., Knockaert, J., Steg, L., 2014. Dynamic consumer heterogeneity in electric vehicle adoption (no. 14-1579).
Brückmann, G., Bernauer, T., 2020. What drives public support for policies to enhance electric vehicle adoption? Environ. Res. Lett. 15 (9), 094002.
CEIC data, 2022. Indonesia Number of Motor Vehicles: Motorcycles: Jakarta. Retrieved August 10, 2022, from https://www.ceicdata.com/en/indonesia/
number-of-motor-vehicle-registered-motorcycles/no-of-motor-vehicles-motorcycles-jakarta.
Choi, H., Shin, J., Woo, J., 2018. Effect of electricity generation mix on battery electric vehicle adoption and its environmental impact. Energy Policy
121, 13–24.
Coffman, M., Bernstein, P., Wee, S., 2017. Electric vehicles revisited: a review of factors that affect adoption. Transp. Rev. 37 (1), 79–93.
Diamond, D., 2009. The impact of government incentives for hybrid-electric vehicles: Evidence from US states. Energy Policy 37 (3), 972–983.
Dimitropoulos, A., Rietveld, P., Van Ommeren, J.N., 2013. Consumer valuation of changes in driving range: A meta-analysis. Transp. Res. A 55, 27–45.
Egbue, O., Long, S., 2012. Barriers to widespread adoption of electric vehicles: An analysis of consumer attitudes and perceptions. Energy Policy 48,
717–729.
Egnér, F., Trosvik, L., 2018. Electric vehicle adoption in Sweden and the impact of local policy instruments. Energy Policy 121, 584–596.
EIA, 2015. International Energy Data and Analysis–Indonesia. Energy Information Administration (EIA), USA, https://www.eia.gov/international/content/
analysis/countries_long/Indonesia/indonesia.pdf.

383
S. Choi, K. Kwak, S. Yang et al. Economic Analysis and Policy 76 (2022) 373–384

GlobalPatrolPrices, 2019. https://www.globalpetrolprices.com, (Accessed 28 June 2020).

Guerra, E., 2019. Electric vehicles, air pollution, and the motorcycle city: A stated preference survey of consumers’ willingness to adopt electric
motorcycles in Solo, Indonesia. Transp. Res. D 68, 52–64.
Habibie, A., Sutopo, W., 2020. A literature review: commercialization study of electric motorcycle conversion in Indonesia. IOP Conference Series:
Materials Science and Engineering. Vol. 943, (1), IOP Publishing, 012048.
Hardman, S., 2019. Understanding the impact of reoccurring and non-financial incentives on plug-in electric vehicle adoption–a review. Transp. Res.
A 119, 1–14.
IEA, 2020. Global EV Outlook 2020: Entering the Decade of Electric Drive?. International Energy Agency (IEA), France.
IEA Data and Statistics, 2021. https://www.iea.org/data-and-statistics?country=WORLD&fuel=Energy%20supply&indicator=TPESbySource, (Accessed 8
January 2021).
IESR, 2019. Indonesia Clean Energy Outlook: Tracking Progress and Review of Clean Energy Development in Indonesia. Institute for Essential Services
Reform (IESR), Indonesia.
IESR, 2020. The Role of Electric Vehicles in Decarbonizing Indonesia’s Road Transport Sector. Institute for Essential Services Reform (IESR), Indonesia.
Jensen, A.F., Cherchi, E., Mabit, S.L., 2013. On the stability of preferences and attitudes before and after experiencing an electric vehicle. Transp. Res.
D 25, 24–32.
Langbroek, J.H., Franklin, J.P., Susilo, Y.O., 2016. The effect of policy incentives on electric vehicle adoption. Energy Policy 94, 94–103.
Liao, F., Molin, E., van Wee, B., 2017. Consumer preferences for electric vehicles: a literature review. Transp. Rev. 37 (3), 252–275.
Mahalana, A., Yang, Z., Posada, F., 2021. Indonesia transport electrification strategy (no. 2021-36).
McFadden, D., 1986. The choice theory approach to market research. Mark. Sci. 5 (4), 275–297.
Mersky, A.C., Sprei, F., Samaras, C., Qian, Z.S., 2016. Effectiveness of incentives on electric vehicle adoption in Norway. Transp. Res. D 46, 56–68.
Moon, H., Choi, H., Lee, J., Lee, K.S., 2017. Attitudes in Korea toward introducing smart policing technologies: Differences between the general public
and police officers. Sustainability 9 (10), 1921.
Moon, H., Park, S.Y., Jeong, C., Lee, J., 2018. Forecasting electricity demand of electric vehicles by analyzing consumers’ charging patterns. Transp.
Res. D 62, 64–79.
Prasetio, E.A., Belgiawan, P.F., Anggarini, L.T., Novizayanti, D., Nurfatiasari, S., 2019. Acceptance of electric vehicle in indonesia: Case study in bandung.
In: 2019 6th International Conference on Electric Vehicular Technology (ICEVT). IEEE, pp. 63–71.
Rahman, A., Dargusch, P., Wadley, D., 2021. The political economy of oil supply in Indonesia and the implications for renewable energy development.
Renew. Sustain. Energy Rev. 144, 111027.
Rietmann, N., Lieven, T., 2019. How policy measures succeeded to promote electric mobility–worldwide review and outlook. J. Clean. Prod. 206,
66–75.
Sheldon, T.L., Dua, R., 2020. Effectiveness of China’s plug-in electric vehicle subsidy. Energy Econ. 88, 104773.
Shin, J., Woo, J., Huh, S.Y., Lee, J., Jeong, G., 2014. Analyzing public preferences and increasing acceptability for the renewable portfolio standard in
Korea. Energy Econ. 42, 17–26.
Sierzchula, W., Bakker, S., Maat, K., Van Wee, B., 2014. The influence of financial incentives and other socio-economic factors on electric vehicle
adoption. Energy Policy 68, 183–194.
Syamnur, F.H., Pambudi, N.A., Biddinika, M.K., Wardani, N.S., 2019. Barriers to the adoption, acceptance and public perceptions of Electric Vehicles
(EV) in Indonesia: Case studies in the city of Surakarta. J. Phys.: Conf. Series 1402 (4), 044061.
Train, K., 2001. A Comparison of Hierarchical Bayes and Maximum Simulated Likelihood for Mixed Logit. University of California, Berkeley, pp. 1–13.
Train, K., 2003. Discrete Choice Methods with Simulation. Cambridge University Press, New York.
Train, K.E., 2009. Discrete Choice Methods with Simulation. Cambridge University Press.
UNEP, 2020. Emissions Gap Report 2020, United Nations Environment Programme (UNEP), Nairobi.
Utami, M.W.D., Yuniaristanto, Y., Sutopo, W., 2020. Adoption intention model of electric vehicle in Indonesia. J. Optimasi Sistem Industri 19 (1),
70–81.
Wicaksono, K.B., Aprianingsih, A., 2021. Electric car penetration potential in Indonesia. J. Penelitian Transp. Darat 23 (2), 142–149.
Winarno, O.T., Alwendra, Y., Mujiyanto, S., 2016. Policies and strategies for renewable energy development in Indonesia. In: 2016 IEEE International
Conference on Renewable Energy Research and Applications (ICRERA). IEEE, pp. 270–272.
WMO, 2020. WMO Statement on the State of the Global Climate in 2019. World Meteorological Organization (WMO).
Woo, J., Choi, H., Ahn, J., 2017. Well-to-wheel analysis of greenhouse gas emissions for electric vehicles based on electricity generation mix: A global
perspective. Transp. Res. D 51, 340–350.
World Bank Data, 2021. https://data.worldbank.org/, (Accessed 8 January 2021).
Yuniarto, M.N., Wiratno, S.E., Nugraha, Y.U., Sidharta, I., Nasruddin, A., 2022. Modeling, simulation, and validation of an electric scooter energy
consumption model: A case study of Indonesian electric scooter. IEEE Access 10, 48510–48522.
Zhang, Y., Qian, Z.S., Sprei, F., Li, B., 2016. The impact of car specifications, prices and incentives for battery electric vehicles in Norway: Choices of
heterogeneous consumers. Transp. Res. C 69, 386–401.

384