Factors Influencing Electric Motorcycle Adoption in Indonesia: Comprehensive Psychological, Situational, and Contextual Perspectives

Abstract

The adoption of electric motorcycles is critical for reducing transportation-related greenhouse gas emissions in Indonesia, which reached 674.54 million t of CO₂ in 2023. This study integrates the Theory of Planned Behavior with situational, contextual, and demographic factors to explore the determinants of electric motorcycle adoption intentions and actual usage. Data were collected from 1602 respondents across ten provinces with the highest motorcycle sales using purposive sampling and analyzed through Partial Least Squares—Structural Equation Modeling. Findings reveal that psychological factors—attitude, subjective norms, and perceived behavioral control—significantly influence purchase intentions, while personal moral norms do not. Situational factors such as technology and cost indirectly affect adoption intentions through attitude and perceived behavioral control. Contextual factors show mixed results; government policies effectively shape attitudes and perceived behavioral control, but infrastructure remains inadequate to influence attitudes directly. Demographic analysis highlights gender as a moderating factor, with men showing higher moral-driven adoption intentions. These results imply that the government and manufacturers need to develop the appropriate strategy to foster public interest in adopting electric motorcycles to increase the adoption rate of pro-environmental vehicles. Government policies such as purchase price subsidies, tax reductions, and charging rate discounts can motivate the intention to adopt electric motorcycles. In addition, manufacturers could improve technical performance and reduce the total cost of ownership, such as the purchase price and battery replacement costs.

1. Introduction

Climate change is a global issue affecting all living species and the environment, primarily caused by the transportation sector [1]. In 2023, transportation in Indonesia produced greenhouse gas emissions of 674.54 million t of CO₂ [2]. Predicting transportation emissions requires detailed traffic data [3]. One of the essential pieces of information needed is the number of vehicles. The increase in the number of vehicles highlights the urgency of transitioning to low-emission transportation. This transition is especially relevant for motorcycles, the most common vehicle in Indonesia, and it significantly contributes to transportation emissions.

Table 1. Number of Vehicles in Indonesia.

Vehicles	2020	2021	2022	2023
Car	15,797,746	16,413,348	17,168,862	18,285,293
Bus	233,261	237,566	243,450	269,710
Truck	5,083,405	5,299,361	5,544,173	6,091,822
Motorcycle	115,023,039	120,042,298	125,305,332	132,433,679
Total	136,137,451	141,992,573	148,261,817	157,080,504

shows that motorcycles dominate the vehicle category in Indonesia [4]. The number of motorcycles has consistently increased yearly, especially from 2020 to 2023. In 2020, the number of motorcycles was recorded at 115,023,039 units, which then increased to 120,042,298 units in 2021. The increase continued in 2022, with the number of motorcycles reaching 125,305,332 and finally 132,433,679 units in 2023. This trend indicates a steady annual growth, with an average increase of over 5 million motorcycles annually. These figures suggest that motorcycles remain the primary choice for the public due to their accessibility, efficiency, and economic appeal to various user groups. However, this significant growth also challenges environmental management and the need to accelerate the adoption of low-emission vehicles.

Mitigating emissions from the transportation sector, the Indonesian government has issued a series of regulations to support accelerating the adoption of electric motorcycles (EMs). Presidential Regulation No. 79 of 2023, which revises Presidential Regulation No. 55 of 2019, governs the acceleration of the battery-based electric vehicle (EV) program [5]. Supporting the implementation of the EV program, the Minister of Energy and Mineral Resources (ESDM) Regulation No. 1 of 2023 regulates the provision of charging facilities for battery-based EVs [6]. As a further step to encourage the adoption of EMs, the government, through the Minister of Industry Regulation No. 21 of 2023, provides incentives for purchasing new battery-based EMs. This incentive is a subsidy that can reduce the upfront cost for consumers and accelerate the transition to EVs [7]. Lastly, the Minister of Environment and Forestry Regulation No. 8 of 2023 establishes emission quality standards for vehicles, setting emission standards to ensure that EVs are energy-efficient and play a role in improving air quality and promoting pollution reduction [8]. These regulations collaborate to accelerate the transition to EVs, create an environmentally friendly ecosystem, and reduce Indonesia’s dependency on fossil fuel-powered vehicles. Opting for EVs instead of fossil fuel-powered vehicles constitutes a substantial advancement in environmental conservation [9].

Studies on the intention to adopt EVs in various countries have different and limited policy implications. Several studies mention that factors influencing the purchase of EVs include vehicle price [10,11], infrastructure [11,12,13], government policies [1], and environmental awareness [9,14,15]. Some studies highlight that factors affecting the adoption intention of EM include sociodemographic, financial, technological, and macro-level factors [16]. Other studies expand the analysis by adding demographic, situational, contextual, and psychological factors [17].

One of the technology acceptance theories that many researchers use to study EV adoption is the Theory of Planned Behavior (TPB). TPB suggests that an individual’s intention primarily depends on three determining factors: attitude, subjective norm, and perceived behavioral control. These determinants rely on a base thought structure: behavioral, normative, and control beliefs [14]. Previous studies on the application of the TPB model to the adoption intention of EVs have been widely conducted in various countries, such as China [18], Turkey [15], India [19,20], and Spain [21].

Previous research utilizing the TPB framework has revealed various factors influencing individuals’ willingness to adopt EM [17]. Earlier research also included additional constructs such as cost, group behavior, and personal norms [19]. A study in India validated the impact of actual usage on consumers’ intention to adopt EVs [20].

Several future research directions are recommended, including developing new conceptual models or integrating existing theories to explore the factors influencing EV adoption [22]. Research should address contradictory findings, as further investigation is needed given the conflicting results of previous studies [23]. Psychological factors are more intricate than demographic, situational, and contextual factors. At the same time, technological variables continue to evolve with advancements in technology, which may pose challenges for the growth of EV research. Therefore, understanding the impact of the evolving attributes of EV technology is crucial [24].

Previous research has extensively discussed psychological factors separately but has not comprehensively examined them. However, we did not find any literature that examines the integration of psychological and external factors in consumer behavior, particularly regarding adopting EM. Rahmawati’s study [17] explored psychological factors as independent variables but did not consider them from the consumer behavior theory perspective, as outlined by Kotler and Keller [25], where psychological factors can mediate between external factors and consumer decisions. Furthermore, external factors such as technology, cost, infrastructure, and policy can function as independent variables influencing consumer behavior, with psychological factors acting as a link that strengthens or modifies their impact.

This study takes motorcycles as an object because their numbers are increasing by around 5% annually [4]. This increase contributes to carbon emissions, which reached 674.54 million t of CO₂ in 2023 [2]. This condition also worsens air quality and negatively impacts the environment, which becomes an ongoing issue that needs to be addressed, especially considering its effects on public health and climate change. Considering the gaps in the literature and the current issues, this study aims to develop a comprehensive model to understand the adoption of EM in Indonesia. This study integrates the TPB with other relevant factors, including technology, cost, infrastructure, and policies relevant to the local context. This model aims to elucidate the key elements that affect the decision to adopt EM in Indonesia, thereby contributing to formulating more effective policies and supporting the use of EM.

There are three contributions to this research. First, we develop a technology acceptance model using the TPB to explain the intention to purchase electric motorcycles (EMs). Second, this study has implications for policy-makers in designing more effective regulations and incentives to promote the use of EMs as a solution to reduce air pollution caused by emissions from motor vehicles. Third, this research provides a broader understanding of how psychological factors play a role in strengthening or modifying the influence of external factors on the adoption of Ems, which the automotive industry and marketers can use to develop more effective marketing strategies.

2. Theoretical Framework and Hypothesis Development

2.1. Theory of Planned Behaviour (TPB)

The Theory of Planned Behavior (TPB), introduced by Ajzen in the late 20th century, is an extension of the Theory of Reasoned Action (TRA) [26]. TPB explains the factors that influence a person’s intention to behave in a certain way. TPB is particularly effective in explaining consumer actions and intentions [27]; the technology adoption model, from a psychological perspective, influences people’s intentions to purchase new technology and helps predict users’ intentions to adopt technological acceptance [28]. TPB states that an individual’s intention to perform a behavior is influenced by three main factors: attitude, subjective norms, and perceived behavioral control. Ajzen [29] suggested that moral norms could serve as an additional element to enrich the understanding of the TPB model. This theory effectively describes consumer characteristics, including in the context of technology or new product adoption [23].

In this study, TPB is integrated with the broader consumer behavior theory, as explained in Marketing Management by Kotler and Keller [25], to enrich the understanding of external factors that also play a role in adoption decisions. The Consumer Behavior Theory emphasizes that consumer decisions are influenced by the value systems and beliefs that underpin their attitudes, as well as how external factors such as technology, costs, infrastructure, and policies can impact consumer decision-making. Technological advancements in particular can modify consumer intentions and behavior in response to market changes or new products [30]. By combining these two theories, this study aims to explore how internal and external factors interact to influence the intention of consumers to adopt electric motorcycles.

2.2. Theoretical Framework

This conceptual model is based on the Theory of Planned Behavior (TPB), expanded by incorporating relevant variables in the context of EV adoption. These variables include attitude (AT), subjective norm (SN), perceived behavioral control (PBC), personal moral norm (PMN), technology (TE), costs (CO), infrastructure (IN), and policy (PO), which are organized based on the results of a literature review related to consumer decisions to switch to EV. Unlike previous studies focusing on EVs, this research emphasizes two-wheeled vehicles, which play a significant role in transportation in developing countries like Indonesia. Figure 1 presents the conceptual model, illustrating how psychological factors such as AT, SN, PBC, and PMN serve as mediating variables. This study also adopts actual adoption (AA), a topic that has not been widely explored and uses factors such as TE, CO, IN, and PO as indirect variables that influence intentions to purchase EM.

2.3. Development of Hypotheses

2.3.1. Attitude (AT)

Attitude refers to an individual’s favorable or unfavorable judgment of a conduct, referring to how strongly an individual has a positive or negative evaluation of a behavior [26]. AT is described as an individual’s psychological emotion, which can be measured based on feelings [18]. AT also plays a role in forming the desire to use an EM, particularly in the decision-making process for purchase [31,32,33], as well as in reducing fuel consumption [31,33], environmental friendliness [31], and addressing climate change [32]. Previous studies have demonstrated that AT positively influences the intention to adopt EVs, as observed in Taiwan [34], and consumer optimism in Pakistan about EVs makes them more likely to be purchased [27]. Based on the previous explanation, it might be inferred that AT considerably impacts the decision to switch to EMs. Therefore, this study proposes the following hypothesis:

Hypothesis 1.

PI is positively influenced by AT.

2.3.2. Subjective Norm (SN)

A subjective norm pertains to how an individual perceives the social expectations or pressures from others, indicating how much they feel pressured by their social environment to engage in or avoid a specific behavior [26]. This social pressure can influence an individual to act or refrain from acting according to the expectations or views of those around them [18]. Subjective norms also play a crucial part in the decision to switch to EMs, including the influence of close individuals [31], social pressure [32], and the influence of people around [33]. Previous research has shown that SN significantly affects the decision to switch to EVs in Taiwan [34]. Subjective norms have also been recognized as the primary driver behind the development of the intention to adopt electric cars in India [19]. Based on the earlier explanation, it can be inferred that SN may impact the decision to transition to EM. As a result, this study presents the following hypothesis:

Hypothesis 2.

PI is positively influenced by SN.

2.3.3. Perceived Behavioral Control (PBC)

Perceived behavioral control is a person’s conviction in their capacity to regulate or perform a given activity. The perception of behavioral control indicates the extent to which a person feels they will encounter difficulty performing a particular action [26]. This indicator measures whether a person can easily consume a product [18]. Previous research has shown that the perception of behavioral control positively influences behavioral intention in Taiwan [34]. Perceived behavioral control is a stimulus for intention only when someone feels competent to drive an EV in India [19]. Several important factors include purchasing decisions related to battery warranties and accommodation of travel needs [31,32], maintenance and repair of EM [31], purchasing decisions [35], and price knowledge control [31]. Based on the previous explanation, PBC is believed to affect the decision to switch to EM. Therefore, the study proposes the following hypothesis:

Hypothesis 3.

PI is positively influenced by PBC.

2.3.4. Personal Moral Norm (PMN)

Ajzen [26] argued that moral norms can be useful to TPB. Moral norms are individuals’ perception of personal responsibility to perform or refrain from a particular behavior [29]. According to Schwartz [36], PMNs are crucial in motivating individuals to engage in behaviors that benefit others or the environment, even without external social pressure. Personal moral norms play an aspect in the intention to embrace EM, including personal responsibility for reducing carbon emissions and enhancing air quality and personal moral commitment [32]. The consumer’s sense of responsibility for the environmental impact of using vehicles plays a significant role when making adoption decisions [31]. Considering the previous explanation, it can be inferred that PMN affects the decision to switch to EM. As a result, this study presents the following hypothesis:

Hypothesis 4.

PI is positively influenced by PMN.

2.3.5. Technology (TE)

It is crucial to understand the continuous technological advancements in EVs [24]. A study in Taiwan [34] discovered that vehicle performance, including safety, reliability, and range, impacts consumer perceptions of behavioral control. Regarding PBC, this encompasses the perception of TE, which affects the ability to adopt and operate EVs. The greater the consumer’s ability to manage these factors, the more likely their intention to adopt such behavior [37]. Therefore, this study proposes the following hypotheses:

Hypothesis 5.

PI is positively influenced by TE through AT.

Hypothesis 6.

PI is positively influenced by TE through PBC.

2.3.6. Cost (CO)

Costs include various aspects such as the purchase price [38], battery replacement costs [39], refueling costs [40], and maintenance costs [41]. The purchase price is a key element that shapes consumer attitudes and perceptions of behavioral control [34]. The greater the consumer’s ability to control factors such as the cost of using and operating an EV, the higher the likelihood of their decision to switch to an EV [37]. Based on the previous explanation, the following hypotheses are proposed in this study:

Hypothesis 7.

PI is positively influenced by cost through AT.

Hypothesis 8.

PI is positively influenced by cost through PBC.

2.3.7. Infrastructure (IN)

Charging infrastructure is generally considered a key obstacle to EVs [42]. The availability of charging facilities is important in alleviating consumer concerns about EV range [34]. Based on the previous explanation, this study proposes the following hypotheses:

Hypothesis 9.

PI is positively influenced by IN through AT.

Hypothesis 10.

PI is positively influenced by IN through PBC.

2.3.8. Policy (PO)

Financial incentives and preferential policies are the most important and relevant factors shaping individual attitudes and demonstrating significant results [18]. Government policies are key in guiding consumers’ intention to use EVs [34]. Based on the previous explanation, the following hypotheses are proposed in this study:

Hypothesis 11.

PI is positively influenced by PO through AT.

Hypothesis 12.

PI is positively influenced by PO through SN.

Hypothesis 13.

PI is positively influenced by PO through PBC.

2.3.9. Purchase Intention (PI) and Actual Adoption (AA)

Consumer intention and attitude toward purchasing have been widely researched and explained in various behavior models, such as the Theory of Reasoned Action (TRA), the Theory of Planned Behavior (TPB), as well as the Technology Acceptance Model (TAM). Based on this, the following hypothesis is proposed in this study:

Hypothesis 14.

PI has a positive influence on AA.

2.3.10. Moderating Effects of Demographics

The association between gender and interest in purchasing EVs has been widely studied, but the results from various studies worldwide tend to be inconsistent. Research in the UK [42] demonstrated that women were more likely to buy EVs. A survey conducted in Russia [43] also found that women were more inclined to purchase EVs than men. Including gender as a moderating element in this research provides a deeper understanding of consumers’ intention to purchase EMs. Several previous studies also used demographic data such as age, gender, and income to analyze factors affecting the intention to adopt EVs, such as in Norway [44].

Hypothesis 15.

Gender roles as a moderator in the relationship between psychological factors and the PI.

3. Methodology

3.1. Item Measurements

Item measurement refers to defining and evaluating specific variables in a study. This research used 10 constructs with 41 measurement items to analyze the factors influencing the adoption of electric motorcycles. Each item is adapted from previous studies and modified to align with the objectives of this research. The details of the constructs and their measurement items are presented in

Table 2. Item Measurement.

Construct	ID	Measurement Items	References
Attitude	AT1	I think buying an electric motorcycle is a good decision.	[31,32,33]
	AT2	In the long term, I think using an electric motorcycle is more cost-effective than a conventional motorcycle.	[32]
	AT3	I think using an electric motorcycle to reduce fuel consumption is very important.	[31,33]
	AT4	I like electric motorcycles because they are environmentally friendly.	[31,33]
	AT5	Using an electric motorcycle will reduce climate change.	[32]
	AT6	I support the idea that this country should implement more policies to encourage people to buy electric motorcycles.	[33]
Subjective Norm	SN1	My close ones think that it is important for me to use an environmentally friendly vehicle.	[31,32]
	SN2	I feel social pressure to buy an environmentally friendly electric vehicle (electric motorcycle).	[32]
	SN3	I think if I buy an electric motorcycle, many people close to me will also be interested in buying one.	[31]
	SN4	If people around me buy an electric motorcycle, it will also encourage me to buy one.	[33]
Perceived Behavioral Control	PBC1	With a good converter kit and battery warranty, I would not worry about adopting an electric motorcycle (1-year battery warranty).	[32]
	PBC2	The electric motorcycle will accommodate my travel needs, even with the limited battery range.	[32]
	PBC3	The maintenance and repair of the electric motorcycle greatly influences my purchase decision.	[31]
	PBC4	I have full control over my decision to purchase an electric motorcycle.	[35]
	PBC5	I think the price of the electric motorcycle is important to me, and I can afford it when I decide to adopt it.	[31]
	PBC6	I will have the ability to buy an electric motorcycle in the future.	[33]
	PBC7	If I want to buy an electric motorcycle, I can purchase it without difficulty.	[35]
Personal Moral Norm	PMN1	I will use an electric motorcycle to reduce carbon emissions and improve air quality.	[32]
	PMN2	I feel a moral commitment to using an electric motorcycle	[32]
	PMN3	I feel responsible for the environmental impact of vehicle use when making adoption decisions.	[31]
Technology	TE1	The maximum range that can be achieved by the electric motorcycle is in line with the standards I want for daily activities.	[30,45,46]
	TE2	The maximum speed that can be achieved by the electric motorcycle meets the standards I want.	[41]
	TE3	The time to fully charge the electric motorcycle meets the standards I want.	[41,46]
	TE4	I feel safe when riding the electric motorcycle, even though the motor is quiet.	[30,47,48]
	TE5	The battery life (after 3 years of use, replacement is necessary) on the electric motorcycle meets the standards I want.	[49]
Cost	CO1	The purchase price of the electric motorcycle is within the budget I set for buying a motorcycle.	[38,50]
	CO2	The cost of replacing the electric motorcycle battery every 3 years is within the budget I set for motorcycle maintenance.	[39,50]
	CO3	The cost of charging an electric motorcycle is cheaper than the cost of gasoline, which makes me want to use an electric motorcycle.	[40,48]
	CO4	The maintenance cost of the electric motorcycle is relatively cheaper than that of a conventional motorcycle because the components in the electric motorcycle are much simpler than those in a conventional motorcycle.	[41,47]
Infrastructure	IN1	The availability of power sources/charging stations in public places that meet battery charging standards makes me want to use an electric motorcycle.	[30,51,52,53]
	IN2	The availability of power sources/charging stations at work that meet battery charging standards makes me want to use an electric motorcycle.	[30,53]
	IN3	The availability of power sources/charging stations at home that meet battery charging standards makes me want to use an electric motorcycle.	[52]
	IN4	The availability of service centers for routine maintenance of common damages makes me want to use an electric motorcycle.	[51]
Policy	PO1	Government subsidies for electric motorcycles make me want to use an electric motorcycle.	[30,50]
	PO2	Government-provided annual tax discounts for electric motorcycles make me want to use an electric motorcycle.	[30,50]
	PO3	Government-provided discounts on public charging fees make me want to use an electric motorcycle.	[30]
Purchase Intention	PI1	I am interested in purchasing an electric motorcycle.	[30]
	PI2	I want to recommend electric motorcycles to others.	[30]
Actual Adoption	AA1	Technological innovation will have a greater impact on my daily life.	[31]
	AA2	I believe that using an electric motorcycle will make my life easier.	[31]
	AA3	I am excited to learn how to use an electric motorcycle.	[31]

.

3.2. Data Collection

The data in this research were acquired by distributing online questionnaires using Facebook Advertisement to respondents who met specific criteria. The study employed a non-probability sampling method, specifically purposive sampling, to ensure that the respondents possessed characteristics aligned with the study’s requirements. The respondents were required to be over 17 years old, possess a motorcycle driving license, and domicile in 10 provinces with the highest motorcycle sales in Indonesia: West Java, East Java, Jakarta, Central Java, North Sumatra, West Sumatra, Yogyakarta, South Sulawesi, South Sumatra, and Bali. The sample size obtained in this study was 1602 respondents. Based on the Pareto principle, about 20% of the total provinces in Indonesia, namely 10 provinces with a total of 95,227,003 motorcycles in 2023 [4], account for almost 80% of the total motorcycle ownership in Indonesia. Therefore, this sample can represent Indonesia’s overall motorcycle ownership conditions.

3.3. Data Analysis Method

The data in this study were analyzed using the PLS-SEM method, which allows researchers to examine relationships between latent variables more comprehensively. PLS-SEM is a useful approach for estimating structural models in research. The complexity level of the PLS-SEM model can handle very complex models with hundreds of observed variables and rarely encounters convergence issues [54]. This study also aimed to test the model, focusing on prediction studies, exploration, and the development of structural model theory. Data analysis was performed using SmartPLS version 4 software, which was executed on a system with the following specifications: an AMD A9-9400 RADEON R5 processor with two physical cores and three integrated graphics cores (2400 MHz), 4 GB of RAM, and Microsoft Windows 10 Pro as the operating system. PLS-SEM allows data analysis with more flexible distributional assumptions. Construct validity and reliability were tested using Composite Reliability (CR) and Average Variance Extracted (AVE), while hypothesis testing was performed using the bootstrapping method. Bootstrapping is a resampling method that randomly selects samples from data to repeatedly estimate the path model with varying data, used to assess the significance of path coefficients. In this analysis, we used a significance level of 0.05 and performed 5000 iterations to ensure the results are robust and reliable [54].

4. Results

4.1. Descriptive Results

Table 3. Respondent demographics data.

Demographics	Attributes	Frequency	Percentage
Age	17–30	881	55.0
	31–45	635	39.6
	46–60	81	5.1
	>60	5	0.3
Gender	Male	825	51.5
	Female	777	48.5
Marital Status	Single	508	31.7
	Married	1085	67.7
	Others	9	0.6
Education	High School	923	57.6
	Undergraduate	594	37.1
	Graduate	85	5.3
Occupation	Student	231	14.4
	Civil Servant	90	5.6
	Private Employee	490	30.6
	Entrepreneur	505	31.5
	Others	286	17.9
Monthly Income	<IDR 2,000,000	503	31.4
	IDR 2,000,000–5,999,999	730	45.6
	IDR 6,000,000–9,999,999	257	16.0
	IDR 10,000,000–19,999,999	82	5.1
	>IDR 20,000,000	30	1.9
Domicile	West Java	461	28.8
	Central Java	308	19.2
	DKI Jakarta	262	16.4
	East Java	221	13.8
	North Sumatra	91	5.7
	DI Yogyakarta	73	4.6
	South Sumatra	63	3.9
	South Sulawesi	47	2.9
	Bali	42	2.6
	West Sumatra	34	2.1

presents the demographic profile of the respondents who are part of the sample in this study. Most respondents are between the ages of 17 and 30 years, accounting for 55%, and are predominantly male, making up approximately 51.5%. Most respondents, with 67.7%, are married and have completed high school education, comprising around 57.6%. Regarding occupation, the majority are entrepreneurs, representing 31.5%, followed by private sector employees at 30.6%. The most significant proportion of monthly income ranges within IDR 2,000,000–5,999,999 at about 45.6%. Most respondents are from West Java, accounting for 28.8%, Central Java at 19.2%, and Jakarta at 16.4%. This profile reflects the diversity of respondents both demographically and economically.

4.2. Measurement Model Assessment

4.2.1. Convergent Validity and Reliability

Convergent validity and reliability are evaluated by examining the value of each indicator. An outer loading value greater than 0.70 indicates that the construct explains more than 50% of the variance of its indicators, which suggests adequate indicator reliability. Internal consistency reliability is measured using Composite Reliability (CR), which reflects the consistency or internal reliability of the construct. The CR value above the minimum threshold of 0.70 indicates that the construct is consistent or reliable in measuring the variable.

Convergent validity shows how well a construct explains the variance of its indicators. It is assessed using Average Variance Extracted (AVE), where an acceptable AVE value is at least 0.50. This result means that the construct accounts for at least 50% of the variation in its indicators [54].

As shown in

Table 4. The constructs’ reliability and validity results.

Construct	Items	Outer Loadings	Composite Reliability (CR)	Average Variance Extracted (AVE)
Attitude (AT)	AT1	0.815	0.924	0.717
	AT2	0.847
	AT3	0.860
	AT4	0.852
	AT5	0.847
	AT6	0.857
Subjective Norm (SN)	SN1	0.800	0.854	0.658
	SN2	0.700
	SN3	0.874
	SN4	0.859
Perceived Behavioral Control (PBC)	PBC1	0.802	0.908	0.643
	PBC2	0.763
	PBC3	0.799
	PBC4	0.818
	PBC5	0.836
	PBC6	0.813
	PBC7	0.780
Personal Moral Norm (PMN)	PMN1	0.891	0.889	0.817
	PMN2	0.913
	PMN3	0.908
Technology (TE)	TE1	0.788	0.844	0.616
	TE2	0.790
	TE3	0.794
	TE4	0.767
	TE5	0.784
Cost (CO)	CO1	0.797	0.800	0.619
	CO2	0.770
	CO3	0.792
	CO4	0.787
Infrastructure (IN)	IN1	0.882	0.901	0.772
	IN2	0.899
	IN3	0.883
	IN4	0.849
Policy (PO)	PO1	0.904	0.902	0.836
	PO2	0.923
	PO3	0.916
Purchase Intention (PI)	PI1	0.932	0.850	0.869
	PI2	0.933
Actual Adoption (AA)	AA1	0.880	0.881	0.802
	AA2	0.901
	AA3	0.906

, the analysis results show that the outer loading values for all items are above 0.70, indicating a significant contribution of the indicators to their constructs. All constructs also have CR values above 0.70, meeting the recommended minimum threshold and indicating good reliability for each construct. All constructs’ AVE values are above 0.50, indicating that each construct’s indicators account for more than 50% of its variance. This result suggests substantial dependability and construct validity.

4.2.2. Discriminant Validity

Discriminant validity assesses the point at which a construct is empirically different from other components in the study model. Discriminant validity is measured using the Heterotrait–Monotrait Ratio (HTMT). For adequate discriminant validity, the recommended HTMT value is below 0.90 [54].

As shown in

Table 5. Discriminant validity using HTMT.

	AA	AT	CO	IN	PBC	PI	PMN	PO	SN	TE
AA
AT	0.845
CO	0.615	0.597
IN	0.628	0.598	0.683
PBC	0.880	0.837	0.685	0.671
PI	0.762	0.779	0.655	0.681	0.772
PMN	0.890	0.865	0.585	0.586	0.854	0.729
PO	0.658	0.691	0.569	0.727	0.683	0.640	0.668
SN	0.797	0.754	0.655	0.646	0.810	0.742	0.768	0.578
TE	0.685	0.683	0.840	0.763	0.754	0.721	0.663	0.655	0.736

, the results of the discriminant validity analysis using the HTMT approach show that all HTMT values between constructs are below the recommended threshold of 0.90. This result indicates that each construct in this study has adequate discriminant validity and can be clearly distinguished from other constructs. These findings confirm the discriminant validity of the constructs examined in the research model.

4.3. Structural Model Evaluation

Hypothesis testing in this study was conducted using PLS-SEM with a bootstrapping procedure, where a t-statistic value above 1.96 or a p-value below 0.05 indicates a significant effect between variables [55]. In addition, PLS-SEM was chosen because this method does not require the assumption of normal distribution [56]. The model’s ability to present the variance of the dependent variables is measured using R squared (R²). The higher the R squared value, the better the model’s ability to predict the fit between the observed data and the estimated model, focusing on the direct difference between the model results and actual data. An SRMR value ≤ 0.08 indicates that the model has a good fit [55]. The NFI (Normed Fit Index) is an incremental fit index developed to assess how well the model fits the data. The NFI value ranges from zero to one, with values closer to one indicating a better model fit [57].

The bootstrap analysis results in

Table 6. Results of bootstrap analysis.

Path	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values
AT -> PI	0.340	0.340	0.038	9.002	0.000 **
SN -> PI	0.206	0.206	0.032	6.492	0.000 **
PBC -> PI	0.231	0.231	0.039	5.879	0.000 **
PMN -> PI	0.050	0.050	0.040	1.245	0.213
PI -> AA	0.654	0.655	0.019	34.913	0.000 **

Note: ** p < 0.001.

show that AT, SN, and PBC significantly influence PI, which also strongly affects AA. However, PMN does not significantly affect PI, as the p-value > 0.05 indicates.

According to the analysis results,

Table 7. Model fit index.

	Saturated Model	Estimated Model
SRMR	0.048	0.145
d_ULS	2.263	20.912
d_G	0.783	1.245
Chi-square	7763.703	10,264.734
NFI	0.847	0.798

presents the model fit index, showing that the Standardized Root Mean Square Residual (SRMR) value for the saturated model is 0.048, which indicates a good model fit since it is below the ideal threshold of 0.08. On the other hand, the estimated model has an SRMR value of 0.145; this implies that the model does not align well with the data. For the NFI, the saturated model has a value of 0.847, indicating a good model fit. In contrast, the estimated model shows an NFI value of 0.798, which is lower and suggests that the model does not fit the data well. In conclusion, these results show that the saturated model offers more suitable data than the estimated model, emphasizing the need for adjustments in the estimated model to improve its fit.

4.4. Mediation Effect of Psychological Factors

Based on the analysis results,

Table 8. The result of mediation affects psychological factors.

	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values
TE -> AT -> PI	0.095	0.095	0.015	6.451	0.000 **
TE -> PBC -> PI	0.071	0.071	0.014	5.139	0.000 **
CO -> AT -> PI	0.041	0.041	0.011	3.834	0.000 **
CO -> PBC -> PI	0.039	0.039	0.009	4.145	0.000 **
IN -> AT -> PI	0.016	0.016	0.011	1.384	0.167
IN -> PBC -> PI	0.030	0.030	0.009	3.190	0.001 **
PO -> AT -> PI	0.132	0.132	0.020	6.683	0.000 **
PO -> SN -> PI	0.104	0.104	0.018	5.650	0.000 **
PO -> PBC -> PI	0.065	0.065	0.013	5.071	0.000 **

Note: ** p < 0.001.

shows several significant mediation effects of psychological factors in the relationships between the studied variables, with p-values well below 0.05. The mediation effect from IN -> AT -> PI is insignificant, with a p-value of 0.167, indicating that this factor does not significantly affect shaping attitudes and PI through mediation. High t-statistics in significant mediation effects, such as PO -> AT -> PI (6.683) and TE -> AT -> PI (6.451), indicate a substantial strength of the relationship. These findings suggest that TE and PO issues significantly impact customer attitudes and purchase intentions.

4.5. Moderating Effect of Demographic

The moderation analysis results show that gender moderates the relationship between psychological factors and PI. The analysis indicates that, of all the interactions tested, gender moderates the relationship between PMN and PI. Other interactions are not significant (p > 0.05). As shown in

Table 9. The result of moderating.

Path Coefficient	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values
GENDER x PBC -> PI	0.067	0.066	0.04	1.693	0.091
GENDER x PMN -> PI	−0.081	−0.08	0.039	2.062	0.039 *
GENDER x AT -> PI	0.041	0.042	0.038	1.09	0.276
GENDER x SN -> PI	−0.032	−0.033	0.034	0.961	0.337

Note: * p < 0.05.

, these results indicate that gender is an important moderator in the relationship between PMN and the intention to purchase EMs in Indonesia.

Table 10. Hypotheses.

Hypothesis	Path	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values	Result
H1	AT -> PI	0.344	0.344	0.037	9.174	0.000 **	Accepted
H2	SN -> PI	0.203	0.202	0.032	6.280	0.000 **	Accepted
H3	PBC -> PI	0.233	0.233	0.039	5.949	0.000 **	Accepted
H4	PMN -> PI	0.047	0.047	0.039	1.204	0.229	Rejected
H5	TE -> AT -> PI	0.095	0.095	0.015	6.451	0.000 **	Accepted
H6	TE -> PBC -> PI	0.071	0.071	0.014	5.139	0.000 **	Accepted
H7	CO -> AT -> PI	0.041	0.041	0.011	3.834	0.000 **	Accepted
H8	CO -> PBC -> PI	0.039	0.039	0.009	4.145	0.000 **	Accepted
H9	IN -> AT -> PI	0.016	0.016	0.011	1.384	0.167	Rejected
H10	IN -> PBC -> PI	0.03	0.03	0.009	3.190	0.001 **	Accepted
H11	PO -> AT -> PI	0.132	0.132	0.02	6.683	0.000 **	Accepted
H12	PO -> SN -> PI	0.104	0.104	0.018	5.650	0.000 **	Accepted
H13	PO -> PBC -> PI	0.065	0.065	0.013	5.071	0.000 **	Accepted
H14	PI -> AA	0.66	0.66	0.018	36.26	0.000 **	Accepted

Note: ** p < 0.001.

presents the results of the hypothesized relationships analyzed through standard paths and statistical significance based on the proposed hypotheses. Several factors, such as AT, SN, PBC, TE, CO, IN, and PO, contributed to PI. These findings support hypotheses H1, H2, H3, H5, H6, H7, H8, H10, H11, H12, and H13. Furthermore, PI was also determined to have a significant impact on AA, supporting hypothesis H14. However, two factors did not show significant relationships with PI. PMN directly and IN through AT led to rejecting hypotheses H4 and H9.

5. Discussion

5.1. Hypothesis Relationship

This study investigates consumers’ willingness to adopt EMs using an adoption model that considers psychological factors (AT, SN, PBC, and PMN), situational factors (TE and CO), contextual factors (IN and PO), and demographic factors (gender). Data analysis using PLS-SEM 4 provides significant insights into the relationships between constructs in understanding purchase intention (PI) and actual adoption (AA). The results provide a better understanding of customer behavior. This study focuses on consumers’ intention to adopt EM, mainly PI and AA. The analysis results reveal several key findings, which are discussed below.

Hypothesis 1, which states that AT significantly influences the decision to adopt EM, is supported. This statement is consistent with previous studies using the TPB, which shows that AT is crucial to adoption intention. The results support prior research, such as those by Dutta and Hwang [34], Shakeel et al. [27], Mohamed et al. [32], Adnan et al. [31], Huang and Ge [33], Rahmawati et al. [17], Deka et al. [19], and Buhmann et al. [21]. This study verifies that an optimistic attitude shapes consumers’ intention to purchase EVs.

Hypothesis 2, which states that SN significantly influences the decision to adopt EMs, is supported. This result is consistent with the research by Shakeel et al. [27] in Pakistan and Yegin and Ikram [15] in Turkey, showing that social support and pressure from family and the surrounding environment can encourage consumers to adopt environmentally friendly vehicles. Other studies supporting this finding include Mohamed et al. [32], Adnan et al. [31], Dutta and Hwang [34], Rahmawati et al. [17], Deka et al. [19], and Buhmann et al. [21].

Hypothesis 3, which states that PBC significantly influences the decision to adopt EM, is supported. This observation is consistent with the studies by Wang et al. [37] in China, which emphasize that the perception of one’s ability to purchase and use EVs influences the PI. This result is also supported by research from Mohamed et al. [32], Adnan et al. [31], Huang and Ge [33], Dutta and Hwang [34], Rahmawati et al. [17], Tiwari et al. [20], Deka et al. [19], and Buhmann et al. [21].

Hypothesis 4, which states that PMN significantly influences the decision to adopt EM, is rejected. This finding contradicts Schwartz’s [36] view, which emphasizes the importance of moral norms in promoting environmentally friendly behaviors. This result suggests that individual responsibility toward the environment is not a significant aspect of the purchase decision in Indonesia. It indicates the need to raise awareness about moral responsibility toward the environment through public education. This finding contrasts with studies by Mohamed et al. [32], Adnan et al. [31], Tiwari et al. [20], and Buhmann et al. [21] but consistent with the research by Deka et al. [19].

Hypothesis 5, which states that TE substantially impacts the decision to adopt EM based on the AT factor, is accepted. This result indicates that technology, particularly vehicle performance such as battery life, positively affects consumer attitudes, enhancing purchase intentions. Although some views suggest that technology can be a barrier to Rahmawati et al. [17], this finding supports the idea that technological innovation improves consumer perceptions and strengthens purchase intentions, consistent with the findings of Dutta and Hwang [34].

Hypothesis 6, which states that TE substantially impacts the decision to adopt EM based on PBC, is accepted. This finding supports the research of Mohamed et al. [32], which indicates that EV technology enhances consumer confidence in its use. This statement is aligned with Ajzen’s definition of PBC [26], which includes perceptions of technology, price, knowledge of using EVs, and the capability to adopt them.

Hypothesis 7, which states that CO significantly influences the decision to adopt EM through AT, is accepted. CO substantially influences consumers’ AT and PBC control toward EM. CO is important in shaping positive or negative perceptions towards EM. As found by Rahmawati et al. [17] and Dutta and Hwang [34]. According to Deka et al. [19], the influence of CO can be mediated by PBC, where cost-reducing policies, such as subsidies, can strengthen the purchase intention if consumers feel comfortable using EVs.

Hypothesis 8, which states that CO significantly influences the decision to adopt EM through PBC, is accepted. This statement is supported by Ajzen’s definition of PBC [26], which includes perceptions of technology, price, knowledge of how to use EVs, and the ability to adopt them. According to Deka et al. [19], the influence of CO can be mediated by PBC, where cost reduction policies, such as subsidies, can strengthen the purchase intention if consumers feel comfortable using EMs.

Hypothesis 9, which states that IN substantially impacts the decision to adopt EMs through AT, is rejected. Based on the analysis, IN was found to have no significant influence on AT toward adopting EMs. This result indicates that the charging infrastructure in Indonesia is insufficient to shape positive attitudes toward EMs. This finding contradicts Krupa et al. [51] in the United States, highlighting that the presence of infrastructure can significantly influence the development of consumer attitudes. One reason for this finding is that the infrastructure for charging in Indonesia is still inadequate, so consumers do not yet perceive it as a significant advantage for adopting EMs. Without sufficient and reliable infrastructure, consumers may hesitate and not be encouraged to change their attitudes toward EMs.

Hypothesis 10, which states that IN significantly influences the decision to adopt EM through PBC, is accepted. This result means that consumers may not consider the existing infrastructure adequate to influence their attitudes, but they feel more confident controlling their purchase decisions when infrastructure is available. For example, when consumers know that there are sufficient charging stations at strategic locations, they feel more confident that their basic EM needs can be met. Therefore, infrastructure development remains crucial in enhancing consumers’ confidence to adopt this new technology. This statement is supported by She et al. [30], who argue that the availability of infrastructure addresses consumers’ concerns before deciding to adopt EVs.

Hypothesis 11, which states that PO significantly influences the desire to adopt EMs through AT, is accepted. This result reaffirms that proactive policies, such as subsidies and financial incentives, can change consumer attitudes to be more favorable toward EMs, as found by Ehsan et al. [18]. Additionally, She et al. [30] show that government policies play a crucial role in influencing consumers’ attitudes toward adopting environmentally friendly vehicles.

Hypothesis 12, which states that attitude strongly determines the desire to use EMs, is accepted. Government policies supporting environmentally friendly vehicles can affect consumer attitudes toward vehicle choices and enhance social support for adopting EVs. Rahmawati et al. [17] state that incentive policies can help reduce consumer concerns regarding costs and charging infrastructure, influencing purchase intention. This statement is supported by the definition of subjective norm explained by Wang et al. [37], which states that when individuals believe they should execute a specific activity, they do so due to societal pressure.

Hypothesis 13, which states that PO exerts a substantial positive impact on the intention to adopt EMs by influencing PBC, is accepted. Government policies directly influence individual attitudes and create social pressure on EMs. For instance, by providing incentives or showing a strong commitment to reducing emissions, the government can influence the public to support the decision to adopt EM more widely. This finding aligns with Dutta and Hwang [34] and She et al. [30], who show that such policies strengthen purchase intentions.

Hypothesis 14, which states that adoption intention significantly positively influences the AA of EM, is accepted. This result means that in Indonesia, a person’s intention to adopt an EM plays a crucial role in determining whether they use the EM. The stronger the individual’s intention to switch to an EM, the more likely they will turn that intention into actual behavior, such as purchasing or using the EM in daily activities.

Hypothesis 15, which states that gender plays a role in strengthening the influence of moral norms on the desire to adopt EMs, is accepted. This finding shows that PMN influences men more than women. Marketing strategies highlighting moral responsibility toward the environment may attract male consumers in Indonesia more effectively. This finding contradicts research by Jayasingh et al. [42] and Habich-Sobiegalla et al. in Russia [43], which suggest that women are more inclined to purchase electric two-wheelers than men. However, it aligns with the study by Esteves et al. [58], which finds that men prefer EVs more because they are more interested in technological innovations and vehicle performance.

5.2. Managerial Implications

Based on the results of this research, we propose various implications for producers, policy-makers, and other stakeholders engaged in developing and marketing electric motorcycles (EMs). We discuss the managerial implications of key factors, including psychological, technological, cost-related, infrastructure, and policy aspects.

Psychological factors are crucial in influencing consumers’ decisions to switch to EMs. Manufacturers must realize that many consumers feel anxious or unsure about the effectiveness and safety of using EMs. Therefore, manufacturers can design marketing campaigns, such as offering free trial programs for consumers. These trials provide an opportunity for customers to experience firsthand the benefits and comfort of riding an EM. Additionally, testimonials from early users or influential public figures can be used to enhance consumer confidence. In terms of technology, developing more efficient batteries and faster charging systems is key to improving the comfort and performance of EMs. Manufacturers should focus on these innovations. Additionally, ensuring easy maintenance and repair services is necessary to provide an optimal user experience.

Cost is also a significant factor, as when the operational expenses of EMs are reduced, the higher initial price still presents a barrier. The government could provide additional incentives or subsidies to reduce the purchase cost, while manufacturers can offer flexible financing schemes. Accessible charging infrastructure is also crucial, so developing charging stations in strategic locations such as shopping centers, tourist destinations, and busy public areas should be encouraged. Providing charging facilities at these locations will make it easier for consumers to recharge their bikes, enhance user comfort, and boost confidence in EM. Government policy also plays a vital role in promoting the decision to adopt EM. Fiscal incentives, regulations supporting the use of EM, and educational campaigns about the environmental benefits of EVs can accelerate adoption.

Collaboration between the government, manufacturers, and the private sector is essential to create an ecosystem that supports the EM market’s growth. By integrating psychological, technological, cost, infrastructure, and policy approaches, it is hoped that the transition to EVs can proceed more smoothly and provide long-term benefits for society and the environment.

6. Conclusions

This study uses psychological, situational, contextual, and demographic approaches to determine the factors influencing the intention to purchase EM in Indonesia. Referring to the psychological perspective, AT, SN, and PBC significantly affect PI, highlighting the importance of beliefs and social pressure in shaping consumer decisions. However, PMN does not have a significant impact, indicating that individual responsibility toward the environment is not yet a major driver for adopting EVs in Indonesia.

Situational factors such as TE and CO are proven to be significant variables through AT and PBC. Technology plays an important role as it continues to evolve, bringing innovations that enhance the performance and comfort of EM. Improvements in technological performance aligned with consumer needs and reductions in purchasing and operational costs have increased consumer interest in adopting EMs.

Based on the contextual perspective, IN and PO have differing impacts on adopting EMs. IN does not significantly influence consumer AT or PI. This result may be due to the limited availability of charging facilities in Indonesia, which has not sufficiently encouraged consumers to adopt EMs. Conversely, government policies significantly contribute to promoting the decision to adopt EMs. Various policies, such as incentives, subsidies, and unique preferences, play key roles in shaping AT, SN, and PBC. These findings underscore the importance of sustained policy support to accelerate the transition toward widespread use of EVs.

Referring to a moderation perspective, the results show that demographic factors, particularly gender, influence the relationship between PMN and PI but do not significantly affect other relationships. PMN influences the PI of EMs as it relates to individuals’ sense of responsibility toward the environment. When someone feels a moral obligation to reduce the negative impact of pollution and climate change, they tend to choose more environmentally friendly alternatives, such as EMs. Therefore, demographic factors must be considered in formulating more targeted marketing strategies to increase the decision to adopt EMs in Indonesia, considering consumer characteristics that may be more sensitive to environmental issues.

This study has limitations as it has not examined travel behavior and environmental awareness, which may influence the decision to adopt electric vehicles. The main focus of this research is limited to electric motorcycles, without considering the geographical differences between urban and rural areas.

Future research should expand its focus to other types of EVs, such as electric cars or electric-based public transportation while considering additional geographical dimensions related to the uneven availability of electricity supply. Remote and mountainous areas, which often face challenges regarding limited electricity supply, should be considered. Understanding these differences is crucial in identifying solutions tailored to the existing geographical conditions, such as relying on renewable energy sources like solar or wind power to support electricity supply for electric vehicles in these regions. Furthermore, expanding demographic dimensions such as age is also important to understand the influence of preferences and knowledge on electric vehicles, particularly among younger generations that are more concerned with technology and environmental issues.