1. Introduction
The autonomous vehicles (AVs) considered in this study are driverless cars that adopt sensing and communication technologies to navigate safely and efficiently without human intervention. They will be an integral part of a future intelligent transportation system that will require a coordinated awareness of the traffic environment, planning decisions, and multilevel assisted driving to ensure public safety. Research into AVs and supporting infrastructure, such as the 5G cellular network, has recently made rapid progress, and the subsequent development of AVs has led to their increased commercialization. Some vehicle manufacturers (e.g., BMW and Volvo) project to incorporate autonomous driving features in their traditional vehicles by 2040 [
1]. By 2040, AVs are expected to represent 25% of the global market for private vehicles, according to previous studies [
2]. Thus, AVs have become a high-potential alternative to positively promote public safety and a promising traffic environment.
A wide range of societal and individual benefits will result from the adoption of AVs in the private vehicle market, including enhanced traffic safety, urban livability, and a better user experience [
2,
3,
4,
5,
6,
7]. Increased AV use would reduce vehicle collisions by 90% because more than 90% of traffic accidents are caused by driver errors and misjudgments [
3]. Urban livability and public health will be enhanced due to alleviated traffic congestion, lower car ownership, reduced needs for transportation infrastructure (e.g., parking spaces), and reduced transport emissions. Massar et al. [
6] revealed that the highest emissions reduction might occur between 60 and 80% AV penetration into the daily mobility, while the expected emissions reduction might not be realized at a lower rate of AV penetration. The use of AVs will enable users to complete more productive tasks during traveling and will provide mobility to people who would otherwise be restricted in travel (e.g., the elderly and the disabled). However, Haboucha et al. [
8] found that the use of AVs was met with significant popular opposition due to the perception of there being a significant technological leap from the use of traditional vehicles. In 2020, the Ministry of Transport of China issued guidelines on promoting the development and application of autonomous driving technology, including enhancing the research of autonomous driving technology, improving the intelligence level of road infrastructure, promoting the pilot application of autonomous driving technology, and designing the risk prevention and control system. However, the application of autonomous driving technology is still very limited. AVs will not realize their benefits if potential users do not adopt them because of perceived drawbacks. Potential users are expected to switch to AVs rather than drive themselves, so their perspectives on AVs and the factors that influence their intentions to adopt AVs merit examination.
Extensive studies have adopted the technology acceptance model (TAM) with extensive factors (trust, relative advantage, and privacy concerns) to identify the psychological factors (e.g., perceived usefulness and perceived ease of use) that influence individual intentions to adopt an AV [
9,
10,
11,
12]. However, some noteworthy factors, such as perceived playfulness, were not introduced or examined. Despite many studies of intentions to adopt AVs, little is known about how policy measures influence attitudes towards AV adoption. Many efforts were made and many policies were introduced to incentivize electric vehicle (EV) use in the early stage of EV development. These efforts were mainly to subsidize EV buyers, construct charging facilities, provide accurate information related to EV performance, and construct dedicated parking spaces for EVs [
13,
14,
15,
16]. Many researchers have examined the effects of financial incentivization on intentions to adopt EVs [
13,
17,
18]. Wang et al. [
19] additionally considered two policy measures (the provision of information and raising the awareness of user convenience) and investigated the effects of these policies on intentions to adopt EVs to provide more insights into EV adoption. These policies directed towards EV promotion might be similarly applied in promoting AV adoption. In the future mixed-traffic environment of driven and driverless vehicles, the use of AVs may introduce legal issues [
20,
21]. For example, the liability for a traffic accident is difficult to define when an AV crashes with another road user or pedestrian. To address these legal issues, the laws, insurance, and oversight related to AVs should be strengthened prior to the widespread use of AVs. Thus, legal review and reform should be a fourth policy measure to promote AV adoption. In this study, financial incentivization was intended to decrease the costs of purchasing and operating an AV and included policies such as direct subsidies, tax exemptions, and road toll exemptions. Policies of AV information dissemination included information about the practicality, reliability, safety, ease of use, fuel consumption, and environmental performance of AVs, among other issues. Policies concerned with the convenience of AV use included dedicated road lanes, higher speed limits, and the removal of restrictions such as even and odd license plate number rules. Legal review and reform policies were intended to provide a supportive legal environment for users of AVs and included measures such as improving the application of laws, insurance, and governance to AVs. We refer to this as the legal normalization of AVs. In addition, user demographic features also play a significant impact because they can demonstrate differences in demographics on user intentions to adopt an AV. Previous studies have considered demographic variables as control variables [
8,
22,
23], but little work has been conducted on their moderating effects on influencing factors related to adoption intention.
The purpose of this study was to (1) examine the influences of TAM-related psychological factors and policy measures on AV adoption intention and (2) explore the moderating effects of demographic variables on the relationships between independent variables and AV adoption intention. Our study contributes to the research literature by providing an increased understanding of individual preferences in AV adoption. The empirical evidence gathered in this research has practical implications for the future wider adoption of AVs in China.
5. Discussion
The findings of this study show that all TAM-related factors and policy measures significantly influenced the adoption intention of potential Chinese AV users and that several demographic factors could moderate the influences of these independent factors on adoption intention.
Among the three psychological factors, PU had a stronger impact on adoption intention than PEU, which is consistent with previous studies of AVs [
9,
32,
33,
34]. The effect of PP has rarely been investigated in AV research. We observed that PP showed the strongest effect on AV adoption intention, which indicates that PP is the most important factor in user decision making. A plausible reason is that users are more motivated to purchase a new technological product (e.g., a foldable smartphone or a wearable VR/AR) by the playfulness it engenders rather than its usefulness or ease of use. Thus, if users perceive that many features of AVs are interesting and using AVs is fun, they will be more likely to adopt an AV. Furthermore, male users will be more strongly impacted by PU and PP in their intentions to adopt an AV than female users because they might favor something that will boost their productivity and entertainment. The effect of PP on AV adoption intention was also influenced by education. Potential users who have higher education levels had a significantly greater impact of PP on the intention to adopt an AV than those who had lower education levels. The reasons might be that users with higher education levels have a better pursuit of life and travel quality, hoping to obtain more enjoyment from various social activities. Higher-income users had a significantly greater impact of PU on adoption intention than lower-income users. One plausible reason is that higher-income users usually pursue more economic efficiency and prefer some instrumental products that can deliver real benefits.
Among the four policy measures, ID had the greatest influence on adoption intention. AVs are vehicles that embody innovative technology and are in their infancy. These findings emphasize the value of adopting informational campaigns to encourage users’ understanding of novel technologies’ capabilities and limitations in their adoption process. This is because many early users of new technologies actively seek out information pertaining to the new technologies’ anticipated characteristics as well as any advantages or disadvantages related to their use. Information about their many attributes may be unknown or misunderstood by potential users, and this may negatively influence adoption intention. Some studies have found that many potential users are inclined to resist AVs or have a wait-and-see attitude due to emerging issues (e.g., trust, privacy, and reliability) around AVs [
8,
63,
64]. Therefore, providing more information may motivate potential users to switch from resisting AVs to acquiring them. FI was the second most important factor. As more financial incentivization is provided, user intentions to adopt an AV are strengthened, which is consistent with previous studies [
49,
50]. The third most important factor was CO. The results suggest that providing dedicated lanes, allowing higher speed limits for Avs, and lifting restrictions of the rules of even and odd license plate numbers for AVs would attractive more users to adopt AVs. The weakest influencing factor was LN. This means that, although legal issues related to AVs are a hot topic in academia and industry [
20,
21], users pay little attention to them in the infancy of AV uptake, so LN has a weak impact on adoption intention. The influence of LN on AV adoption intention might become more topical in the future when AV users are confronted with significant legal issues. Furthermore, more highly educated users will be more greatly impacted by LN on their adoption intention because they will be aware of the legal risks associated with using AVs and will pay more attention to legal issues related to AVs. Expectedly, lower-income users had a significantly greater impact of FI on adoption intention than higher-income users, as they are more price sensitive to products. This also explains why potential users who live in lower-tier cities had a greater impact of FI on adoption intention than those who live in higher-tier cities. Additionally, the influence of CO on adoption intention was stronger among the higher-tier city users because these cities are well developed and can provide more infrastructure to promote the commercialization of AVs. Interestingly, this study revealed that the potential users who have less private cars will be more greatly influenced by ID in their intention to adopt AVs. One plausible reason is that they have less awareness of car-related information; hence, providing more positive information about AVs might be more beneficial to encourage them to adopt AVs.
6. Conclusions
Our study reveals how psychological factors and policy measures influence Chinese users to adopt AVs and shows how demographic factors moderate these relationships. Three important theoretical conclusions and some managerial applications are provided to motivate AV safety and ecofriendly development:
(1) The present study extended the TAM model by introducing an extensive factor, PP, when examining the psychological factors that might influence AV adoption intention. The findings revealed that the importance of the three psychological factors is ordered (high–low) PP (β = 0.524, p < 0.001), PU (β = 0.278, p < 0.001), and PEU (β = 0.141, p < 0.001). This ranking shows that early adopters of AVs are concerned most with the enjoyment gained from using an AV. Therefore, to facilitate AV safety and maturing, in addition to improving the functionality and technology of AVs, manufacturers and AV developers are advised to develop more in-vehicle entertainment features, which might make them more attractive to potential adopters.
(2) Different policy measures have different influences on AV uptake. We identified four policy areas, including finance, information, convenience, and legislation, in which policy can influence adoption. The importance of the four policy measures is ordered (high–low) ID (β = 0.348, p < 0.001), FI (β = 0.216, p < 0.001), CO (β = 0.121, p < 0.001), and LN (β = 0.113, p < 0.001). Thus, the government should put more effort into providing information related to the benefits of AVs (e.g., the practicality, reliability, safety, ease of use, fuel consumption, and environmental performance) to promote the early adoption of AVs.
(3) Our study investigated the moderating effects of demographic factors on the relationships between independent variables and AV adoption intention. Several demographic variables (gender, education, income, the number of private cars owned by a family, and the types of cities) showed differences in the influences of psychological factors and policy measures on AV adoption intention. Therefore, policymakers should tailor interventions and policies to potential AV users with different demographic characteristics. In the future, the above psychological investigation findings may be used to significantly promote AV development.