Analysis of Factors Influencing the Choice between Ownership and Sharing: Qualitative and Quantitative Survey Results on Car Sharing Service Users Conducted in Japan

Abstract

The purpose of this study is to propose new conditions with which to classify consumers who choose new modes of consumption other than ownership that influence the growth of the sharing economy in Japan. Our findings highlight the characteristics of consumers who choose the new consumption mode from a consumer behavior perspective, which is different from previous studies that have focused on ownership to date. By analyzing the results of a qualitative survey of car sharing services (CSS) users through group interviews, hypotheses were formulated regarding the dimensions that classify the characteristics of consumers who choose to use CSS, and these hypotheses were verified by analyzing the results of a quantitative survey. CSS is considered a substitute for the private car and is expected to contribute to the effective use of resources, but it is used as a complement to the private car and to realize a commitment to the car. The method proposed in this study to categorize and understand the factors that influence consumers’ decisions to choose new modes of consumption other than ownership may have implications for marketing strategies, allowing them to respond to the growth of the sharing economy in other countries as well. This will support sustainability and the various experiences of consumption.

1. Introduction

Research has suggested that “during the last decade, observers have noted that markets are giving way to networks, and alternative modes of acquisition and consumption are emerging beside ownership” [1] (p. 881). Sharing is defined as an action or process involving a social and cultural context that is distributed from oneself to others and vice versa [2]. This does not only cover modes of transactions that do not involve the transfer of ownership, as seen in representative cases such as Uber, Airbnb, and Zipcar. Bardhi and Eckhardt (2012) [1] also defined market-mediated transactions that do not involve the transfer of ownership as Access-Based Consumption and distinguished it from sharing. Access has characteristics that contrast to those of pure ownership and pure sharing and is located in the middle of a continuum between the two extremes [1]. With the emergence of various types of access models, including individual peer-to-peer sharing rather than corporate assets, the boundary between sharing and access remains blurred. Therefore, further research is needed on the conditions for classifying them [1].

In 2015, the term “sharing economy” was added to the Oxford English Dictionary as “an economic system in which assets or services are shared between private individuals, either for free or for a fee, typically by means of the Internet” [3] (p. 277). In Japan, the market size of the sharing economy is estimated to grow from JPY 2,419.8 billion in FY2021 to JPY 14,279.9 billion in FY2030 if issues such as the uncertainty caused by COVID-19 and lack of awareness are resolved [4].

While there are various understandings of new modes of consumption as alternatives to ownership, there is a common understanding of sharing resources. The sharing economy is also closely related to the product service systems (PSS) debate in the sense that it reduces environmental impacts through the efficient use of resources. PSS arose from a Nordic research community that studied competitiveness and sustainability [5]. In general, PSS are defined as a “product(s) and service(s) combined in a system to deliver required user functionality in a way that reduces the impact on the environment” [6] (p. 1545), with an emphasis on “sale of use” rather than “sale of product” [7] (p. 1545). In PSS, the value provided is either product-centric or service-centric, and the continuum between these two extremes (pure product or pure service) is further classified into three PSS: “Product-oriented”, “Use-oriented”, and “Result-oriented”. Although access is considered to fall under “Use-oriented” or “Result-oriented” PSS, again, the conditions for classifying individual cases of various consumption modes on the continuum are not clear.

Research shows that “the sharing economy is growing rapidly while disrupting many traditional industries such as the hotel and transportation sectors” [3] (p. 277): for example, Airbnb is perceived by consumers as an alternative to hotels [3,8], and ownership in product categories that were once considered assets, such as cars and houses, is declining [8]. Consumers use Zipcar, Airbnb, and sharing resources among family members as alternatives to ownership [1,9].

Thus, “a shift in the established politics of consumption is emerging in which access is gaining symbolic capital as a more economically and ecologically viable, flexible, and freeing consumption mode” [1] (p. 895). However, “the consumer behavior literature to date has focused primarily on solid consumption” [10] (p. 582) and “the sharing economy’s implications for marketing thought and practice remain unclear” [11] (p. 5), and it has been argued that new concepts of consumer behavior need to be defined [10] (p. 582). Prior research has dealt primarily with the issue of substitutability, either interbrand choice [12] or substitution between households and markets [13]. The consumer behavior research on ownership has dealt with interbrand choice within product categories [12]; however, in considering contemporary environmental issues, it has become necessary to deal with consumer choice beyond product categories [14].

Service marketing research has proposed that offerings consist of a combination of tangible and intangible goods. Intangible goods contain four elements: usage rights of material goods, information, usage rights of information, and human services. Tangible and intangible goods are classified into five elements (material goods, usage rights of material goods, information, usage rights of information, and human services) based on these two axes by (1) whether the subject producing utility is material or immaterial, and (2) whether the ownership of the subject producing utility is transferred or not [15,16]. Offerings generate choices across product categories, with the main goods that comprise them changing their consumption mode as they are substituted for other goods. In other words, the main factors that give rise to substitutability between these goods allow us to distinguish between ownership and the different modes of consumption that replace it. This theory of substitutability of goods proposes, for example, that in the case of private cars and car sharing services (CSS), the main goods that constitute the offerings are material goods and the usage rights of material goods, respectively, and that the main factors that generate substitution relations between these goods are the cost of ownership, the number of times the goods are used, and uncertainty in the use situation [17].

The idea is that these factors influence the consumer’s decision, i.e., choosing whether to own or rent. In choosing either mode of consumption, consumers compare and consider the cost of owning and maintaining a private car versus the cost of CSS depending on the frequency of use [14]. When considering the purchase of a private car, consumers perceive uncertainty about whether they will use it in the future for an extended period of time. On the other hand, a private car provides them with the security of knowing that it will always be available [14]. When they use CSS, however, they perceive uncertainty in that they will not be able to rent it freely at any time [18]. This also requires the ability to use information technology, such as making reservations on a website [14]. These factors will influence their decision making.

Additional options for solving the problem of moving from place to place can include other usage rights of material goods, such as using cabs, renting cars, or carpooling. By analyzing the factors that influence consumer decision making regarding these various options across product categories, we can identify the conditions that categorize different modes of consumption on a continuum from pure ownership to pure sharing [12]. This will provide a basis for firms to respond to the evolution of the sharing economy in their marketing strategy.

The CSS market in Japan continues to grow, with 19,346 CSS vehicle stations in Japan (+1.2% YoY), 43,460 vehicles (+7.9% YoY), and 2,245,156 members (+9.7% YoY) as of March 2021 [19]. The number of stations and the number of vehicles have both increased. The metropolitan areas of Tokyo, Osaka, Kanagawa, and Hyogo prefectures account for approximately 70% of the nation’s total number of stations and vehicles [20]. The number of stations and the number of vehicles in these metropolitan areas are increasing. It is assumed that the well-developed public transportation system in these metropolitan areas, coupled with the skyrocketing parking costs, has encouraged the spread of CSS as an alternative to the private car.

The purpose of this study was to propose the new conditions to classify consumers who choose new modes of consumption other than ownership that influence the growth of the sharing economy in Japan. By motivating people to make choices in a meritorious manner, it is believed that their understanding of environmental issues will be deepened if the scope of their use can be expanded as a social habit [14,21]. We analyzed the factors that influence consumers’ decisions to choose or continue to use CSS from the perspective of consumer behavior, and by confirming and classifying the relationship with consumer characteristics, it aimed to formulate and substantiate hypotheses regarding the factors that govern choosing the consumption mode of renting rather than owning a private car. The former analysis was based on qualitative findings from group interviews with CSS users, and the latter on quantitative findings from a questionnaire survey of CSS users, both of which were conducted in Japan. The analysis data and methods are described in the next section.

2. Materials and Methods

The investigation in this study focused on the alternative relationship between private cars and CSS, even though there are various alternative consumption modes across product categories, such as access and sharing in lieu of ownership. The main item that constitutes each offering is the same car, which in itself eliminates its impact on decision making. In addition, because the main goods are the material goods and the usage rights of material goods, which are different goods, it was possible to identify the factors that generated the substitutive relationship between these goods.

Study 1 was a qualitative study based on group interviews with CSS users or those who had used CSS. We hired a research firm to conduct a screening survey of 11,805 monitors, who were asked to register, reserve a vehicle, and pay for it on a website in order to use CSS. Considering this, the screening survey targeted monitors aged 20–59, who were able to use computers, smartphones, and other devices and Internet information technology. The screening survey asked about CSS companies they now used or had used and the length of time they have used CSS. Other questions asked were whether they had a private car and how often they used CSS. These corresponded to the ability to use information technology, cost of ownership, and frequency of use, factors that are thought to produce the substitutive relationship between material goods and the usage rights of material goods discussed earlier.

Of the 11,805 monitors, 11,694 were between the ages of 20 and 59. Of these, 442 were CSS users or had used CSS before, accounting for 3.8% of the total. Considering that the population aged 20–59 was 61.12 million [22], the ratio of 2.25 million registered CSS users to that number was 3.7%; thus, it is fair to say that the extracted monitors reflected the penetration rate of CSS in Japan. The research firm asked 442 selected CSS users or experienced CSS users to participate in interviews, and 16 were selected. Of the 16 people selected, four were grouped into one group with similar age, gender, family structure, and other attributes. We conducted group interviews with each group for 120 min (

Table 1. Demographics of group interviewees (conducted 9–12 March 2022).

Subject 1	Gender	Age/Occupation	Family Structure	Residential Area	Availability of Private Car
C201	female	24/Nurse (full-time job)	partner	Akita (Tohoku area)	have
C202	female	43/System development (self-employed)	single	Tokyo (metropolitan area)	owned in the past
C203	female	27/Residential design (full-time job)	single	Saitama (Kanto area)	no ownership experience
C204	female	37/Childcare worker (full-time job)	partner, children	Osaka (metropolitan area)	have
C301	male	40/Software (full-time job)	partner, parents	Hyogo (metropolitan area)	have
C302	male	30/Administrative corporate (full-time job)	wife, son (age 0)	Tokyo (metropolitan area)	have
C303	male	56/IT-related office (full-time job)	single	Kanagawa (metropolitan area)	have
C304	male	32/Research (Forest resources) (full-time job)	parents	Tokyo (metropolitan area)	have
C401	female	40/Architectural design (full-time job)	single	Osaka (metropolitan area)	owned in the past
C402	female	30/IT Gym (full-time job)	single	Tokyo (metropolitan area)	no ownership experience
C403	female	32/N.A.	partner, 2 children	Kanagawa (metropolitan area)	have
C404	female	38/Nurse (full-time job)	partner, child	Fukuoka (Kyushu)	have
C501	male	23/Warehouse work (full-time job)	parents and siblings	Tokyo (metropolitan area)	no ownership experience
C502	male	45/Security (full-time job)	mother	Kanagawa (metropolitan area)	owned in the past
C503	male	28/Sports club (full-time job)	single	Kanagawa (metropolitan area)	have
C504	male	37/Stock investment (self-employed)	single	Aichi (Chubu area)	have

¹ Third digit is group number.

). Transcripts of the statements were recorded in text (31 pages in total) and used in subsequent analyses.

We used the theories-in-use (TIU) as our research method, which has been proposed as one way to build marketing theories for new phenomena [23]. Since CSS is a new mode of consumption, different from the private car, the questions were designed with the research objectives of identifying the concept and practical definition of CSS, new consequences of CSS use, positive and negative antecedents of CSS use, moderating variables, and mediating variables as the survey objectives. Since private cars and CSS have an alternative relationship [1,9,18], questions assessing the boundaries between private car ownership and CSS use and their relationship were included. In addition, since private cars and CSS are choices that transcend product categories, it was assumed that highly abstract motivations and values would dictate the choice of CSS [24,25,26], so soft-laddering questions were also administered (

Table 2. Interview flow developed based on the theories-in-use (TIU).

Research Objectives	Questions
Clarify the use of car sharing services (CSS) (concept)	How did you come to use CSS?
	Why do you think CSS is useful and valuable? Why is it important? (soft-laddering)
	What does CSS mean to you?
Find a general precedent	What information was important to you when deciding to use CSS? What was the deciding factor?
Find the ‘positive’ antecedents (main drivers) of CSS use Find the ‘negative’ antecedents (main barriers) of CSS use	What were the good things about using CSS? Conversely, what were the bad points?
Evaluate ownership–CSS boundaries and ownership–CSS relationships	You have your own car. Why do you use CSS?
	(If you do not have a car) Would you like to buy a car?
	Has your use of CSS made you want to buy a car? On the other hand, have you been tempted not to buy a car?
	You said you had a car before, why did you give it up, what made you decide to use CSS?
	What is the difference between renting with CSS and buying or owning a car?
Evaluate construct boundaries (practical definition)	Does the fact that you can use the car only by yourself, or anytime you want to use it, affect whether you buy or rent it?
	Does renting a car affect your use of CSS in terms of reducing the number of cars produced, reducing the number of times you drive, or being environmentally friendly?
	Does the fact that you are connected to society by sharing your CSS experience with other users on sites and social networking sites affect your use of CSS?
	Do things, such as being connected to society by thinking about other users of CSS, influence your use of CSS?
Link CSS to new results	What were some of the results of using CSS that surprised you or that you did not expect?
Find the moderating and mediating variables	Would you continue to use CSS in the future?
	Would you like to use CSS more often?
	What would you need to do to increase the number of times you use CSS?
	Are you planning to stop using CSS?
	Why did you stop (or would you stop) using CSS?
	When were you happy with CSS?
	When did you feel bad using CSS?
	What else would you buy or rent a car for?

).

Study 2 was a survey-based quantitative study of those who were registered or had used CSS. This survey was conducted earlier than Study 1, as part of a study analyzing factors influencing the selection and continued use of CSS, but with a different purpose than Study 1. We hired a different research firm than the one described above to conduct a screening survey of 15,913 monitors (mid to end of February 2019). Of the 15,913 monitors, 12,845 were between 20 and 59 years old. The 576 people who now used or had used CSS comprised 4.5% of this total, which was not far from the penetration rate of CSS among the same age group in Japan mentioned earlier. We therefore concluded that the data from Study 2 generally reflected the penetration rate of CSS in Japan and could be used as data to substantiate Study 1. The Study 2 screening survey was similar to Study 1. Furthermore, the Study 2 survey asked 300 of the above monitors who use or have used CSS to complete a questionnaire related to their evaluation of the quality of CSS offerings and companies and of their experience using CSS. The items of the questionnaire were developed based on the SERVQUAL scale for the purpose of measuring CSS service quality [27] and the cocreation value scale for the purpose of measuring the experience of using CSS [28]. Both scales were adjusted to accommodate CSS [29]. To ascertain the consumer characteristics, we also asked about their evaluation of the value constructs [28]: “It is important to get rewards that match the effort of helping with car sharing”, “It is important to get status and certification with the car”, “It is more important to get a car that meets my needs and service”, “It is important to connect with others through the use of a car”, “It is important to get knowledge about cars, driving, and car sharing”, “It is important that using a car is fun”, using a 5-point Likert scale method (1 disagree, 5 agree). The Cronbach alpha reliability coefficient for this construct was 0.818. Since the questions in this survey included questions related to dimensions of the factors that determine the selection or continued use of CSS, as hypothesized in Study 1, we used this existing survey data for our analysis as Study 2 to make effective use of the limited research budget.

3. Results

3.1. Study 1

3.1.1. Qualitative Interviews Analysis

We used a KH Coder (3.Beta.04b) to perform an exploratory analysis of the transcripts of the aforementioned group interviews (First Cycle coding). KH Coder has a function that organizes the number of word occurrences in each data set to enable a multivariate analysis [30]. As in Higuchi (2019) [31], we extracted words that were characteristic of each subject (in the same direction and at a greater distance), since the analysis of the correspondence between the words extracted for each group and the subjects confirmed that there were differences in the characteristics of the subjects in each group. We independently identified first-order categories (Axil Coding) by reviewing the characteristic utterances containing those extracted words and found that they were saturated with no further new information or interpretations [32]. As in Challagalla, Murtha, and Jaworski (2014) [33], we further identified abstract second-order categories that encompassed these codes (

Table 3. Characteristic words corresponding to the subject, First Cycle Coding, first-order categories, second-order categories, and characteristic statements of the subject corresponding to these categories.

Subject	Extracted Feature Word	First Cycle Coding	First-Order Category	Characteristic Statements	Second-Order Category	Characteristic Statements
C201	private	objective	substitution	Since I’m in a state where I do not have my own car, it’s indispensable for private day trips and overnight trips and especially business trips.	Low commitment	I prefer to car share light cars in my private life, but there is no particular difference there.
C202	metropolitan area	cost	substitution	I live in the city and do not plan to purchase.	Low commitment	I am not looking for what kind of car I want to drive in a car share at all. It is enough if I can drive it as a means of transportation. At this stage, I don’t need a car that I can drive when I want to drive myself in Tokyo.
C203	buy	commitment	substitution	I have narrowed down the criteria for what I want to buy and would like to buy a car as soon as I can find one that meets those criteria, and the cost is acceptable in various ways.	High commitment	The majority of the cars I use for car sharing are in colors that are good, because they are perfectly functional, but not to my liking, and I am in the mood to not buy this color if I were to buy one myself.
C204	go out (e.g., on an excursion or outing)	proper use	complementary	We have one at home, but it’s not always at home because my husband basically drives it to work, and I use it when I want to use the car to go out with friends.	Low commitment	I use car share to go out when I have a lot of luggage and need a bigger car.
C301	rent	cost	substitution	I moved out, bought a car, rented a parking space, and then stopped using it because I had a car.	High commitment	If it’s a rare or luxury car or one that looks too awesome, I’m afraid of crashing it, so on the contrary, I can’t rent it. If you are going to rent a car that is likely to be damaged, it is better to rent and test drive the car the right way, even if you have to spend a little more.
C302	family	one’s (living) environment	complementary	I think a large part of this is due to the fact that the family structure requires a car.	Low commitment	Enjoying family time and travel is bigger and more important than that (social connections). I am 100% aware that they enrich my personal life.
C303	get in a car	person who craves the limelight	complementary	I might change my mind and take a ride if I see a Cooper.	High commitment	I was driving a so-called sports car type.
C304	be useful	means	substitution	I stopped using Times (a CSS company) about 3 years ago when my father returned and was able to use the car.	Low commitment	For me, having a car is a value in itself, so being able to use a car when I don’t have one is a value in itself.
C401	frequency	frequency	substitution	Since a car that is used infrequently will naturally have a lower priority, it is better to temporarily pay for a car share in order to reduce expenses when considering maintenance costs, etc.	High commitment	Considering how often I drive, it’s hard for me to think about owning a car at this point, but eventually I would like to own a car that I want to drive.
C402	Hold (in one’s hand)	cost	substitution	When you think it’s more expensive to own a car, consider whether there is a car share nearby.	High commitment	When I use a car share, I actually drive it from the perspective of what it would be like if I actually owned it.
C403	child	sense of security	complementary	They only make reservations at the last minute when they think that their child might catch a cold or something, so they end up not being able to make a reservation.	Low commitment	I don’t think I would use it unless there was a disaster or something like that and I had to send my kids or something.
C404	house	one’s (living) environment	substitution	I used to live in a car share place because it had a well-developed transportation network and I didn’t have to worry about not having a car; but since I moved, the environment has changed because it is a little inconvenient to have a car in the house without one in the family.	Low commitment	Since we only have one car at home, I use it to commute to work, and since my husband would be without a car when he wants to go golfing on weekdays, he has to borrow it or go to his parents’ house to borrow it, so there is a possibility of using it for such overlapping situations.
C501	Hold (in one’s hand)	one’s (living) environment	substitution	I think that as my living environment changes, my decision to have or not to have a car will change. Again, I don’t see the benefit of having a car when I live in a place where everything is just a short walk away.	High commitment	If I had my own car, I would take care of it for the rest of my life, because I think that’s an important attitude to have when you own things.
C502	frequency	frequency	substitution	I had my own car for work, but after a family member had a self-inflicted accident, I switched to car sharing because I didn’t use my car that often, considering the frequency of use.	Low commitment	The frequency of use is two to three times a month. When I need to go shopping in the neighborhood or when I feel like driving, I use the Times car share service in my neighborhood.
C503	allow passengers to board a car	objective	complementary	I often carry people, but I also carry large luggage, but I am over-specified for one person, so I feel that it would be a waste to use only one car to get to my destination, so I want to use them together.	High commitment	Not that I don’t want to make the car share dirty, but the Alphard I’m using now carries people most of the time and I don’t want to make the back seat dirty.
C504	buy	assets	substitution	I think there is quite a bit of wallet considerations in buying, renting, or giving it away. The cost aspect is significant.	Low commitment	When I bought mine, I thought about when I would sell it, so I decided to keep the car a safe color. Only I use the car, and the performance of the car will not change, and even if the car is a little older now, it will run properly.

). The process by which these categories were derived was as follows.

Consumers consider several preconditions when making the distinction between the use of material goods and the usage rights of material goods. First, since material goods are owned, they can be used at any time in principle, but consumers must bear the costs associated with their ownership. In addition, the ability to choose one’s preferred brand of goods may increase the benefits of such goods [34]. Thus, it was assumed that there is an alternative relationship between choosing material goods and choosing the usage rights of material goods. If the costs of ownership increase, such as the cost of parking and property insurance in the case of automobiles, people will choose the usage rights of material goods, and there are also means to substitute their own “human service” of driving, such as cabs. This would be the option of choice if the cost of acquiring driving skills and physical fitness became significant [34].

Thus, our choice of services will be diverse and will itself be an object of choice. As a result, for sharing to be useful, the conditions allowing it to be the first choice would need to be in place. In this study, we define this relationship as an alternative relationship.

The other relationship concerns the diversity of consumers’ needs regarding the product class. The substitution relation is the idea that material goods and the usage rights of material goods are incompatible, but consumers may purchase these two types of goods together for a category of products in which they would like to buy and use a second or third product. In this case, there is a substitution relationship with the second and third goods [14], but since consumers do not consider ownership from the beginning and find value in using many brands, we distinguish this from the aforementioned substitution relationship and organize it under the concept of complementarity [21]. For consumers, this is expressed in this way because their needs are met by using multiple brands. These two reasons for choice comprise one suggestion for advancing the sharing economy. Although this survey was conducted for the B2C market, in the B2B market, where product assortment is important, sharing by complementation is common, and the option of owning everything is considered to be rarer.

As theoretically described above, this study assumed two motives for consumer use. The first axis of the analysis of the interview content was, as expected, whether the relationship was alternative or complementary. Subjects make decisions based on the resources they own (private cars), costs, and purposes. In this interview, we found that there were subjects who owned a private car but also used CSS and used them differently depending on their purposes and desires, so we extended this dimension and derived a complementary dimension [35].

The second newly discovered axis was commitment to the category. Although the use of CSS for events and short periods of time was assumed based on the aforementioned factors that create substitutive relationships between goods, since many of the subjects’ statements about the subject of choice appeared in their statements, we established a dimension of commitment in which they made their own decisions about brands and other factors [36]. At the opposite end of the spectrum, subjects who delegated decision making to others, such as family members, depending on the purpose of use, were captured, confirming the existence of consumers with low levels of commitment [36].

Using these dimensions as a classification axis, we typified the factors that governed the choice or continued use of CSS. By identifying the characteristics of the subjects corresponding to each of the categorized quadrants, we named each quadrant as saving-oriented, minimalist, usage purpose-oriented, and waste-control type (Figure 1).

The saving-oriented type can be understood as the theoretically inferred users of CSS. This saving type has a low level of car-related involvement and chooses between owning a private car or using CSS depending on factors that create a substitutive relationship between material goods and the usage rights of material goods: cost of ownership, uncertainty of use, and number of times the goods are used. Minimalists are the type who use CSS to save money but are highly committed to their car and are able to make decisions about their choice to purchase a private car. Usage purpose-oriented consumers own a private car but also use CSS. This type has a low level of car-related commitment, and their choice is influenced by their family and purpose. The waste-control type owns a private car and is highly committed to car-related activities. Digital Virtual Consumption is analogous to bringing new experiences into daily life, helping people to manage and control the risks of daily life and their own needs and consumption [37].

3.1.2. Interpretation

We reviewed the attributes and characteristic statements of the subjects corresponding to these four types and interpreted each type as follows.

Saving-oriented type and minimalists, who are classified in the alternative dimension of the classification axis, are users or experienced users of CSS. That is, they either still use CSS because they do not own a private car, or they do not currently use CSS because they have purchased a private car or have access to a family-owned car. The presence or absence of a private car influences the choice of CSS and the decision to continue using it.

The usage purpose-oriented and waste-control types, classified by the complementary dimension, own a private car but still use CSS, and the presence or absence of a private car does not affect their choice of CSS or their decision to continue using CSS. The saving-oriented and usage purpose-oriented, which are classified in the low-commitment dimension, are less involved with the car, and they view the car as a means of transportation and are more concerned with the purpose of using the car, i.e., having their needs met.

On the other hand, the minimalist and waste-control types, classified in the high commitment dimension, have a high level of involvement regarding the car and view the car as a means of self-actualization, such as commitment and status, rather than as a means of transportation. However, these interpretations are only hypotheses derived from the interview transcripts in an exploratory manner regarding the factors that govern the choice of CSS. Therefore, in the following Study 2, we attempted to test and substantiate these hypotheses by analyzing the results of our previous survey of CSS users or those who have used CSS.

3.2. Study 2

3.2.1. Hypothesis Formulation

We tested the hypotheses set forth in the previous section by analyzing the results of the aforementioned survey, but first we confirmed the theoretically inferred hypotheses. Since CSS and private cars have an alternative relationship [1,9,18], if one does not own a private car, one is more likely to use CSS. Conversely, if they own a private car, they are less likely to use CSS. Therefore, we confirmed the following hypotheses. These hypotheses concerned the factors that govern the substitution between CSS and private cars.

Hypothesis 1 (H1).

If you do not own a private car, you are more likely to use CSS.

Hypothesis 2 (H2).

If you own your own car, you are less likely to use CSS.

A chi-square test of independence was calculated on 773 of the 15,913 monitors registered for CSS to determine the percentage of those who have used CSS among those who own a private car, used to own a car but gave it up, or never owned a car. A significant interaction was found (χ2 (2) = 15.27, p < 0.001). Those who used to own a private car but gave it up, or those who did not originally own a car, are more likely to use CSS (89.9% or 79.1%, respectively) than those who own a private car (77.8%). Conversely, those who own a private car are more likely to be registered with CSS but have never used it (22.2%) than those who owned a private car but gave it up or did not originally own a car (10.1% and 20.9%, respectively). Hypotheses 1 and 2 were therefore supported.

If the frequency of car use is high, private cars are likely to be selected, and if the frequency of car use is low, CSS is likely to be selected. These are the hypotheses inferred from the factor of frequency of use of the goods among the factors that generate substitution relationships among goods. The following hypothesis was confirmed by the following factors.

Hypothesis 3 (H3).

The number of times private cars are used is greater than the number of times CSS is used.

Of the 630 respondents who had used CSS, we asked how often they used CSS and how often they used their own car. The most frequent use of CSS was about one day per month, which was 23.3% of the total number of respondents and, in contrast, among the 353 respondents who owned a private car, the most frequent use of a private car was almost daily (34.3%). A summary t-test was conducted on the frequency of use for 630 CSS users and 353 private car users, and the results were statistically significant (t(981) = −20.032, p < 0.001). Given that Hypotheses 1 through 3 were supported, we confirmed that the survey data reflected the theoretically postulated hypotheses [17].

In this interview survey, it was understood that the presence or absence of a private car and the factors of whether the respondent focuses on matching various needs and whether he or she is highly or moderately committed to the product make a difference in consumption style. Hypotheses 1 and 2 were supported, indicating that respondents are more likely to choose or continue to use CSS if they do not own a private car. However, based on the characteristics of the subjects identified in Section 3.1.2, the usage purpose-oriented and waste-control types continue to use CSS despite owning a private car. This is contrary to Hypotheses 1 and 2, which were substantiated. Since it is theoretically assumed that subjects will continue to use CSS as a result of their needs being satisfied, it can be inferred that the factor of high or low level of commitment serves as an adjustment variable, moderating the impact of having a private car on the decision to continue to use CSS. Therefore, when we examined the statements made by subjects who were highly committed to the product, we assumed that the question items found in the questionnaire survey were characteristics of highly particular consumers, since statements regarding product attributes, such as status, were observed. If this tendency were high, consumers would be highly committed to the product, which would motivate them to use CSS even if they own a car. The following hypothesis was therefore formulated.

Hypothesis 4 (H4).

CSS users who own their own cars will increase their intention to continue using CSS when they place a higher value on status.

3.2.2. Quantitative Survey Analysis and Findings

The Study 2 survey asked 300 CSS users, or those who had used CSS, to rate their experience with CSS as part of their evaluation of the importance of status, meeting their needs, and their intention to continue to use CSS by asking them to rate the following statements: “It is important to get status and certification with the car”, “It is more important to get a car that meets my needs and service” and “I would like to continue to use CSS from this company”, using a 5-step Likert scale method (1 disagree, 5 agree). First, a multiple regression analysis was conducted to identify variables that provide for the intention to continue using CSS, including various control variables (age, gender, annual income, presence of children, and marital status) included in the questionnaire in addition to the variables of private car ownership, importance of status, and need matching. Although the regression equation was significant (F(8, 291) = 10.326, p < 0.001, R² = 0.221), the additional variables (age, gender, annual income, presence of children, and marital status) had no significant impact on the dependent measure (p > 0.5 respectively), and we did not include them in the following final model.

To test Hypothesis 4, another regression analysis was conducted using the data from these survey results with the presence of a private car as the independent variable (X), intention to continue using CSS as the dependent variable (Y), importance of status as the moderating variable (W) and need matching as the covariate (C). Model 1 of SPSS PROCESS version 4.1 (Hayes 2022) was used for the analysis [38] (Figure 2). This model contains the OLS regression coefficients along with their standard errors, t- and p-values, and 95% confidence intervals [38]. The regression equation results were R = 0.490, R² = 0.240, MSE(0.640), F(4, 295) = 23.295, p < 0.001, which were significant, and the coefficients of determination are shown in

Table 4. Test results of the moderation model of Figure 2.

	Coeff.	SE	t	p	LLCI	ULCI
Constant	2.549	0.259	9.843	0.000	2.039	3.059
Private Car (X)	−1.217	0.300	−4.056	0.000	−1.807	−0.626
Focus on Status (W)	−0.033	0.070	−0.475	0.635	−0.172	0.105
X × W	0.342	0.098	3.481	0.001	0.149	0.536
Focus on Various Needs (C)	0.350	0.055	6.364	0.000	0.242	0.458

. The covariance between each variable is sufficiently low that the results of the regression analysis are reliable (

Table 5. Covariance matrix of regression parameter estimates.

	Constant	Private Car (X)	Focus on Status (W)	X × W	Focus on Various Needs (C)
Constant	0.067	−0.049	−0.012	0.016	−0.009
Private Car (X)	−0.049	0.090	0.013	−0.028	0.003
Focus on Status (W)	−0.012	0.013	0.005	−0.005	−0.001
X × W	0.016	−0.028	−0.005	0.010	−0.001
Focus on Various Needs (C)	−0.009	0.003	−0.001	−0.001	0.003

). We also note that the model with the interaction of X × W provides the goodness of fit of the regression model with an R² of 0.240, which was increased by R² of 0.031 from the model without the interaction (

Table 6. Test of highest order unconditional interaction.

	R2 Change	F	df₁	df₂	p
X × W	0.031	12.114	1	295	0.001

). The goodness of fit, while not satisfactory, is acceptable because this model was intended to demonstrate H4 and did not explain all of the intentions for continued use of CSS.

For users who did not own a private car, a higher emphasis on status indicated a lower intention to continue using CSS, but this effect was offset by the significant interaction between the presence of a private car and emphasis on status, confirming that for users who owned a private car, a higher emphasis on status indicated a higher intention to continue using CSS than for users who did not own a private car. Hypothesis 4 was therefore supported.

Figure 3 shows the mean values of the mean and ±SD from the regression equation predicted by the moderator, with the focus on status (W), for the intention to continue using CSS (Y). This figure shows that when the focus on status (W) was relatively low, the presence of private car ownership (X) had a significant effect on the intention to continue using CSS (Y). Users who owned a private car when the importance of status was relatively low had a significantly lower intention to continue using CSS than users who did not own a private car. On the other hand, users who owned a private car with a relatively high emphasis on status had a higher intention to continue using CSS than users who did not own a car, but the effect was not significant. Figure 3 shows that users who did not own a private car had a relatively higher intention to continue using CSS, which was consistent with Hypotheses 1 and 2 and also consistent with the supported results. In other words, it was confirmed that consumers with strong car preferences have the same or an even higher intention to continue using CSS than consumers who do not own their own cars. Thus, the hypothesis that consumers with strong car preferences are motivated to use CSS even if they own a private car, as assumed from the interview results, was supported.

As Figure 3 shows, users who were relatively status-conscious and owned their own cars had the highest intention to continue using CSS. This corresponded to the waste-control type among the four types classified in Study 1. However, after checking the focus on status as a moderator value defining Johnson–Neyman significance regions, we found that the focus on status was relatively higher and that owning a private car had a significant positive effect on the intention to continue to use CSS. It should be noted that only about 5% of subjects were in the significance region at values exceeding +SD from the mean of the moderator by focus on status.

As the goodness of fit of the model indicates, the intention to continue using CSS cannot be explained solely by an emphasis on status and matching needs; user evaluations, such as quality ratings and satisfaction with the CSS use experience, are also possible factors that influence consumer characteristics on the intention to continue using CSS, but in this empirical study, not included in the model as a limitation. In future research, it is required to model and empirically study the factors that influence consumer characteristics on the decision to use CSS.

4. Discussion

Although prior research has proposed that new modes of consumption, access, and sharing, as alternatives to ownership, fall into one of two polar continua, pure ownership and pure sharing, there is a need to propose conditions for classifying individual cases. For example, as a result of interviews with users of a U.S. car-sharing service (Zipcar), Bardhi and Eckhardt (2012) [1] identified the access-based consumption in the context of car-sharing as a lack of identification, varying significance of use and sign value, negative reciprocity resulting in a big-brother model of governance, and a deterrence of brand community. As a result of an empirical study of Zipcar users in the U.S., Lamberton and Rose (2012) [18] found that in the car-sharing context, transaction utility, the degree of substitutability, knowledge of sharing, technology cost, mobility utility and consumers’ perceived risk of product scarcity are determinants that influence the likelihood of participating in Zipcar. Similarly, as a result of an interview survey and empirical study of users of Zipcar in the U.S., Habibi, Kim, and Laroche (2016) [3] found that transaction utility of sharing, degree of substitutability and materialism were factors that positively influence the likelihood to participate in Zipcar. In contrast to these previous studies, this study, although based on the analysis of the survey of CSS users in Japan, attempted to understand the new form of consumption, different from ownership, by identifying the consumer characteristics that influence the intention to use CSS from the perspective of consumer behavior.

In Study 1, based on an exploratory analysis of the statements made in the qualitative group interviews, we proposed that consumers who choose or continue to use CSS can be categorized into four types by typifying the factors that influence their decision making: saving-oriented, minimalist, usage purpose-oriented, and waste-control type. In Study 2, we analyzed the results of a quantitative survey to test the hypotheses formulated in Study 1, i.e., that consumers who own and are more committed to their cars are more motivated to use CSS.

Among the four proposed types, the saving-oriented and minimalist types use CSS as a substitute for private cars, supporting the arguments of previous studies. On the other hand, the usage purpose-oriented and waste-control types use CSS not as a substitute for private cars, but to supplement their desires that cannot be satisfied by a single private car, i.e., their commitment to cars and diverse needs, that may represent new conditions to classify consumers who choose new modes of consumption. However, these types may also be using CSS as an alternative to a second or third private car [14].

Understanding the attributes of these four types of consumers provides a foundation for companies to implement marketing strategies to respond to the evolution of the sharing economy. CSS firms should seek out and expand their services to customers classified as the waste-control type, who have the highest intention to continue using their products and services, as well as to customers who focus on the usage purpose. For example, the waste-control type could be served by offering access to a brand of car that is conducive to status, or the usage purpose-oriented type could be served by offering access to different types of cars depending on their purpose, and so on. The service content could be customized for each type of user. In another paper, the authors proposed that the effect of the availability of a private car on the frequency of CSS use is moderated by the usage purpose of the CSS [17]. However, after confirming the size of each type of user, an appropriate additional fee would be required to cover the cost of the expanded service and still make a profit.

The saving-oriented and minimalist types were classified by 11 of the 16 interview subjects as necessary consumers and customers, and the key issue is how to keep them. The saving-oriented type requires service improvements in response to customers’ evaluation of service quality, which will be discussed later, in order to avoid unfulfilled needs and withdrawal from the service. On the other hand, the minimalist type may carefully consider purchasing a private car as the frequency of use of CSS increases, so if a car manufacturer operates CSS, a program that leads customers to purchase their own car may be effective. Engaging users of their own vehicles could help them produce more efficiently. By classifying customers into four types, CSS companies can consider marketing strategies to determine which type of service to offer to customers to encourage them to continue using the service, or how to select potential customers by classifying them into this category.

The interviews also confirmed statements related to holding costs, uncertainty of use, and the number of times the goods are used as factors that create a substitution relationship between material goods and the usage rights of material goods, as well as statements related to quality [27], such as smell and dirt, as factors that determine continued use. Service quality affects satisfaction as a precondition [39], and it is generally agreed that service quality and higher customer satisfaction are positively correlated with firm performance [40]. Therefore, we cannot overlook these factors, which have been proposed in previous studies, as regulating the choice and continued use of CSS.

However, the findings in this paper are the result of analyzing only survey data on CSS users or those who have used the service in Japan and is not necessarily applicable to other sharing services in other countries in the same way. For future research, we will use the proposal in this paper as a starting point to model hypotheses about the factors that govern the decision to select and continue using sharing services, and to test these hypotheses, we will conduct a survey of a wider range of subjects besides CSS users. We will then attempt to generalize the factors that motivate consumers to choose and continue to use sharing services.

Notwithstanding above, this paper provides a thought-provoking look at the diffusion of CSS in Japan. Because of the relatively well-developed public transportation system in Japan, many people can live without a passenger car, and we believe that the branching of the data that includes such consumers reflects a wider range of consumer choices and shows scientifically informative results. We also proposed how to understand consumers who choose a new mode of consumption different from ownership by typifying the factors that influence consumers’ decision to use the service from the viewpoint of consumer behavior. This proposal may neither be limited to the Japanese market nor for CSS alone, but also be an attempt to extend the traditional marketing and consumer behavior research that has been developed for ownership and understand the new modes of consumption emerging in other countries as well.

5. Conclusions

The main element of the CSS offering is the usage rights of material goods, e.g., automobiles, but it is not the only one; usage rights of information on websites and human services at call centers also constitute an important part of the offering as ancillary goods. The access model should be extended to cover not only the usage rights of material goods but also other goods, e.g., information and human services [1]. Each type of consumer evaluates not only the quality of cars, but also the information systems and services. Given that their quality evaluations and decisions evolve into the effective use of sustainable resources, it is essential to understand consumer behavior in choosing new modes of consumption.

In this study, none of the subjects in the interviews expressed the intention that the act of sharing would solve environmental problems. The reasons for use were based on convenience and ease of self-determination, and consumers were not aware of the positive outcomes for the environment that would result. This indicated that consumers must first be motivated by convenience to participate in the sharing economy, which is consistent with [1,3,18]. We also observed that consumers who are highly committed to their products actively use CSS. This result differed from previous studies. Thus, consumers are not using CSS with the intention of contributing to environmental issues. However, it is important to categorize and understand the factors that influence the decision to choose or continue to use CSS, as proposed in this study, as this can help CSS companies’ marketing strategy to respond to the sharing economy, leading to more resource sharing and, consequently, to the promotion of “Use-oriented” and “Results-oriented” PSS.