Keywords

1 Background and Motivation

Originally developing from redistribution of idle resources, the sharing economy has expanded rapidly since its emergence. It enables individuals to use resources at a lower cost, while resource owners can more or less receive some benefits in return. The sharing economy functions through the concept of utilizing an online third-party platform to temporarily transfer, and further enable resources to be shared and reused, and this mechanism can be seen in examples of sharing of scooters for public and vacant rooms for tourists. The sharing economy enhances the efficiency of asset utilization through paid rentals of surplus resources, creating a win–win situation between users and resource owners, and the subversive business opportunities brought by its business model attract global attention. The overall profit generated by the sharing economy is estimated to be around $40.2 billion in 2022, with at least $18.6 billion estimated for platform providers in 2017, as stated in Juniper’s Sharing Economy: Opportunities, Impacts & Disruptors 2017–2022.

Sharing economy business is highly integrated with mobile applications that can be used anytime and anywhere, which further impacts traditional B2C and B2B2C service models. Because the third-party platform acts as a matchmaker and guarantees the supplies quality by crowdsourcing mechanism, the concern of risks derived from this condition has great influence on realizing the transactions between supply and demand sides. As illustrated in the Transaction Cost Theory (TCT), the concept of transaction cost explains how both parties of a transaction consider risk will determine the realization of transaction, which means that when transactions are under high uncertainty and complexity, both transaction parties will tend to be more conservative, resulting in the failure of transaction. Besides, for the sharing economy mobile services (M-services), most processes are based on mobile networks. Everything from identity verification, transaction records and payment procedures will all be documented by platform operators. With the possibilities of individual data being disclosed for usage, users may adopt an even stricter approach to assess the risk of the sharing economy M-services. In fact, for users, their adoption of sharing economy M-services is supported by reduced transaction costs. That is to say, users are willing to accept the M-services if the operator can provide easy access to the information of the sharing economy M-services and make sure high quality and on time delivery of their orders with a guarantee of the enforcement of the transaction contract through mobile network connectivity.

On the other hand, the Signaling Theory proposed by Spence (1973) stated that if online transaction markets are not mature enough, serious information asymmetry will occur between the transaction parties, and the delivery of information quality will have certain degree of influence. That is to say, even under the condition of information asymmetry, if sellers can establish a high quality signal through some form of activity to rationalize the compensation buyers get from the activity and further induces buyers’ willingness to make additional payments, a transaction can be realized (Lankton and McKnight 2011). For example, when the sharing economy market is still at the early development stage, its M-service has an unstable price-performance ratio because users have lower degree of familiarity with it, which makes price and quality become the major factors of realizing the transaction. In other words, when users’ concern for risks outweighs the use value of the sharing economy M-services, users’ early adoption intention may be affected, and they may even start to slowly reject the continuous adoption of that services. This also means that delivery of value can help improve users’ level of tolerance for risk. The concept of perceived value is also introduced by Davis (1989) in his Technology Acceptance Model (TAM), which has been widely applied by many researchers in the information systems (IS) setting. The common conclusion is that the willingness to adopt IS services for buyers who tend to avoid a high level of uncertainty might be enhanced through their perceived value of how relevant features can improve performance, even though they are likely to consider IS as risky and uncertain (Perez-Alvarez and Paterson 2014). In short, perceived value and perceived price are two factors that may affect the influence state of perceived risk on adoption behavior.

Understanding the engagement and behavior of users is important for the success of businesses offering sharing economy M-services. Especially sharing economy M-services still in the early stages of development, knowing about the motivation of users will enable businesses to build a more detailed user profiling, devise more precise marketing strategies and design adaptive services. However, related research is still at a premature stage in the sharing economy M-services field. Even though existing literature has looked at the impact of perceived price, perceived value, and perceived risk on users’ adoption of various IS services, the discussion about how these three factors influence users on adopting IS in the sharing economy is still rather scarce. Hence, this study focuses on integrating the theories on perceived price, value, and risk to establish a relational sharing economy M-services adoption model to analyze the important motives for users to continue to adopt services by including contingencies in which low and high user experience existed.

In the following sections, we first explore how past research and theories look at perceived risk, perceived value, perceived price, and user adoption intention. Next, we propose possible hypotheses and illustrate the methodology used in this research, followed by the results of this study. Finally, we conclude this paper with a discussion section identifying the implications and limitations of our findings as well as possible directions for future research.

2 Theoretical Background

2.1 Perceived Risk

TCT was proposed by Coase (1937). It emphasizes that the realization of a transaction is determined by consideration of risks, which means when a transaction is under high uncertainty and high complexity, it’s difficult to draw up a long term contract that includes all possible situations, and hence leads to market failure. Williamson (1975, 1979) later expanded Coase’s original theoretical framework and proposed that market and vendors are the mechanisms that realize transactions. That is to say, when transaction cost doesn’t exist or is very low, an economic agent will prefer market governance. On the contrary, if the cost is higher than the costs and benefits of production in the market, the economic agent will prefer internal organization. Hence, market failure explains why vendors exist. Market failure is not only based on behavioral assumptions (i.e., bounded rationality and opportunism), but is also influenced by the specificities of transactions, including asset specificity, uncertainty, and frequency of transaction.

Although TCT was first used to discuss market and vendor governance, more and more research later also explains user behavior based on TCT and consider user behavior as an issue of exchang between sellers and buyers. In addition, uncertainty is considered as a main specificity of transactions influencing users’ strategic decision making in the realm of strategic management research (e.g., Belkhamza and Wafa 2014; Castaño et al. 2008; Maduku 2014; Perez-Alvarez and Paterson 2014; Teo and Yu 2005). Because of the breakthrough guarantee mechanism from crowdsourcing and asymmetry of information towards sharing economy M-services, it’s difficult to predict the behaviors of the other transaction party, and further results in uncertainty. Users’ transaction costs may escalate with the increase of uncertainty because users will be required to be engaged in time-consuming activities such as gathering information about vendors, products and services as well as paying attention to transaction process (Teo et al. 2004). Especially at the early stage of the sharing economy M-services, information asymmetry means that users need to take on more uncertainty. Hence, when information delivery and communication fail to efficiently reduce users’ concerns of uncertainty, the possibility of realizing the sharing economy M-services will dramatically decrease.

2.2 Perceived Value

Users’ motivation of behavior is mainly determined by their perception of benefits. Especially when users are more unfamiliar with new services or products, less experienced, or lack of knowledge, subjective value judgment becomes the basis of users’ decision making. The theoretical base most commonly adopted by scholars is the TAM introduced by Davis (1989), which mainly explains users’ behavioral model of adopting new IS using the ideas of perceived usefulness and perceived ease of use. Perceived usefulness refers to the subjective perception of users have in believing their work efficiency can be enhanced if they use certain systems. Especially in IS services, technology functions need to fulfill the task users expect from them, and provide users with appropriate feedback and assistance through operating with accuracy continuously (Lankton and McKnight 2011). Perceived ease of use refers to the degree to which an individual believes that using a particular IS would be free of effort. Thus, when users consider that system to be easy to use, they will believe it can improve their efficiency.

In recent years, many researchers (e.g., Godoe and Johansen 2012; Rahman and Sloan 2017) pointed out that perceived usefulness can be the key factor that leads to differentiation, and is often proved to be the most important variable in the model, while perceived ease of use has an inconsistent conclusion. Therefore, this study considered perceived usefulness as the crucial element constituting perceived value.

2.3 Perceived Price

In the context of information quality in Signaling Theory, perceived price can be interpreted as the important element. This is especially true in an immature online service. In this scenario, in addition to sending signals to users to enhance their confidence through marketing strategies such as return and exchange guarantee, privacy protection, and branding, the keys to more effectively and directly impact transactions under information asymmetry are building signals that convey a reasonable price-performance ratio of the product or service to make users agree on the product or service quality reflected in its price, or even be willing to pay a higher price to obtain this product or service. There are two paradigms that define perceived price. The first paradigm is the objective monetary price, referring to the actual cost of buyers’ purchase (Dillon and Reif 2004). The second paradigm is the encoded price of the purchase perceived by buyers. Because online services continue to adopt new pricing strategies for matching and selling services, users’ perceived price is vital for sharing economy M-services. If sellers implement a good price matching strategy, there would be a decrease in competition for price because competitors will fail to gain profits from lowering prices, which is similar to well-designed price-matching policies (Srivastava and Lurie 2001).

3 Research Hypotheses

3.1 The Effect of Perceived Risk on Adoption Intention

In accordance with TCT, this study used uncertainty to explain how users evaluate the risk of adopting sharing economy M-services. A large amount of research employing TCT stresses how uncertainty impacts decisions with regard to the scope of the enterprise, focusing on the decision of online services. For example, Teo and Yu (2005) used TCT to propose a model for explaining users’ online purchase behavior and conducted empirical testing with Singaporean users. Their study found that perceived transaction cost could be used to explains users’ willingness to make online purchases, and can be explained by uncertainty, dependability and purchase frequency. In other words, users are more likely to make online purchases if they consider the online store to have higher dependability, see online shopping as having lower uncertainty, and have more online experiences. Utilizing data from users engaged in e-banking services offered by four main retail banks in South Africa, Maduku (2014) also found out that the most important determinant for users to adopt mobile banking and online services is their trust in the e-banking system, which implies that reducing users’ uncertainty in these systems can speed up their adoption of online and M-services on a larger scale.

Many IS researchers define uncertainty as a multi-faceted structure and favor the elements in examining uncertainty. Cunningham (1967) originally proposed six uncertainty dimensions to form the perceived risk: (1) performance, (2) financial, (3) opportunity/time, (4) safety, (5) social and (6) psychological loss. A rich stream of literature supports the usage of this measurement to understand e-service evaluations and adoption (e.g., Featherman and Pavlou 2003; Luo et al. 2010). According to Grewal et al. (1994), performance uncertainty is posited to be the most critical because it presents the overall possibility of ‘‘the product malfunctioning and not performing as it was designed and advertised and therefore failing to deliver the desired benefits.” Their research adopted the performance uncertainty and further advanced the study on perceived risk by including security/privacy uncertainty. Luo et al. (2010) stated that transaction security/privacy uncertainty has to be considered in e-service investigation. The security/privacy uncertainty is defined as “potential loss of control over personal information” by Featherman and Pavlou (2003). Obviously, the concern about the security/privacy is a threat that is often theorized as a cause of user’s self-protective behavior, and this argument has been evidenced by many other e-service research. Hence, this study proposes the following hypotheses:

Hypothesis 1:

Users with high “performance uncertainty” will perceive low adoption intention toward sharing economy M-services.

Hypothesis 2:

Users with high “security/privacy uncertainty” will perceive low adoption intention toward sharing economy M-services.

3.2 The Effect of Perceived Value on Adoption Intention

The TAM is a popular IS theory that models how users come to accept and use an IS service. The TAM has been continuously studied and expanded because it has been evidenced with strong behavioral elements and assumes that when someone forms an intention to act, they will be free to act without limitation. The TAM suggests that when users are presented with a new IS service, two factors influence their decision about how and when they will use it are perceived usefulness and perceived ease of use. Especially, a large number of studies stress the importance of the role perceived usefulness occupies in predicting the IS services usage and acceptance of users. Using a model built on TAM and TRI, Godoe and Johansen (2012) analyzed the reason users reject a system even if they are positive toward technology in general, and concluded that a low degree of perceived usefulness and perceived ease of use could contribute to users’ rejection of a system in spite of their optimistic attitude toward the technology. Additionally, in their study conducted in Bangladesh, Rahman and Sloan (2017) also identified perceived usefulness to be the determinant factor of users’ adoption of mobile commerce, and concluded that companies should continue to optimize their technologies and services to facilitate users in meeting the demands of contemporary ever-changing lifestyle. Hence, this study proposes the following hypothesis:

Hypothesis 3:

Users with high “perceived usefulness” will perceive high adoption intention toward sharing economy M-services.

3.3 The Effect of Perceived Price on Adoption Intention

Perceived price is pointed out in several studies focusing on how it influences user behavior. For instance, Varki and Colgate (2001) investigated the relation between price, user behavioral intentions, and perceived value in the finance industry and found out that perceived price is the major determinant of value perceptions and has a stronger influence on user behavioral intentions, more than their mediated outcome via perceived value. Jiang and Rosenbloom (2005) also analyzed how satisfaction and price influence customer retention in various stages, and found that online retailers should focus on the quality of post-delivery services and improving customers’ price perceptions to attain a higher customer retention rate.

Moreover, looking at how perceived price affects customers’ decision on buying mobile communication services in the mobile and combined customer segment, Munnukka (2008) reached a conclusion that customers’ buying intentions and perceived price are positively and significantly related. Another study exploring the influence of fairness and transparency of price on customer behavior by Rothenberger (2015) also discovered that customers will consider their offer to be superior if they perceive higher price fairness from the clarity of information they receive about product or service price. This results in a more positive attitude and satisfactory reaction of the customers about recommending or repurchasing the services or products. In conclusion, perceived price is considered as an important factor impacting buyers’ decision making and a valid equivalence of economic expenditure that customers need to sacrifice to participate in a transaction of purchase. Hence, this study proposes the following hypothesis:

Hypothesis 4:

Users with positive “perceived price” will perceive high adoption intention toward sharing economy M-services.

3.4 The Relationship Between Perceived Risk, Value, and Price

Businesses need to enhance consumers’ level of confidence and assist them with reducing their concerns about uncertainty to ease their anxiety when making transactions through online third-party platforms. In their cross-level research that investigated how national culture and social institutions relate to the way users perceive the usefulness of the IS services, Parboteeah et al. (2005) found that users’ perceived usefulness is significantly influenced by the degree of social inequality, industrialization, masculinity, and avoidance of uncertainty. To understand how online user behaved when they adopted e-commerce service, Babin et al. (2009) discovered that the intangible quality of online transaction is what contributes to online users’ anxiety. Hence, uncertainty decreases the usefulness users attain from online shopping. They also found that uncertainty has a significant influence on how users perceive the value of the way their needs of information are satisfied, which means that how users perceive the value of standards is decided by the cause of uncertainty. Ntsafack et al. (2018) also pointed out in their research that well-structured services are needed to decrease the level of anxiety in a culture that is characterized with high uncertainty avoidance, which could be explained by the impact of uncertainty on reducing users’ perceived usefulness and decreasing users’ adoption intention. Hence, this study proposes the following hypotheses:

Hypothesis 5:

Users with high “performance uncertainty” will negatively affect “perceived usefulness” toward sharing economy M-services.

Hypothesis 6:

Users with high “security/privacy uncertainty” will negatively affect “perceived usefulness” toward sharing economy M-services.

Previous studies considered risk as influential for consumers’ price perception. For instance, Tellis and Gaeth (1990) suggested that if consumers have a higher level of uncertainty about product quality, they are likely to be price-seeking, suggesting under-weighting of price, or price-averse, suggesting over-weighting of price. This implies that the accessibility of quality information will influence consumers’ sensitivity of price. Recent research by Demirgüneş (2015) also pointed out that the perceived risk of consumers has vital impacts on their willingness to pay a higher price, and concluded that there’s a smaller chance for consumers to pay more for the service or product if they perceived uncertainties with it. They also proposed that offering consumers more information will enable enterprises to decrease their perceived uncertainty and thus lead to a higher profit. Hence, this study proposes the following hypotheses:

Hypothesis 7:

Users with high “performance uncertainty” will negatively affect “perceived price” toward sharing economy M-services.

Hypothesis 8:

Users with high “security/privacy uncertainty” will negatively affect “perceived price” toward sharing economy M-services.

4 Research Results

4.1 Measurement and Psychometric Properties

The objective of this study was to understand user’s adoption behavior from three theories: perceived risk (measured by security/privacy uncertainty (SUN) and performance uncertainty (PUN)), perceived value (measured by perceived usefulness (PU)), and perceived price (PP) in the context of sharing economy M-services. A survey was conducted to obtain primary data with the survey questionnaire comprising 17 items measured on a 7-point Likert scale (where 1 = extremely disagree, 2 = disagree, 3 = somewhat disagree, 4 = neither disagree nor agree, 5 = somewhat agree, 6 = agree and 7 = extremely agree). The data was collected from Taiwan. The online sampling method was used, and a total of 509 questionnaires were completed and returned.

In this study, the multi-item scale was used to assess the various constructs. According to the judgment criteria proposed by Hair et al. (2006), the internal consistency of latent constructs, the factor loading of each observed item, and the average extracted variance (AVE) are the main methods used to evaluate reliability and validity. Specifically, Cronbach’ α and composite reliability (CR) are used to represent internal consistency of latent constructs. A higher composite reliability value means that true variance accounts for a higher percentage of the total variance, with its value suggested to be above 0.6. The square of the factor loading of each observed item can be used to represent the explanatory power latent constructs have on the observed item, and is the basis of evaluating convergent validity, with its value suggested to be above 0.7. AVE is the total explanatory power the latent constructs have on the observed items. According to Fornell and Larcker (1981), if the AVE value presents the degree of association in latent constructs to be above 0.5 and higher than the degree of association between latent constructs, it indicates the discriminant validity of measured variables. Finally, the multi-group SEM approach was adopted to examine the moderating effects of contingencies of different user experiences.

Table 1 shows the results of reliability statistics. In terms of reliability, the values of Cronbach’s α for the latent constructs were between 0.688 and 0.882 with all the values of CR above 0.69. As for validity, the samples had a factor loading above 0.6 between each measured variable, while all the AVE value were above 0.5 and larger than the square of correlation coefficients between other latent constructs, showing the convergent and discriminant validity of this study.

Table 1. Measurement and reliability statistics
Full size table

4.2 Results of Structural Equation Modeling

Fit of the Full Structural Model

First of all, clustering analysis was used to classify the samples into groups with different user experience levels for contingency analysis. Based on the k-means method, the whole sample was classified into two groups: low- and high-experienced users, and the F-value revealed that there were significant differences among the two groups on attribute scores. Next, this study adopted the structural equation modeling (SEM) to estimate the structural model using the maximum likelihood (ML) method. The results showed that the chi-square statistics for low- and high-experienced users were 297.81 (df = 110) and 243.99 (df = 110) separately, and the p-values were less than 0.001, so the model failed to fit in an absolute sense. However, because the chi-square test was known for its sensitivity to sample size (Hair et al. 2006), even a good fitting model (i.e., only small discrepancies between observed and predicted covariance) could be rejected. Thus, researchers recommend complementing chi-square with other goodness-of-fit measures. As noted previously, GFI = 0.86 and 0.91, CFI = 0.86 and 0.91 met the 0.85 cutoff, and the point estimate of RMSEA = 0.088 and 0.065 was less than 0.1. Also, the parsimonious fit measure χ2/df = 2.71 and 2.22 were under the recommended threshold limits (< 5) for this measure (Jöreskog 1970). Thus, the overall proposed structural model was sufficiently supported.

Hypotheses Testing

This study adopted the multi-group SEM to test the direct and indirect effects of antecedent variables (SUN, PUN, PU, and PP) on adoption intention of sharing economy M-services in different user experiences contingencies. Figure 1 – (1) indicated the path coefficients and their significance for each hypothesis. Overall, two out of the eight hypotheses were significant by low-experienced users: the direct effects of PUN and SUN on adoption intention were nonsignificant, while the main construct - PU (0.67, p-value < 0.01) and PP (0.29, p-value < 0.01) had positive direct effect on adoption intention; therefore, H3 and H4 was supported but H1 and H2 were not supported. Moreover, all the effects of PUN and SUN on PU and PP were nonsignificant. Thus, the results didn’t support PU and PP to be the mediator of the structural model. Therefore, H5, H6, H7, and H8 were not supported.

Fig. 1.
figure 1

Multi-group structural equation modeling results

Full size image

However, it is worth noting that the high- and low-experienced users had great differences in structural relationships. Figure 1 – (2) showed that six out of the eight hypotheses were significant by high-experienced users: the direct effects of PUN, SUN, and PP on adoption intention were nonsignificant, while PU (0.81, p-value < 0.01) had positive direct effect on adoption intention; therefore, H3 was supported but H1, H2, and H4 were not supported. It was interesting to find that the effect of PUN on PU (0.73, p-value < 0.01) being positive while the effect of SUN on PU (−0.71, p-value < 0.01) being negative. Conversely, the effect of PUN on PP (−0.75, p-value < 0.01) was negative while the effect of SUN on PP (0.69, p-value < 0.01) was positive. Thus, the results supported mediated roles of PU on the relationship between PUN, SUN, and adoption intention. However, because of the negative hypotheses of PUN and SUN on PU and PP, H6 and H7 were supported, but H5 and H8 were not supported.

5 Conclusion

Management and Practical Implication

The sharing economy opened up new ways of accessing goods and services in sustainable ways. It required a new level of thinking and creativity, disrupting traditional habits of doing things. Even though the concepts of perceived risk, value, and price have long been central components of a number of theories of user behavior, there is only a scarce amount of literature focusing on the adoption of sharing economy M-services at this stage. Moreover, there is no literature to further observe what changes will occur in the interrelationships between perceived risk, value, price towards adoption intention under different contingencies. Therefore, the contribution of this study is to identify how contingencies formed by user experiences moderate the structural model.

The result of this study indicated that the mediation of PU remains a significant concern in the sharing economy M-services. Users’ PU is a major driver to the widespread adoption for both low and high-experienced users. This finding is consistent with previous researches (e.g., Huang and Wilkinson 2014; Mou et al. 2017). Not surprisingly, if the enterprise provides service to their customers without having to inspect everything customers get strictly every time, the whole system would come to a screeching halt. Enterprises must find ways delivering the direct benefit to users. This concept is even more critical in the sharing economy because users tend to hold a more conservative attitude of specific utility for new services implemented in the early stage, shared values may become a foundation for increasing the likeliness of success in creating commitment.

Secondly, the result of our study showed that SUN is the main barrier for PU and PUN is the main barrier for PP. Obviously, given the sensitive nature of the information shared when using online services, the adoption of these services are impeded by consumers concern about the overall security of the technology (Cope et al. 2013). Sharing economy M-services providers therefore need to ensure that online systems are sound with state-of-the-art safety management in place to minimize potential risks that end-users may be exposed to. To decrease users’ SUN in sharing economy M-services, enterprises will also need to comply with legal structures to provide transparent guarantees if any dispute occurs. These risk control activities should be part of a communication strategy aimed at instilling trust in users regarding the services (Maduku 2014). On the other hand, the results showed that the higher the PUN, the lower the PP. This demonstrated that compared to the security uncertainty caused by technology, consumers believe that the uncertainty of supplies quality is more controllable by suppliers, and therefore, when PUN is increased, the consumer’s tolerance for price is lower.

The third observation in this study was that there is a distinct discrepancy in the price judgment of sharing economy M-services between low and high-experienced users. For the low-experienced users, the price is a driver to determine user’s intention towards adopting sharing economy M-services. In comparison, when user experience is increased to a certain level, the consideration of price ceases to significantly influence users’ adoption behavior. The practical implications of the findings refer to the fact that the use of pricing schemes needs to not only take service life cycle stages into consideration, but also enhances users’ judgment of price reasonableness through delivering signals of service value to prevent pricing schemes from becoming the obstacle of adoption behavior. In their study of pricing objectives over the service life cycle, Avlonitis et al. (2005) also pointed out a similar concept and considered the stage of services’ life cycle and the sector of operation to have an influence on the pricing objectives pursued. Managers could have much to gain by taking a “situation specific approach” when they set prices. Therefore, various objectives of pricing need to be determined when a service moves from one life cycle stage to the next, and different services also require a different pricing strategy.

Limitations

Our results and findings should be interpreted in light of the following limitations that are inherent in this research. First, the survey respondents in this study are 20–40 age group; hence, caution should be exercised in generalizing these results to the general population. Second, the research model is designed to discuss the influences on perceived risk, value, and price in adopting sharing economy M-services by using Taiwan users. Future research can further explore the cross-culture investigation to obtain insights for users’ continuous adoption.