2.1 Mobile Payment

Mobile payment is considered a revolutionary technology that transforms consumer buying processes as well as business models of providers. Several authors such as Au and Kauffman [2] and Kousaridas, et al [29] agree on the idea that mobile payment involves buying goods or services by using a mobile device inside a mobile communication platform. Therefore, two main elements constitute the definition of mobile payment: 1) the mobile technology or service to conduct the payment and 2) the network or platform needed for the payment to take place.

With regard to the first element, several technologies for mobile payment can be used. For example, Zhou [73] underline that mobile phones are the tool to conduct mobile payment, whereas Dewan and Chen [15] extend the domain to include personal digital assistants (PDA) or radio frequency (RF) devices. Cheng and Chang [8] pay attention to the potential benefits of NFC technology. In addition, mobile payment studies have been applied to several domains: mobile ticketing [37], mobile Internet payments [38], and mobile payment tools in Virtual Social Networks [32].

The second element that defines mobile payment is related to the platform or network through which the payment takes place. Pousttchi [46] state that in order to initiate, authorize, or complete the payment, mobile devices should be used in conjunction with mobile communication techniques. Therefore, as stated by Dahlberg, et al. [12], firms must take advantage of wireless and other communication technologies. Miao and Jayakar [39] explain that different mobile platforms coexist in the market, offered by (1) mobile network operators, (2) banks (or financial institutions), (3) third-party platforms, and (4) hybrid models from strategic alliances of banks and mobile operators. These platforms coexist as there are several options for where to position the secure element that authenticates mobile payment users: in the SIM card of the device, on the motherboard of the device or in a cloud emulator [13]. Consequently, players like mobile network operators, device manufacturers, banks, card schemes and over-the-top providers are all offering alternative solutions for mobile payment, and struggle to come up with joint solutions [41]. There are several examples of these suggested platforms in the market, and some providers have created their own mobile payment platform, becoming independent from financial institutions [4].

In summary, as explained by Pousttchi [46], these interdependent factors (technical, human and market) should be combined adequately in order to be successful in any mobile payment process. For example, [22] argues that, based on technical factors, limited availability of NFC initially slowed down the adoption process of mobile payment. In this paper, we combine multiple theories for analyzing the determining factors of intention to use mobile payments: value (perceived value approach), facilitating conditions (the Unified Theory of Acceptance and Use of Technology, UTAUT) and learning costs (in the broader discourse on switching costs).

2.2 Perceived Functional Value of Mobile Payment

The perceived value approach is rooted in the field of economics and utility theories. Perceived value is the consumer’s overall assessment of the utility of a product/service based on perceptions of what is received and what is given [71]. Value has functional and emotional dimensions. Functional value refers to the utility resulting from quality perception and performance expectation, while emotional value is defined as the utility derived from feelings or affective states that a product generates [62]. [24] develop a Value-based adoption model of mobile Internet, in which emotional value is related to enjoyment and functional value relates to usefulness. Some disagreement exists in how perceived value should be proxied in the mobile literature. Three alternatives can be identified: as diverse individual dimensions [52], as a second-order construct made of several value components [10], or as a unidimensional concept [70]. Based on the different dimensions of perceived value, and the intercorrelations among the different dimensions of perceived value [57], this paper focuses on perceived functional value.

Perceived value of mobile payment services is relevant on its functional dimension. Mobile payment is an advanced mobile service mainly adopted for its multiple functionalities [23] [53] in the first place, not for its perceived enjoyment or emotional value. Like other mobile transaction services that are more functional and less categorized as mobile entertainment [25], mobile payment is considered here to have functional rather than emotional value. Perceived value is closely related to consumer’s evaluations of what she/he receives and sacrifices [42] and has traditionally been defined as the trade-off between benefits and costs, i.e. between what the user gets and gives when using a service [71]. Cost-benefit theory has also been applied to mobile payment research, although most studies only consider monetary perceived costs and risk [5]. In their pioneering paper on perceived value of mobile channel, [25] conclude that consumers focus more on the costs of the mobile channel than on its benefits. More recent conceptualizations of value support the idea that value is more than just a comparison between costs and benefits and acknowledges that value is situational and can change before, during, or after an experience [71]. Moreover, additional costs may arise when considering specific mobile services [35]. For mobile Internet, [24] propose technicality of the mobile service to be a cost facet of perceived value. That might be also the case for mobile payment services. Following prior research, the present paper adopts a cost-benefit approach and analyses learning costs (cognitive effort or non-monetary costs), facilitating conditions and perceived functional value of mobile payment.

Perceived functional value has been shown to have a positive effect on the behavioural intention to use a new technology, for instance regarding mobile location-based services [48], SMS [62], mobile internet [24], mobile payment [70], mobile channel [25], mobile coupons [34], and mobile services [52]. In a similar way, some studies have focused on the role of perceived functional value of a new mobile technology on explaining the intention to pay for it, purchase intention [20], commitment [48], word of mouth [20] or attitude toward those services. Mobile payment is a new technology for many people. It is crucial that users perceive a positive overall evaluation to be motivated to use this and more importantly to balance the set of features and functionalities [55]. In line with above discussion, we posit that perceived functional value of mobile payment might have a positive influence on the user´s intention to use it:

H1: Perceived functional value has a positive effect on the behavioural intention to use mobile payment services.

2.3 Facilitating Conditions of Mobile Payment

Facilitating conditions are defined as “the degree to which an individual believes that an organizational and technical infrastructure exists to support use of the system” [63]. This concept is rooted in the Perceived Behavioural Control and Innovation diffusion theories and a major predictor of the acceptance of new technologies in the Unified Theory of Acceptance and Use of Technology (UTAUT) [63]. Prior research following cost-benefit theory sees facilitating conditions as a potential reducer of sacrifice component of perceived functional value [24] and a construct made of response cost and self-efficacy [56].

In the mobile context, the concept of facilitating conditions is seen as a very relevant determinant of mobile services use, thus playing an important role in the acceptance of mobile payment. Following UTAUT, most studies focus on the impact of facilitating conditions on actual use of a new technology [63]. Some recent articles, though, have modified UTAUT and started to analyse the direct effect of facilitating conditions on behavioural intention to use specific mobile services [60].

In the last decade, many studies have analysed the direct effect of facilitating conditions on intention to use specific mobile services. For instance, prior research on mobile payment has found a positive effect of facilitating conditions on mobile shopping [69], mobile advertising [40] and mobile TV [68]. In addition, facilitating conditions has been shown to have a positive effect on usefulness of mobile banking, performance expectations of mobile shopping [69] and wireless trust which in turn affects behavioural intention to use mobile Internet [35].

Specifically, we propose a positive influence of facilitating conditions on the intention to use mobile payment services. These facilitating conditions could be the support offered by payment providers for using services and the infrastructure to conduct mobile payments in practical situations such as points of sale. Such facilitating conditions are especially important given the multitude of standards, platforms, technologies and offerings on the market. When an individual believes that the needed infrastructure and support for using mobile payment services exists, he or she is more likely to use mobile payments, which in turn will influence the actual use of mobile payment services [61]. Accordingly, we propose the following hypothesis:

H2: Facilitating conditions has a positive effect on the behavioural intention to use mobile payment services.

Learning Costs of Mobile Payment

In marketing, management, and economics literature, a consensus has emerged that switching costs are prevalent in a wide variety of industrial and consumer contexts [26]. Switching costs are defined as the one-time costs that customers associate with the process of switching from one product to another [49]. While switching costs must be associated with the switching process, they need not be incurred immediately upon switching. Furthermore, switching costs need not be limited to objective, economic costs. When consumers simplistically state that it’s just not worth it to switch providers, they may perceive impediments ranging from search costs, transaction costs, learning costs, loyalty-card discounts, customer habit, emotional cost and cognitive effort, coupled with financial, social, and psychological risk on the part of the buyer. Given their importance, it seems natural that firms would want to manage their customers’ perceptions of switching costs

Switching costs have been decomposed in different types. [26] proposed three types of switching costs: transaction costs; learning costs; and artificial or contractual costs. In [21], six dimensions are identified: performance costs uncertainty costs; pre-switching search and evaluation costs; post-switching behavioural and cognitive costs; setup costs; and sunk costs. [5] categorizes procedural switching costs along four dimensions: economic risk costs; evaluation costs; set up costs; and learning costs. In the smartphone market, [43] analysed several categories of switching cots: searching costs; learning costs; artificial costs; sunk costs; and uncertainty costs. They found that searching, learning and uncertainty dimensions of switching costs are the most relevant in the smartphone market.

This paper focuses on Learning costs. This type of costs includes the time and effort needed to acquire, exchange and evaluate information [21]. Learning costs are considered transition costs that are incurred while shifting from one situation to a new situation, for example, learning to use a new information system [24]. Learning costs are non-monetary perceived costs which include sacrifices (time, effort and search costs) needed to obtain or use a product or service [34]. The negative effect of learning cost on behavioural intention to use mobile payment services has been proposed and tested in few studies in the past. There is some evidence of a negative impact of learning costs on loyalty to SMS, similarly to the evidence obtained in other studies in this area [35].

Lee, et al. [30] found that absorptive capacity contributes to the intention to use mobile services. Reversely, the lack of prior knowledge requires potential users to learn how to use the new mobile payment service, thus incurring higher learning costs, higher switching costs and reducing their intention to use it. Similarly, [35] found that perceived costs had a significant negative effect on behavioural intention to use mobile payment services. There is evidence that switching costs have a significant impact on repeat choice behaviour [67]. Thus, if a critical mass of users is not reached, due to the learning costs required, mobile payment solutions will not be adopted [55].

Due to the special relevance of learning costs in the mobile context and the still low acceptance of mobile payment systems, the focus here is on studying the impact of learning costs on the acceptance of mobile payment services. Therefore, based on previous arguments, we propose the following hypothesis:

H3: Learning costs has a negative effect on the behavioural intention to use mobile payment services.

In prior research, perceived costs have been found to have no direct impact on intention to use mobile ticketing [37]. Also other studies found no significant effect when testing the direct path from learning costs to intention as specified in our H3 [31]. For this reason, we consider that the effect of learning costs on the intention to use mobile payment services may also be indirect and mediated by other concepts. Specifically, we focus on perceived functional value and facilitating conditions as factors that may play a role in the influence of learning costs on the intention to use mobile payments.

We suggest that perceived functional value can be influenced by learning costs. Higher perceived costs are expected to evoke lower value, although prior studies found no significant effect [66]. Conversely, in their study on mobile channel, [25] found that a greater cognitive effort reduces perceived functional value of the mobile channel and, in turn, the intention to use it. Based on previous arguments we propose that learning costs might have a negative impact on user´s intention to use mobile payments through perceived functional value as the learning costs might reduce the overall perceived functional value of this technology. Therefore, we propose the following hypothesis:

H4: The negative influence of learning costs of mobile payments in intention to use mobile payment is mediated by perceived functional value of mobile payment.

Learning costs may reduce users’ perceptions with regards to facilitating conditions available to use mobile payment services. Time, effort and search costs sacrifices hinder the perceptions and intentions toward using new systems. Indeed, cognitive effort, understood as an information search cost, has been found to be the strongest barrier to the use of the mobile channel [25]. In a study on mobile health services, [56] consider learning costs as part of response cost.

Thus, we propose that learning costs might impact negatively the infrastructure to support the system and in turn user intentions to use mobile payment. In the initial periods of a new mobile service, users incur evaluation costs (to gather information and assess that information before switching to the new service), and cognitive costs (to become familiar with the new service after switching). Those learning costs may reduce user perceptions of external conditions facilitating the acceptance and support of the new technology or mobile service. Prior research states that users require more effort and time to learn how to use mobile payment tools [60], thus incurring higher learning costs and having scarce time to spend on evaluating facilitating conditions available. Therefore, the presence of facilitating conditions is posited to increase user intentions to use mobile payments, while the existence of learning costs reduce user perceptions of facilitating conditions. Based on prior discussion, we posit the following:

H5: The negative influence of learning costs of mobile payments in intention to use mobile payment is mediated by facilitating conditions of mobile payment.

A theoretical model with proposed hypotheses is presented in Figure 1.

Figure 1: Theoretical model 

3.1 Data Collection

Data used in this research was obtained from a study conducted in a major university. University students were considered as an appropriate sample due to their familiarity with the use of mobile phones and mobile services nowadays. They are frequently heavy users of mobiles devices and are aware of all the technological advances in this field. Moreover, recent studies demonstrate that these respondents are often willing to try new features on their mobile phones and pay for them [16].

A questionnaire was developed that self-administered through an online tool. All questions were tested with academic experts to improve readability and understanding. Also, a pre-test was conducted with respondents.

A total of 463 valid responses were obtained. 60% of our sample were men and 40% were women. The average time they spend with a mobile device over the day is between 2 and 5 hours. More than 70% of the students uses mainly a mobile phone, whereas 30% mainly uses other devices such as tablets. Over 55% of the students uses a credit card and 45% uses PayPal or similar technologies.

The measures used in this research were obtained from well-known tested scales, and are all measured on a Likert scale (1 - 5) from totally disagree to totally agree. In order to measure perceived functional value we took into account recent contributions in [14]. The literature on switching costs is very prolific and we obtained our measures for learning costs from the work of Burnham, et al. [5] and also considered other contributions from Kohli [27]. This adaptation of the learning cost scale was confirmed in the pre-test as well as the reflective measures used. Facilitating conditions were extracted in [8]. Finally, Intention to use mobile payment was adapted from [37] [50] . A detailed list of scales is presented in Table 1.

Table 1: Measurement model 

3.1 Reliability and Validity

We conducted a confirmatory factor analysis (CFA) including the independent and dependent constructs with Lisrel 8.8. As can be observed in Table 1, the results of the four-factor model suggest a good fit (2 (29) = 78.08; CFI =.97 , NNFI =.95 , RMSEA = .07 ). Convergent validity was confirmed as all of the loadings for the items on their respective constructs were significant [3]. The reliability of the multi-item scales was assured by calculating the Bagozzi and Yi [3] composite reliability (SCR) and the average variance extracted (AVE) suggested by [17].

Two different approaches were used to test discriminant validity: 1) the correlation-AVE comparison [3] and 2) the heterotrait-monotrait (HTMT) ratio of discriminant validity. The first method is widely acknowledged in literature to test discriminant validity and compares the correlations of the constructs with the average variance extracted. However, a new stream of research suggests to include the heterotrait-monotrait (HTMT) ratio of discriminant validity [65] with a .85 cut-off point [19]. The HTMT test requires the calculation of a ratio of the average correlations between constructs and the geometric mean of the average correlations within items of the same constructs. The results of both approaches are reported in Table 2.

Table 2: Discriminant validity (AVE-Correlations and HTMT) 

Data was collected with a single step approach and from the same respondents. This could be an important threat in terms of common method variance (CMV) for our research [36]. To assess the potential risk of CMV, we conducted several distinct tests:

Harman one-factor test. if common method bias poses a serious threat to the analysis, a single latent factor would account for the majority of the covariance among the measures [45]. A Harman one factor tests showed that a single factor has a decidedly inferior fit compared to our multi factor structure.

As a further test, confirmatory factor analysis of the one-factor test [45] also revealed that a multi factor structure had a better fit in our model.

The Lindell and Whitney [33] marker variable technique also showed that common method bias was not problematic.

Finally, we used a latent method factor approach [45] that involves having all the items load on both their constructs and simultaneously in a common factor. This common factor did not improve the fit of our model.