Does Central Bank Digital Currency (CBDC) Payment Create The Opportunity For the Tourism Industry?

Central Bank Digital Currency (CBDC) Payment and Perceived Risk

Since money represents value, it is used as an essential means of transaction and functions as an economic system (Kydland, 1989). In human history, transactions began with barter, that is, the exchange of goods or services for other goods or services, such as food, gold, silver, copper, or paper notes. Financial transactions have now developed to involve electronic and digital systems such as online debit/credit card payment, electronic cash or checks, and internet banking (Davies et al., 2010; Peluso, 2021). The use of the internet has accelerated extensions to financial systems, for example, the replacement of money with virtual currency, and other transaction products such as cryptocurrency and stablecoins (Shen & Hou, 2021). Finally, the most pervasive currency, known as Central Bank Digital Currency (CBDC), has been realized.

The use of these payment systems has been slowly increasing in developing countries and has become the foundation of futuristic financial transactions (Ligon et al., 2019). The preference for the fast transfer of emergency payments during COVID-19 has increased the demand for digital payment methods (Rennie & Steele, 2021). However, research in India found that the adoption level of digital payment systems is still low due to various factors, particularly demand-side factors, which include increasing tax liability records of mobile payments and customers’ low trust levels for digital payments (Ligon et al., 2019).

Very recently, the World Bank and the International Monetary Fund (IMF) have considered implementing the direct exchange of CBDC (BIS et al., 2021). In a situation in which governments are debating the adoption of CBDC due to the complexity of the system (He et al., 2022), China has a keen interest in international governance of the CBDC payment system because its popularity could be used to forestall the preemptive position of the U.S. dollar as a world currency (Wang, 2022). However, the adoption of the system can be influenced by numerous factors, such as transparency of transactions, protection of privacy, domestic political situations, legal and regulatory support, and customers’ trust in the economic system.

As a theoretical foundation, the current study focuses on customers’ perceived risk associated with tourists’ intention to adopt CBDC. The customer’s expectation of loss or a mismatch between the expected and actually purchased product/service is an axis of perceived risk theory (Bauer, 1960; Mitchell, 1992, 1999). This theory has attracted the attention of researchers and marketing practitioners for its versatility and applicability in different research contexts. The vital concept of this theory is that a consumer’s primary goal is to minimize anticipated risks rather than to maximize the expected utility from purchased services/products (Chang & Chen, 2014). In the context of electronic payment systems, the psychological risk dimension occurs when a specific payment method induces psychological anxiety or irritation (Ho & Ng, 1994). It also occurs when customers perceive when purchasing that they did not really want an item or made the wrong purchase. Ariffin et al. (2018) found that psychological risk significantly and negatively influenced consumers’ online purchase intention because consumers are concerned about security risks.

Based on perceived risk theory, this study uses the perceived risk dimensions proposed by Littler and Melanthiou (2006) in the internet banking context. This study considers perceived risk to involve financial, performance, privacy, social, time, and psychological risks and examines their influences on customers’ attitudes toward CBDC payment. In a CBDC transaction situation, the concept of perceived risk is essential in analyzing consumer behavior, which is affected by concerns relating to undesirable consequences, primarily in online transaction situations (Bindseil, 2019; Koziuk & Ivashuk, 2022; Okuwoga, 2021).

Relationship Between Perceived Risk and Attitude Toward CBDC Payment

Internet banking institutions require stricter security compared to other industries since financial services deal with personal and financial risk issues (Baker et al., 2011). Ho and Ng (1994) conceptualized five perceived risk dimensions pertinent to electronic payment systems: performance, time loss, psychological, physical, and financial. Since financial risk refers to consumers’ concerns about the security of internet banking transactions, it is a suitable variable with which to examine consumers’ attitudes to the CBDC payment system (Roy et al., 2017).

S. Park and Tussyadiah (2017) developed a multidimensional instrument for assessing perceived risk in the mobile travel booking context, in which the dimensions of perceived risk included performance, time, privacy/security, financial, physical, psychological, and device risks. Lee (2009), in contrast, studied the perceived risk of internet banking adoption and divided the perceived risk dimensions into time, performance, financial, social, and security risks. Five dimensions of risk were used by Shuhidan et al. (2017) in the mobile banking context. According to Zhao et al. (2008), in Chinese internet banking services adoption, only four perceived risk dimensions, including privacy, security, finance, and performance risks, were considered. In the context of retail banking services, other researchers have assumed that perceived risk encompasses financial, performance, time, psychological, and security risks (Featherman & Pavlou, 2003; Littler & Melanthiou, 2006). In general, the number and types of dimensions may vary based on the context of risk occurrences. For example, Han et al. (2019) dealt with the perceived risk associated with electric airplanes, encompassing performance risk, physical risk, psychological risk, and financial risk. Choe et al. (2021) found that performance risk and psychological risk constituted the perceived risk relating to drone-based delivery services in terms of evaluating the image of such services. The existing literature implies that the significance of each facet of perceived risk differs depending on the field. This indicates that examining the multidimensional perceived risk in CBDC payment is a necessity.

This study adopted a multidimensional structure for perceived risk in the context of CBDC payments that tourists may use during travel. The first risk dimension, financial risk in internet banking, is involved with concerns about financial loss as a result of providing wrong specifications for account numbers, legal issues triggered by a user’s incorrect transfer, exposure of personal account details, and scamming (Hanafizadeh & Khedmatgozar, 2012). Radic et al. (2022) noted that customers’ use of CBDC requires the removal of perceived potential financial risks. Tong and Jiayou (2021) elaborated that CBDC is vulnerable to financial risks related to nonbank financial intermediation and issuance of currency. The negative relationship between perceived financial risk and attitude toward internet banking systems is supported by previous studies (Hong et al., 2016; Zou, 2021). Therefore, perceived financial risk-generating factors can increase negative views about using CBDC payments.

As the second dimension, performance risk can occur when a specific payment is inadequate to complete a transaction because of additional charges, failure to access an account, or rejection by the bank (Ariffin et al., 2018; Ho & Ng, 1994). In the context of internet banking, performance risk relates to inefficient internet banking transaction processes (Roy et al., 2017). As addressed by Hanafizadeh and Khedmatgozar (2012), the efficiency of internet banking is determined by factors including inefficient access and internet operation, and low performance of internet banking compared to its advertised capabilities. Wu et al. (2022) empirically investigated consumers’ intention to adopt digital currency electronic payment in China and found that perceived risk significantly and negatively influenced consumers’ intention to use digital currency. Interestingly, national identity and sense of belongingness were found to be significant moderators of the relationship between perceived risk and usage intention. Likewise, Chauhan et al. (2019) found that perceived security risk significantly and negatively determined consumers’ intention to adopt digital payment methods in internet banking.

As the third dimension, privacy in the internet banking security context refers to customers’ confidentiality and their ability to offer the right personal information at the required time (Ouaddah et al., 2016). Accordingly, the privacy risk in online banking or transactions arises from the disclosure of consumers’ personal information during internet banking, and at the time of providing private data to internet banking facilities, and misuse of these private data by the bank (Hanafizadeh & Khedmatgozar, 2012; Roy et al., 2017). The protection of privacy is more complex in the case of digital currency because it requires commitment from agents to secure customers’ privacy or anonymity (Koziuk & Ivashuk, 2022). Rennie and Steele (2021) defined four types of privacy risks pertinent to CBDC models, including loss of anonymity, loss of liberty, loss of individual control, and loss of regulatory control. Similarly, Tronnier and Biker (2022) noted that privacy concerns stem from customers’ past experiences, demographic characteristics, awareness about privacy practices, lack of trust, and personal beliefs.

Finally, the time risk dimension refers to consumers’ perception of the probability of wasted time, effort, and convenience when a product is wrongly purchased, mended, or replaced (Ko et al., 2004; Masoud, 2013). Considering internet banking services, Hanafizadeh and Khedmatgozar (2012) described time risk as the consumer’s perception of the time spent learning how to use internet banking, solving the problems relevant to internet banking, and implementing transactions in internet banking. Ariffin et al. (2018) found that time risk was a significant predictor of consumers’ intention to use online purchasing. Hanafizadeh and Khedmatgozar (2012) identified time risk as the mediating variable in the relationship between the consumer’s internet banking awareness and their intention to adopt internet banking. In a similar vein, Husain’s (2011) study noted that perceived risk related to time loss may be higher for internet banking services. To the best of our knowledge, there is no research that has examined the effects of perceived risk on tourists’ attitudes toward CBDC payment. To answer this question, the following hypothesis was postulated:

Relationship Between Attitude Toward CBDC Payment and Intention to Adopt CBDC Payment

A number of prior studies have widely applied the technology acceptance model (TAM) in diverse technology contexts, including online shopping (Ghani et al., 2022; Kesharwani and Bisht, 2012; Rose & Fogarty, 2006), smartphones (Özbek et al., 2014), and robotics (Go et al., 2020; Lee et al., 2018). This model was adopted to assess perceived ease of use, perceived usefulness, and behavioral intention to use. For instance, based on the TAM, Kesharwani and Bisht (2012) showed that perceived risk negatively influences the consumers’ intention to adopt internet banking. Lee et al. (2018) also revealed that attitude toward robot technology positively affected the technology acceptance of the users with perceived usefulness, while perceived ease of use positively influenced perceived usefulness and attitude toward robot technology. However, the TAM has not been widely tested in the context of CBDC payment.

Recently, fintech has enhanced the convergence of CBDC with the global financial network (GFN) and contributed to alleviating strict regulations on financial sectors that hinder internetization (Wang & Gao, 2021). Some studies (Boar et al., 2020; Shkliar, 2020) have asserted that since the central banks are keen on blockchain technology and cryptocurrency, CBDC payment systems will gain popularity in tourism and hospitality businesses. The relationship between attitude toward a certain technology and behavioral intention has been found to be positive when using online travel agencies (Ponte et al., 2015), online banking (Kaur & Arora, 2020; Martins et al., 2014), and mobile banking (G. Kim et al., 2009).

Furthermore, Zhu et al.’s (2012) findings indicated that consumers’ attitude toward online games has a positive correlation with their intention to use computer games on the internet. With regard to smartphone chatbot adoption for shopping, the intention to use the technology was influenced by customers’ attitudes toward chatbots (Kasilingam, 2020). Similarly, a study that focused on the use of mobile devices for private clubs showed that attitude influenced the intentions of club members to adopt a mobile device (Morosan & DeFranco, 2014). Scholars have repeatedly found significant relationships between attitude and behavioral intentions toward technological innovations in the hospitality and tourism sectors (Han et al., 2019; Hwang, Kim, Kim, et al., 2021). Therefore, those who have a favorable attitude toward CBDC payment will demonstrate a high level of intention to adopt CBDC payments. On this basis, the following hypothesis was proposed:

Moderating Effects

The present study identified other factors and tested their moderating role in the adoption of the CBDC payment system. The factors used in this study are related to consumers’ personal characteristics (innovativeness and previous experience) and product characteristics (structural assurance and media encouragement). The present study investigated the effects of these characteristics on the relationship between perceived risk and attitude toward the use of central banking digital currencies, because consumers’ purchasing of new products is influenced by both internal factors (personal attributes) and external factors (product-related factors; Turkyilmaz et al., 2015). Considering the circumstances or situations in which the individuals or products exist helps with understanding the potential buying behavior of consumers in CBDC transactions (Peluso, 2021; Radic et al., 2022; Shkliar, 2020).

Moderating Role of Personal Characteristics (Consumer Innovativeness and Previous Experience)

Individual innovativeness has been defined as an individual’s propensity to accommodate new knowledge and make an independent decision without others’ influence (Midgley & Dowling, 1978; F. Zhang et al., 2020). Jeong et al. (2017) found that consumer innovativeness influenced consumers’ perceived novelty regarding wearable technologies. Z. Zhang and Hou (2017) found that innate consumer innovativeness moderated the relationship between perceived risk and intention to purchase innovative products. Consumer innovativeness was also considered a pervasive factor, which moderated the relationship between the antecedents for using mobile commerce and consumers’ intention to use mobile commerce technology (Anwar et al., 2021). Hirunyawipada and Paswan (2006) noted that consumer innovativeness strengthened intentions to use new products, and Aldás-Manzano et al. (2009) found that consumer innovativeness was a major predictor of intention to adopt electronic banking. These results help to promote the positive role of consumer innovativeness in dealing with perceived risks (privacy, security, social, time, performance risks). Chauhan et al. (2019) also showed that consumer innovativeness significantly and positively influenced the intention to adopt internet banking services.

CBDC, as a cashless transaction, influences unbanked consumers’ involvement in finance systems (Seth, 2022). Those who aspire to adopt this novel technology are likely to reveal positive attitudes toward it, even though they perceive diverse risks stemming from using that technology. Hence, the following hypothesis was posited:

Another underpinning variable that may help to allay the effects of perceived risk is past experience. According to Kaur and Arora (2020), past experience was conducive to fostering customers’ ability to decide independently, while those who did not have past experience demanded approval or support from others in order to decide. Taylor and Todd (1995) empirically confirmed that there is a strong relationship between the intention to adopt technology and past experience. Similarly, John (2013) confirmed that an individual’s past experience had a positive effect on self-efficacy and intention to use social networking programs. Straub’s (2009) findings indicated that personal characteristics such as age, gender, and prior experience influenced technology adoption. Lu et al. (2011) also found that customers’ previous experience with internet payment technology positively affected their perceptions of payment services.

Kaur and Arora (2020) established that prior experience was a moderator between perceived risk in using online banking and consumers’ intention to adopt internet banking. In exploring consumers’ intention to use mobile technology, trust and previous experience with internet use moderated the relationship between perceptions of utilitarian and hedonic values in mobile technology and their intention to adopt the technologies (J. Park and Yang, 2006). Previous experience is thus likely to act as a moderator in the relationship between perceived risks and attitude toward CBDC payment. As a result, the following hypothesis was postulated to test the relationship between the related variables:

Moderating Role of Product Characteristics (Structural Assurance and Media Encouragement)

McKnight et al. (2002) indicated that trust enables consumers to reduce perceived risk and insecurity. In an initial trust formation stage, institutional-based trust nurtures an individual’s feeling of security through safety nets and guarantees or other conditions (McKnight et al., 1998; Shapiro, 1987; Zui et al., 2022). As conceptualized by McKnight et al. (1998), institution-based trust has two dimensions: structural assurance belief and situational normality belief. Structural assurance refers to a belief that regulations, promises, guarantees, and contracts lead to success (Shapiro, 1987; Zucker, 1986). Filotto et al. (2021) found that structural assurance played a decisive role in whether customers adopted online banking. G. Kim et al. (2009) also observed that structural assurance significantly influenced customers’ trust in and favorability toward mobile banking. Zhou (2011) noted that structural assurance helped customers be more confident in their mobile banking payments and strengthened trust in the system. In summary, structural assurance is likely to serve a moderating role in the relationship between consumers’ perceived risk and attitudes toward CBDC payment. Consequently, the following hypothesis was proposed:

Those who initially adopt a technology tend to display confidence in seeking new technology and avoid perceived risk by seeking interpersonal communication channels, participating in social events, and acquiring information (Rogers, 1995). Zolait and Sulaiman (2009) found that word of mouth and exposure to mass media had a significant effect on Yemeni consumers’ acceptance of internet banking, because mass media coverage helped in allaying consumers’ perceived risks. Likewise, YenYuen and Yeow (2008) found that there was a positive relationship between mass media opinion and consumers’ adoption of internet banking, with customers noting that mass media helped enhance the credibility of the system and alleviate perceived risk. Recently, Radic et al. (2022) ascertained that mass media contributed to creating positive attitudes in tourists to the adoption of digital currencies, because consumers believed that the content in mass media helped enhance the comprehensiveness of digital currencies and dissipate distrust.

However, there is no study that has empirically explored the moderating role of media encouragement in predicting the relationship between perceived risks and attitude toward CBDC payment. Thus, the following hypothesis was proposed:

A conceptual framework including relevant constructs and hypotheses was developed as shown in Figure 1. Perceived risk was conceptualized as encompassing five domains. It was hypothesized that the perceived risk affects attitude toward CBDC payment, which determines the intention to adopt CBDC payment. Four moderating variables are situated between perceived risk and attitude toward CBDC payment.

OPEN IN VIEWER

Figure 1.

Proposed theoretical model.

Measurement Development

The measurements for all study variables were adapted from previous studies and were adjusted to the context of CBDC payment innovation. The measurement items for multidimensional perceived risk were developed from previous studies (Chen, 2013; Choe et al., 2021; Han et al., 2019; Martins et al., 2014). Each of the three items for the varieties of perceived risk was measured using a 7-point Likert scale ranging from strongly disagree (1) to strongly agree (7). To adopt Han et al.’s (2020) study, we used four measurements for attitude with bipolar semantic-differential scales (i.e., “Negative” [1]–“Positive” [7]). Another four measurements using a Likert-type scale were devised based on the studies conducted by Ajzen (1991) and M. Kim (2021) in order to assess the intention to adopt CBDC payment.

In addition, measurement items for moderating variables, which were consumer innovativeness, structural assurance, and media encouragement, were established based on a 7-point Likert scale. The four questionnaires were developed to assess consumers’ general innovativeness on the basis of earlier studies (e.g., Goldsmith & Hofacker, 1991; Hwang, Kim, & Lee, 2021; Z. Zhang & Hou, 2017). The survey included questions related to the extent to which respondents tended to accept novel offerings sooner than other people and the degree to which they were willing to adopt new products and services. With regard to structural assurance, we created the following four items according to Delgosha and Hajiheydari’s (2021) study: (1) “I would feel okay using CBDC because it is backed up by a monetary authority (central bank)”; (2) “Monetary stability would make me feel all right about using CBDC”; (3) “Favorable-to-consumer centralized structures would help me feel safe using CBDC”; and (4) “Having the backing of a central bank would make me feel secure in using CBDC.” Also, we developed two items for media encouragement from Juschten et al.’s (2019) study. These items were formulated to estimate how individuals perceive media coverage in promoting a positive image of CBDC payment and how they are motivated to use CBDC by such media coverage. To assess previous experience, we used a nominal scale (Yes or No) to answer the question, “Have you ever held cryptocurrency or virtual currency and used these for buying products/services?” Overall, we used various scales in order to estimate study constructs, which helped reduce common method bias.

Survey Design

The survey was structured as follows. The first section contained the purpose of the study and the screening questions. Two questions were used to restrict the survey to participants who were aware of the emergence of CBDC and who had often purchased products/services relating to the tourism industry. Participants were asked to answer yes or no to the question of whether they had heard about central bank digital currency, and those who answered yes were eligible to participate in the survey. Also, they were asked to indicate the frequency of purchasing tourism products based on the question “How often do you consume tourism products (e.g., hotel, tourism package purchase, airline, cruise, car rental) on average?” and only those who confirmed frequent buying experiences were permitted to complete the survey. The second part of the survey began with detailed information about CBDC, which also emphasized its legal tender status as declared by the government. Thus, participants became more knowledgeable about the subject and understood how it is distinguished from cryptocurrency. Multiple questions then followed in order to assess the multidimensional perceived risks associated with CBDC payments. Also, participants were asked to respond to questions that had been developed to measure attitudes toward CBDC payment, intention to adopt CBDC payment, and four moderating constructs. The last section included questions to identify the demographic characteristics of the survey participants.

The survey was initially developed in English and subsequently translated into Chinese and Korean. We then hired multiple academics who were native Korean and Chinese speakers and asked them to review the questionnaire document to ensure there were no grammatical errors. We also invited scholars in tourism departments and frequent international travelers to complete a pretest of the survey, and the questions were fine-tuned to improve clarity.

Data Collection

PwC (2021) listed the most mature countries with respect to retail CBDC, and both mainland China and Korea were ranked in the top 10. These two countries are also known as innovative nations, which implies that people living in these countries have significant knowledge of payment innovations. Thus, Chinese and South Koreans were determined to be suitable survey targets, and we arranged online data collection with help from research companies in these respective countries. Both companies are known to have a substantial number of panels, and they use several tools, such as an incentive program, in order to obtain genuine responses from participants. The survey yielded a total of 820 responses, which included data from 397 Chinese and 423 Koreans, over 3 weeks in September 2021. Of these, we identified 25 multivariate outliers using the Mahalanobis distance, which is known as a useful multivariate distance metric. As a result, 795 responses were retained and used for the data analysis.

Analytical Technique

Following a two-step approach (Anderson & Gerbing, 1988), the measurement model was initially assessed using confirmatory factor analysis (CFA), and then the research hypotheses were examined using structural equation modeling (SEM). A metric invariance test allows researchers to substantiate multigroup comparisons of factor variances and covariances, and it is often adopted to determine the difference or similarity across groups. For example, Han et al. (2021) divided airline travelers into a high-perceived-risk group and a low-perceived-risk group and conducted a metric invariance test in order to examine the moderating effect of perceived risks on eco-friendly technology product acceptance behaviors. Hwang, Kim, Kim, et al. (2021) used a metric invariance test to assess the different degrees of effect of attitude on behavioral intentions across low versus high levels of product involvement in the robotic restaurant context. Therefore, a metric invariance test, along with a chi-squared test, was used to identify the moderating effect of personal characteristics (i.e., consumer innovativeness and previous experience) and product characteristics (i.e., structural assurance and media encouragement) in the relationship between perceived risk and attitude toward CBDC payment. The analyses were conducted using SPSS 25.0 and AMOS 23.0.

Profiles of Survey Respondents

First, this study performed frequency analysis to understand respondents’ profiles. The respondents’ demographic characteristics are as follows. The total sample of 795 comprised 414 Koreans and 381 Chinese. These respondents were divided almost equally between males and females: 49.9% (N = 397) males and 50.1% (N = 398) females. Their average age was 37.3 years. With regard to their educational background, 51.1% had graduated from a 4-year university, and 21.3% had finished graduate school. In total, 21.0% reported that their income was between 55,000 and 69,999 U.S. dollars per year, 18.4% indicated between 25,000 and 39,999 U.S. dollars, and 17.9% indicated between 70,000 and 84,999 U.S. dollars. With regard to the frequency of purchasing tourism products, 35.1% indicated that they usually bought products two to three times a year, and 23.5% reported four to five purchases a year.

Measurement Model

The validity of the measurement model was determined using CFA. The results showed an appropriate fit to the current study data: χ² = 447.983, df = 186, p < 0.001; χ²/df = 2.409; incremental fit index (IFI) = 0.986; comparative fit index (CFI) = 0.986; Tucker-Lewis index (TLI) = 0.982; root-mean-square (RMSEA) = 0.042 (Byrne, 2001). All measurement items for constructs were significantly loaded except one of the measurements for time risk, and we removed this specific item from the measurement model. As Table 1 shows, the standardized factor loadings for all items ranged from 0.848 to 0.912, which were greater than the commonly suggested level of 0.6. The skewness and kurtosis of each measurement were between −0.932 and 0.294, with a standard error of 0.087, and between −1.030 and 0.747, with a standard error of 0.173, respectively, which lay within the acceptable ranges following Kline’s (2005) recommendation. Hence, the normality assumption was supported.

OPEN IN VIEWER

Table 1.

Results of Confirmatory Factor Analysis: Items and Loadings.

Construct and Scale Item	Standardized Loading a
Financial risk
CBDC payments are likely to cost more than I thought	0.848
I might get overcharged if I use CBDC payment.	0.898
The chances of losing money if I use CBDC payment are high.	0.894
Performance risk
The performance of the CBDC payment might create problems with my credit.	0.860
CBDC payments do not seem to perform well.	0.884
Considering the expected level of performance of CBDC payment, it would be risky to use it.	0.872
Privacy risk
Using CBDC payment may not protect my personal information (e.g., phone number, address, etc.).	0.930
Others may steal personal information when using CBDC payment.	0.931
Personal information could be exposed when using CBDC payment.	0.929
Psychological risk
The usage of CBDC payment would lead me to a psychological loss.	0.897
Using CBDC payment would not fit in well with my self-image.	0.866
Using CBDC payment makes me feel nervous.	0.900
Time risk
It takes time to learn how to use CBDC payment.	0.859
It takes time to learn the rules of CBDC payment.	0.912
Attitude toward CBDC payment
For me, using CBDC payment for hospitality/tourism product purchases is. . .
Bad ↔ Good	0.858
Foolish ↔ Wise	0.863
Unpleasant ↔ Pleasant	0.868
Harmful ↔ Beneficial	0.901
Intention to adopt CBDC payment
I am willing to adopt CBDC payment in a hospitality/tourism consumption situation in the future.	0.898
I plan to adopt CBDC payment for hospitality/tourism product purchases in my daily life.	0.903
I intend to use CBDC payment for hospitality/tourism product purchases in the future.	0.909
I am willing to use CBDC as a means of payment for hospitality/tourism product purchases in the future.	0.897
Goodness-of-fit statistics: χ2 = 447.983, df = 186, p < .001; χ2/df = 2.409; IFI = 0.986; CFI = 0.986; TLI = 0.982; RMSEA = 0.042.

Note. CBCD = central bank digital currency; NFI = normed fit index, IFI = incremental fit index, CFI = comparative fit index, TLI = Tucker-Lewis index, and RMSEA = root-mean-square error of approximation.

a

All factors loadings are significant at p < .001.

According to Hair et al. (2006), convergent validity can be established based on each construct’s average variance extracted (AVE). AVE values were calculated, and all values were greater than the recommended value of 0.5. Also, we computed composite reliability (CR) values for each variable in order to assess internal consistency. The CR values range was between 0.733 and 0.909, which exceeded the minimum level of 0.70 (Hair et al., 2006), and therefore, internal reliability was established. Discriminant validity was estimated following Fornell and Larcker’s (1981) criterion. As Table 2 displays, the results showed that the AVE value for each construct was greater than the square of the correlation between each pair of variables, except for the link between financial risk and performance risk. According to Bagozzi and Yi’s (1988) guideline, a χ² difference test between a constrained model and an unconstrained model was performed for this exception, and the difference of χ²(1) was 76.488, which is far greater than 3.941 (df = 1), established discriminant validity.

OPEN IN VIEWER

Table 2.

Results of Measurement Model: Correlations, AVE, CR, Mean, and SD.

	Mean (SD)	AVE	(1)	(2)	(3)	(4)	(5)	(6)	(7)
(1) Financial risk	3.952 (1.522)	0.562	0.794 a	0.828 b	0.726	0.721	0.722	−0.033	−0.055
(2) Performance risk	3.834 (1.478)	0.552	0.686 c	0.787	0.815	0.799	0.805	−0.078	−0.106
(3) Privacy risk	4.061 (1.638)	0.687	0.527	0.664	0.868	0.759	0.792	−0.154	−0.154
(4) Psychological risk	3.483 (1.553)	0.636	0.520	0.638	0.576	0.800	0.741	−0.121	−0.192
(5) Time risk	4.012 (1.552)	0.669	0.521	0.648	0.627	0.549	0.733	−0.134	−0.145
(6) Attitude toward CBDC payment	5.174 (1.167)	0.658	0.001	0.006	0.024	0.015	0.018	0.885	0.709
(7) Intention to adopt CBDC payment	5.202 (1.228)	0.713	0.003	0.011	0.024	0.037	0.021	0.503	0.909

Note. CBDC = central bank direct currency; CR = composite reliability; AVE = average variance extracted.

a

Composite reliabilities are along the diagonal.

b

Correlations are above the diagonal.

c

Squared correlations are below the diagonal.

SEM

This study used SEM to test the two proposed hypotheses (see Table 3). The goodness-of-fit statistics, χ² = 546.859, df = 200, p < 0.001; χ²/df = 2.734; IFI = 0.981; CFI = 0.981; TLI = 0.978; RMSEA = 0.047, confirmed an acceptable level. The analysis results showed that perceived risk significantly and negatively influenced attitude toward CBDC payment (β = −0.126, t = −3.346, p < 0.001). Attitude toward CBDC payment had a significant and positive effect on the intention to adopt CBDC payment (β = 0.754, t = 23.772, p < 0.001). Therefore, both Hypothesis 1 and Hypothesis 2 were supported. In addition, we computed the weights of first-order constructs on designated second-order constructs, and the results were as follows for financial risk (β = 0.891), performance risk (β = 0.984), privacy risk (β = 0.916), psychological risk (β = 0.898), and time risk (β = 0.901).

OPEN IN VIEWER

Table 3.

Results of Structural Model Evaluation.

	Independent Variable	Dependent Variable	β	t-value	Hypothesis Testing
H1	Perceived risk	→ Attitude toward CBDC payment	−0.126	−3.346**	Supported
H2	Attitude toward CBDC payment	→ Intention to adopt CBDC payment	0.754	23.772**	Supported

Note. CBDC = central bank direct currency; goodness-of-fit statistics: χ² = 546.859, df = 200, p < 0.001; χ²/df = 2.734; incremental fit index = 0.981; comparative fit index = 0.981; Tucker-Lewis index = 0.978; root-mean-square error of approximation = 0.047.

Total variance explained (R²): R² for attitude toward CBDC payment = 0.016, R² for intention to adopt CBDC payment = 0.569.

**

p < .001.

Metric Invariance Test

The moderating effect of personal characteristics, including consumer innovativeness and previous experience, was estimated using a metric invariance test (see Table 4). First, we divided the group depending on the degree of consumer innovativeness, using its mean value. As a result, we classified the respondents into a low level of consumer innovativeness (N = 380) and a high level of consumer innovativeness (N = 412). We then compared the structural invariance between the freely estimated model and the equally constrained model. The goodness-of-fit statistics of the baseline model were χ² = 1202.694, df = 404, p < .001; χ²/df = 2.977; IFI = 0.952; CFI = 0.952; TLI = 0.945; RMSEA = 0.050, while the goodness-of-fit statistics of the nested model were χ² = 1211.987, df = 405, p < .001; χ²/df = 2.993; IFI = 0.952; CFI = 0.952; TLI = 0.945; RMSEA = 0.050. Thus, the result showed a significantly different chi-squared value of 9.293 (Δdf = 1), p < .05, confirming Hypothesis 3. The distinct difference between the two groups was that perceived risk was a significant antecedent of attitude toward CBDC payment for the low consumer innovativeness group (β = −0.237, t = −4.107, p < .001), whereas it was insignificant for the high consumer innovativeness group (β = −0.078, t = −1.486, p > .05).

OPEN IN VIEWER

Table 4.

Results of the Invariance t-test: Personal Characteristics.

Linkages			Low Consumer Innovativeness (n = 380)		High Consumer Innovativeness (n = 412)		Nested Model (equally constrained)	Chi-Squared Difference
			β	t-value	β	t-value
H3	Perceived risk	→ Attitude toward CBDC payment	−0.237	−4.107**	−0.078	−1.486	χ2 (405) = 1211.987	Δχ2 (1) = 9.293*
Goodness-of-fit statistics for the baseline model: χ2 = 1202.694, df = 404, p < .001; χ2/df = 2.977; IFI = 0.952; CFI = 0.952; TLI = 0.945; RMSEA = 0.050.
Linkages			With Previous Experience (n = 287)		Without Previous Experience (n = 508)		Nested Model (equally constrained)	Chi-Squared Difference
			β	t-value	β	t-value
H4	Perceived risk	→ Attitude toward CBDC payment	0.065	1.021	−0.180	−3.857**	χ2 (405) = 1180.482	Δχ2 (1) = 7.450*
Goodness-of-fit statistics for the baseline model: χ2 = 1173.032, df = 404, p < .001; χ2/df = 2.904; IFI = 0.958; CFI = 0.958; TLI = 0.952; RMSEA = 0.049.

Note. CBCD = central bank digital currency; IFI = incremental fit index; CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root-mean-square error of approximation.

*

p < .05, **p < .001.

With regard to the moderating role of previous experience, we divided the respondents into those who had ever held and used cryptocurrency or virtual currency for buying products/services (N = 287) and those who had never held them (N = 508). The goodness-of-fit statistics of the baseline model were χ² = 1173.032, df = 404, p < .001; χ²/df = 2.904; IFI = 0.958; CFI = 0.958; TLI = 0.952; RMSEA = 0.049, confirming an acceptable level. Structural invariance comparison showed a significant difference in the chi-squared values at 7.450 (Δdf = 1) and p < .05. The adverse effect of perceived risk on attitude toward CBDC payment became severe for individuals without experience of holding cryptocurrency or virtual currency. Hence, Hypothesis 4 was supported.

In order to determine the moderating effect of product characteristics, we conducted the same procedure that was used to identify the moderating effect of personal characteristics. We divided the group depending on how respondents perceived the level of structural assurance related to CBDC. Using the mean value, we grouped 375 responses showing a low level of structural assurance and 420 responses with a high level of structural assurance. Then, the structural invariance of the baseline model was compared with a nested model. The goodness-of-fit statistics of the freely estimated model were χ² = 1107.780, df = 404, p < .001; χ²/df = 2.742; IFI = 0.960; CFI = 0.960; TLI = 0.954; RMSEA = 0.047, which indicated an acceptable fit to the data. The results also showed a significant difference in the chi-squared value (Δχ²(1) = 13.899, p < .001) from the equally constrained model. The distinct difference between the two groups was that perceived risk was a significant predictor of attitude toward CBDC payment for the low structural assurance group (β = −0.234, t = −4.259, p < .001), whereas it was insignificant for the high structural assurance group (β = −0.015, t = −0.293, p > .05).

Similarly, we divided the group depending on how respondents perceived the level of media encouragement related to CBDC. We classified the respondents into a low level of media encouragement (N = 387) and a high level of media encouragement (N = 408). The goodness-of-fit statistics of the freely estimated model were χ² = 1397.636, df = 404, p < .001; χ²/df = 3.459; IFI = 0.938; CFI = 0.937; TLI = 0.928; RMSEA = 0.056. The goodness-of-fit statistics of the equally constrained model were χ² = 1405.391, df = 405, p < .001; χ²/df = 3.470; IFI = 0.938; CFI = 0.937; TLI = 0.928; RMSEA = 0.056. Thus, there was a significant difference (Δχ²(1) = 7.755, p < .05) between the chi-squared values of these two models. The negative effect of perceived risk on attitude was prominent for the group who perceived a low degree of media encouragement. Both Hypothesis 5 and Hypothesis 6 were accordingly supported (see Table 5).

OPEN IN VIEWER

Table 5.

Results of the Invariance Test: Product Characteristics.

Linkages			Low Structural Assurance (n = 375)		High Structural Assurance (n = 420)		Nested Model (equally constrained)	Chi-Squared Difference
			β	t-value	β	t-value
H5	Perceived risk	→ Attitude toward CBDC payment	−0.234	−4.259**	−0.015	−0.293	χ2 (405) = 1121.679	Δχ2 (1) = 13.899**
Goodness-of-fit statistics for the baseline model: χ2 = 1107.780, df = 404, p < .001; χ2/df = 2.742; IFI = 0.960; CFI = 0.960; TLI = 0.954; RMSEA = 0.047.
Linkages			Low Media Encouragement (n = 387)		High Media Encouragement (n = 408)		Nested Model (equally constrained)	Chi-Squared Difference
			β	t-value	β	t-value
H6	Perceived risk	→ Attitude toward CBDC payment	−0.177	−3.049*	−0.024	−0.452	χ2 (405) = 1405.391	Δχ2 (1) = 7.755*
Goodness-of-fit statistics for the baseline model: χ2 = 1397.636, df = 404, p < .001; χ2/df = 3.459; IFI = 0.938; CFI = 0.937; TLI = 0.928; RMSEA = 0.056.

Note. CBCD = central bank digital currency; IFI = incremental fit index; CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root-mean-square error of approximation.

*

p < .05. **p < .001.

Discussion

CBDC has been introduced as a likely innovation in the foreseeable future, and it is highly likely to promote diversity in payment options and increase financial inclusion (Auer & Böhme, 2020). Moreover, as economies go digital, its practical role of making cross-border and online payments faster and cheaper will be invaluable, and CBDC is therefore perhaps greatly to be welcomed. However, novel offerings are generally associated with various risks, which may have an adverse effect on adoption behavior (Chauhan et al., 2019; Han et al., 2019; Wu et al., 2022). To date, there have been few efforts to estimate consumers’ risk perceptions of CBDC payments from the point of view of travelers. The current study set out to bridge this research gap by exploring the effect of a multidimensional risk factor in the formation of consumer acceptance of CBDC payments. In addition, this study paid particular attention to personal characteristics and product characteristics linked to CBDC, which can potentially influence the relationship between perceived risk and attitude toward CBDC payment.

Based on a multidimensional approach to perceived risk associated with CBDC payment, our study revealed that the most significant first-order dimension was the performance risk in the international tourism sector. Also, this study found that perceived risk greatly influenced attitudes toward CBDC payment. The results indicated that lower perceived risk would result in more favorable attitudes toward CBDC payment. This outcome can be explained on the basis of risk perception theory. It is also in accordance with earlier studies (e.g., Choe et al., 2021; Hwang, Kim, Kim, et al., 2021), which observed that individuals with a high-risk perception toward innovations tended to exhibit avoidance behavior. The CBDC payment system is the most recent payment system to have been developed and many people see it as a novel alternative to existing payment methods (Radic et al., 2022). There are multiple unsolved issues associated with CBDC, which has yet to roll out in many places (Okuwoga, 2021). The current status of CBDC may make people feel unclear or uncertain about several aspects that entail risk, including financial, performance, privacy, psychological, and time aspects. A close attitude-to-behavior relationship was evident. That is, attitude toward CBDC payment influences the intention to adopt this specific payment innovation. This result resonates with extant studies that have documented the need for building a positive attitude in order to increase intention and engage in a specific behavior (Kaur & Arora, 2020; Martins et al., 2014).

This study found a significant role the for moderating variables that reflect personal characteristics, namely, consumer innovativeness and previous experience. The findings indicated that the negative effect of perceived risk on attitude is strong for individuals with low innovativeness. Existing studies have underlined the positive role of consumer innovativeness in adoption behavior across different sectors (Hwang, Kim, & Lee, 2021; Jeong et al., 2017; F. Zhang et al., 2020). Our finding specifically echoes the prior finding that determined its role in mitigating perceived risks of novel payment systems (Aldás-Manzano et al., 2009). Another factor this study addressed, which could strengthen or weaken the link between perceived risk and attitude toward CBDC payment, was previous experience. The analysis showed that the negative effect of perceived risk on attitude was significant for people without previous experience. One possible reason for this outcome is that people who have ever held cryptocurrency or virtual currency may be more familiar with the use of digital currencies to complete transactions. This is in line with Kaur and Arora’s (2020) study that confirmed the moderating role of prior experience in building consumers’ intention to adopt internet banking.

In addition, this study identified the positive role of product characteristics, in terms of structural assurance and media encouragement, in minimizing the negative effect of perceived risk on attitude. This result supports the existing studies (e.g., Filotto et al., 2021; Zhou, 2011) that have noted the positive influence of structural assurance in inducing consumer adoption behavior with regard to payment innovations. This implies that governance and accountability are important in the context of payment innovation. Also, it was found that the negative effect of perceived risk on attitude was greater for those with a low degree of media encouragement. Digital forms of money, specifically CBDC, are supposed to improve various aspects of current financial transactions. Despite that, customers may feel a lack of confidence in using CBDC as it has not yet been commercialized, and our results indicated that media messages related to CBDC play an essential role. This finding aligns with the extant literature, which supports the crucial role of media encouragement in forming acceptance behavior toward innovations (Radic et al., 2022; Zolait & Sulaiman, 2009).

Theoretical Implications

Four major research objectives framed this study. They were all based on quantitative procedures and yielded the following theoretical implications. First, J. J. Kim et al. (2022), who used the five-stage consumer adoption model, showed that the proposed model predicted consumer acceptance of CBDC for buying products/services and revealed that consumer innovativeness moderated the relationship between interest and evaluation and evaluation and trial intention. Radic et al. (2022) also explored the effects of attitude toward CBDC payment, relative advantage, perceived risk, subjective norm, and perceived behavioral control on tourists’ behavioral intention to adopt CBDC payment. Prior studies also did not address the multidimensionality of perceived risk in using CBDC. For example, while Radic et al. (2022) used perceived risk as one of the antecedents to the adoption of the CBDC payment system, they did not consider the multidimensional structure of the perceived risk. However, the current study conceptualized perceived risk as five components, adopting the findings of Littler and Melanthiou (2006), and futher tested whether perceived risk has a significant impact on tourists’ attitudes toward CBDC.

Second, in line with the first implication, this study conceptualized a facet-based perceived risk, which supports the parsimonious second-order typology of risk perception (e.g., Kaur & Arora, 2020; Martins et al., 2014; S. Park & Tussyadiah, 2017). Our results validated a second-order composite of perceived risk and helped to establish a better understanding of risks associated with CBDC in a readily comprehensible manner. In addition, it was evident that the significance of risk perceptions may not necessarily be the same for every innovation. For example, Yadav et al. (2015) found that perceived risks toward internet banking did not influence consumer behavior. The relationship between perceived risk and online purchase intention was not statistically supported in Ventre and Kolbe’s (2020) study. Yi et al. (2020) determined that risk perception related to the performance of the sharing economy in the tourism industry did not adversely affect the adoption intention. In this regard, our findings provided evidence that the perceived risk of CBDC payment plays a critical role in influencing consumer behavior.

Second, there are only a few research studies that describe the influence of perceived risk on travelers’ attitudes toward CBDC payment. Therefore, it was worth investigating whether there is a relationship between travelers’ perceived risk and their attitudes toward CBDC, one of the latest digital payment systems. CBDC can contribute to modernizing the current monetary system; however, this study observed that people perceive risks, specifically financial, performance, privacy, psychological, and time risks, which in turn affect attitude. The role of each facet of perceived risk differs considerably between novel products/services. For example, physical risk is a large component of the perceived risk of electric airplanes (Han et al., 2019), whereas privacy and security risks are more relevant in the field of mobile travel booking platforms (S. Park & Tussyadiah, 2017). Nonetheless, prior studies (e.g., Radic et al, 2022) that regarded perceived risk as a unidimensional construct in the tourism sector do not provide evidence that a specific facet largely accounts for a perceived risk toward CBDC. The current study fills voids left by previous research as it found that performance risk played a salient role in shaping the perceived risk relating to CBDC. However, financial risk had a minimal impact. Accordingly, it suggests the primary focus is performance risk, in order to reduce risk perceptions relating to CBDC payment. Furthermore, the finding of a significant association between attitude and intention supports the related theory proposed by Ajzen and Fishbein (1977).

Third, this study underlines the influencing role of consumer innovativeness and previous experience in the context of payment innovation. Numerous studies have indicated the critical role of personal characteristics in consumer adoption behavior toward innovation. A unique contribution made by the present research is the finding that consumer innovativeness and previous experience moderated the relationship between perceived risk and attitude in the area of CBDC payment innovation. This study also makes a contribution to the current literature regarding the impact of individual characteristics on customer responses to payment innovations.

Fourth, this study improved our understanding of the moderating impact of product characteristics. The importance of structural assurance cannot be overstated, and it is an important step toward developing more positive attitudes toward CBDC payment. As stated earlier, this study provided evidence of the negative effect of perceived risk associated with CBDC on attitude to the use of CBDC payment systems. This result is similar to previous payment innovations, such as cryptocurrency (M. Kim, 2021). However, this study paid extra attention to the particular difference between CBDC and the current cryptocurrencies available in the market. CBDC is centralized and regulated by national financial institutions (Bindseil, 2019; Shkliar, 2020), which should increase individuals’ beliefs and trust. This research was one of the first attempts to examine structural assurance related to CBDC in forming consumer behavior and showed that personal and product characteristics have a decisive effect on the link between perceived risk and attitude. Furthermore, this study identified the prime role of media encouragement, which diminishes the negative effect of perceived risk on attitude. This study produced empirical evidence on how media can contribute to the successful implementation of CBDC payment innovation.

Practical implications

The CBDC payment system is expected to lead to a paradigm shift for international financial transaction systems and it will improve the current GFN and facilitate the creation of a cashless society (Wang & Gao, 2021). Many parts of the world are experiencing a revival in international tourism. The growth in international tourism demands easy, inexpensive, and seamless cross-border payments, and the CBDC system could play an important role in revitalizing the tourism industry (J. J. Kim et al., 2022; World Bank, 2021). Therefore, this research tried to address issues specifically related to tourists’ adoption behavior toward CBDC payment.

This study also provides important managerial suggestions for stakeholders involved in this modern payment system. First, this study has demonstrated that the CBDC payment system is affected by perceived risks, and the risk dimensions have been clearly identified. Therefore, financial institutions can take this into consideration, and make improvements in collaboration with developers of the system to downsize the perceived risks of transactions and build the confidence of customers to rely on CBDC payment systems. Efforts should be made to lower the performance risk related to CBDC payment. In order to establish that technologies around CBDC payment are reliable, online booking platforms in the tourism industry could, for example, introduce a video clip that demonstrates how travelers can use CBDC payment without problems.

Second, this study showed that low consumer innovativeness (Aldás-Manzano et al., 2009) and a lack of past experience (Kaur and Arora, 2020) negatively affect customers’ attitudes and their intention to adopt the CBDC payment system. This implies that there is a need to promote the payment system and increase customers’ technical skills. This can be done by financial institutions, and other end users such as tourism organizations, by producing short, free podcasts that create awareness and improve the technical skills of users such as tourists. Therefore, CBDC institutions need to provide customers with more and correct information on CBDC so that customers are more exposed to the payment system and can perceive reduced risk.

Third, the results of this study can contribute to the enhanced use of CBDC payments in the travel and tourism industry, which is currently under high pressure as a result of the effects of the COVID-19 pandemic. The current study identified some determining factors of this contactless system, which can enhance international or border-cross travel (Gross et al., 2021) and reduce the risk of the pandemic (Rennie & Steele, 2021) through the provision of contactless services (Seth, 2022). The existing literature has also shown that CBDC is advantageous for ease of use, safety, speed, stability, and reliability (Bindseil, 2019; Radic et al., 2022) in financial transaction processes. The findings of this study also provide information for central banks worldwide to help with the issue of CBDC by integrating it with their own monetary policies and transaction systems. Furthermore, this study provides evidence of how structural assurance, or regulations, promises, guarantees, and contracts (Shapiro, 1987a; Zucker, 1986) affect attitudes toward adopting the CBDC payment system. To effectively implement this payment system, financial institutions need to build systems their customers can trust and use payment methods that support perceptions of low risk. The adoption of the system will depend on diverse factors, including the transparency of transaction, the protection of privacy, the domestic political situation, legal and regulatory support, and customers’ trust in the economic system (Wang, 2022).