PiChain Ghana

Financial literacy, behavioral traits, and ePayment adoption and usage in Japan

Abstract
This study investigates how financial literacy and behavioral traits affect the adoption of electronic payment (ePayment) services in Japan. We construct a financial literacy index using a representative sample of 25,000 individuals from the Bank of Japan’s 2019 Financial Literacy Survey. We then analyze the relationship between this index and the extensive and intensive usage of two types of payment services: electronic money (e-money) and mobile payment apps. Using an instrumental variable approach, we find that higher financial literacy is positively associated with a higher likelihood of adopting ePayment services. The empirical results suggest that individuals with higher financial literacy use payment services more frequently. We also find that risk-averse people are less likely to adopt and use ePayment services, whereas people with herd behavior tend to adopt and use ePayment services more. Our empirical results also suggest that the effects of financial literacy on the adoption and use of ePayment differ among people with different behavioral traits.

Introduction
In recent decades, the progress of financial technology (fintech) has helped consumers access financial markets and services more easily than before (Kou et al. 2021). As a result, the number of financial products offered has increased, and simultaneously, such products have become more complicated. The literature shows that adopting fintech services helps increase financial inclusion and improve consumers’ financial well-being (Suri 2017). Among fintech services, consumers, firms, and governments commonly use electronic payment.Footnote1 Adopting ePayment services helps develop a cashless economy, reduces business costs for all parties, and increases economic efficiency (Rogoff and Rogoff 2017; Xu et al. 2019). Using ePayment services helps reduce the estimated gross domestic product (GDP) losses due to the use of cash: 0.12% of the GDP in Germany (Cabinakova et al. 2017); 0.45% of the GDP in Canada (Kosse et al. 2017); 0.29% of the GDP in Japan (Fujiki 2022).

While fintech services play an increasingly important role in individuals’ financial activities, they entail numerous risks, including both traditional financial risks and new Internet-related risks (Li et al. 2021; Morgan et al. 2019). Thus, to overcome these risks, consumers should have adequate knowledge to make financial decisions about adopting and using fintech services. However, among the various studies examining the factors that affect the adoption of ePayment services (Jack et al. 2013; Afawubo et al. 2020), only a few have explicitly investigated the role of financial literacy in the adoption of fintech products (Li et al. 2020; Morgan and Long 2018; Foster and Johansyah 2021). Moreover, these studies provide mixed empirical results on the role of financial literacy in ePayment and fintech adoption, while others conclude that financial literacy is positively associated with fintech/ePayment adoption (Morgan and Long 2018; Foster and Johansyah 2021), while others find a negative relationship (Li et al. 2020). In addition, previous studies show that the effect of financial literacy on financial decisions may vary by sex, rural–urban location, or education level (Xu et al. 2022). Moreover, the relationship between behavioral traits such as risk aversion or herd behavior and the heterogeneity of the effects of financial literacy on fintech adoption are yet to be studied. Therefore, answering questions such as whether the effect of financial literacy differs for individuals with different behavioral traits may provide empirical evidence to help develop targeted financial literacy programs.

This study attempts to fill these gaps in the literature by examining the effects of financial literacy on the adoption and use of ePayment services in Japan. First, we construct a financial literacy index that reflects financial knowledge, financial behavior, financial attitudes, and knowledge of the practical management of financial assets. We then investigate the relationship between this financial literacy index and the extensive and intensive use of two types of payment services: chip-based electronic money (e-money)Footnote2 and mobile payment apps.Footnote3 We also examine how two behavioral traits (herd behavior and risk aversion) may affect the relationship between the financial literacy index and the adoption and usage of ePayment services. Finally, we use the instrumental variable (IV) approach to address the potential endogeneity of financial literacy. The Bank of Japan collected the dataset in 2019, with a sample size of 25,000 adults. The sample was designed to represent the adult Japanese population.

Our empirical results show that higher financial literacy is positively associated with a higher likelihood of adopting and using both ePayment services. The results also suggest that risk aversion is negatively associated with the adoption and usage of ePayment services, whereas herd behavior is positively correlated with ePayment adoption and use. We also find that the effects of financial literacy on the adoption and usage frequency of ePayment services differ between individuals with and without risk aversion. For those with herd behavior, higher financial literacy tends to be associated with a lower adoption rate of mobile payment apps but does not affect e-money adoption.

This study extends the existing literature in several ways. First, it is related to the growing literature on the role of financial literacy in adopting and using fintech services. To the best of our knowledge, few studies have examined the effect of financial literacy (or financial knowledge) on the adoption and use of fintech services (Li et al. 2020; Morgan and Long 2018; Foster and Johansyah 2021), and the results are mixed. For example, Morgan and Long (2018) find that financial literacy positively correlates with awareness of fintech services in Lao PDR. In addition, Foster and Johansyah (2021) find that financial literacy positively correlates with adopting chip-based e-money in Indonesia. By contrast, Li et al. (2020) find a negative relationship between financial knowledge and ePayment adoption in the US. Our study provides further evidence of the effects of financial literacy on the adoption and intensity of fintech use.

Second, this study contributes to the literature on consumer adoption of technology. While previous literature has used several theories of technology adoption, such as the Diffusion of Innovation (DOI), Technological Adoption Model (TAM), and Unified Theory of Acceptance and Usage of Technology (UTAUT2), none of these models explicitly consider the role of financial literacy and behavioral traits.Footnote4 Moreover, only a limited number of empirical studies have examined how these two factors affect the adoption of fintech products (Li et al. 2020). Our study adds to the literature by examining the effects of financial literacy, risk aversion, and herd behavior in Japan.

Third, our study contributes to the literature on the heterogeneity of the effects of financial literacy. Specifically, we investigate whether financial literacy affects individuals with different behavioral traits. While previous literature has demonstrated that behavioral traits and financial literacy play important roles in individuals’ financial decisions (Almenberg and Dreber 2015; Gathergood and Weber 2017; Grohmann 2018; Hsiao and Tsai 2018; Van Rooij et al. 2011), only a few studies have examined the heterogeneity of the effect of financial literacy across individuals with different behavioral traits. For example, Jiang et al. (2020) show that the effect of financial literacy on financial well-being is greater among risk-averse individuals. This study extends this strand of the literature by examining how behavioral traits may affect the relationship between financial literacy and fintech adoption and use.

Japan is an interesting case study to examine the role of financial literacy and fintech adoption. On the one hand, as a highly developed economy, Japan has adequate foundations (in terms of financial regulation, financial structure, and technical knowledge) and policies (e.g., the Japanese government’s recent policy to subsidize cashless payments) to promote the use of fintech (Fahey 2019). On the other hand, the adoption of fintech is limited, especially compared to China and South Korea. Ernst and Young (2019) show that the fintech adoption rate in Japan is low (approximately 34% in 2019 vs. 87% in China, 67% in South Korea, and 46% in the US). Moreover, the gap in the adoption rate between Japan and the global average widened from 19 percentage points in 2017 to 26 percentage points in 2019 (Ernst and Young 2019). According to the Central Council for Financial Services Information (CCFSI; 2016), the financial literacy of the Japanese population is slightly lower than that of Americans, Germans, and the British. Therefore, researchers and policymakers want to understand the relationship between financial literacy and ePayment adoption in Japan.

The remainder of this paper is organized as follows. First, we briefly review the literature and propose testable hypotheses. “Empirical approach” section presents the data, empirical approach, and descriptive data. “Estimation results” section presents the empirical results. Finally, “Discussion and concluding remarks” section discusses the results and their theoretical, practical, and policy implications.

Literature review and hypothesis development
Financial literacy and ePayment adoption
Previous studies have treated fintech services as technological innovations and analyzed their adoption using various theories, including the DOI theory (Rogers 2003), Initial Trust Model (Kim and Prabhakar 2004), TAM (Davis 1989), UTAUT Model (Venkatesh et al. 2003), and UTAUT2 Model (Venkatesh et al. 2012). Additionally, previous studies have examined other factors that explain the usage of fintech services. These studies show that an individual’s perceived usefulness and ease of use, performance expectancy, perceived risk, perceived trust, and facilitation conditions are major determinants of fintech service adoption (e.g., Laukkanen and Pasanen 2008; Baptista and Oliveira 2016; Malaquias and Hwang 2016).

Financial literacy may also affect the adoption and use of ePayment services. Given the increasing responsibilities that consumers need to assume in planning for retirement and using credit, there is an increasing focus on whether consumers are sufficiently well-equipped to deal with financial matters. Financial literacy is the knowledge and understanding of financial concepts used to make effective financial choices.Footnote5 Lusardi et al. (2017) develop a theoretically augmented stochastic life cycle model that endogenizes the decision to acquire financial literacy. The model predicts that different levels of financial literacy account for sizable differences in wealth holdings across education groups. Many studies have empirically shown a strong correlation between financial literacy and financial behavior, such as daily financial management skills, participation in financial markets, investing in stocks, and engaging in precautionary savings (Hilgert et al. 2003; Christelis et al. 2010; van Rooij et al. 2011; de Bassa Scheresberg 2013; Yoshino et al. 2017). Research has corroborated these results in countries, such as Japan (Yoshino et al. 2017), Cambodia, Laos, Vietnam (Morgan and Long 2019, 2020), and Bangladesh (Hasan et al. 2021).

Research suggests that high financial literacy motivates individuals to process information, set up a business, acquire new financial knowledge, and search for what is available in the market. These characteristics pave the way for the adoption of new services such as fintech. Financial literacy affects the adoption of fintech services through two channels. First, higher financial literacy lowers the information costs incurred owing to the use of a new financial product. The literature shows that the likelihood of using a financial product, especially a risky one, is crucially affected by the costs and benefits of acquiring information (Hsiao and Tsai 2018). Therefore, lowering the cost of acquiring such information influences the decision to use that product (Vissing-Jorgensen 2003; Guiso and Jappelli 2005). Second, financial literacy may help mitigate a customer’s risks when using fintech services. Individuals with higher financial literacy are more likely to avoid such risks because they have a higher propensity to choose suitable financial products (Agarwal et al. 2020; Gathergood and Weber 2017), detect risky services (Engels et al. 2020), and detect fraud (Wei et al. 2021). Several studies have examined the effects of financial literacy on the adoption of fintech services, including ePayment services (Li et al. 2020; Morgan and Long 2018; Foster and Johansyah 2021; Lo Prete 2022), mobile money, digital banking (Yates 2020; Frimpong et al. 2022; Chen and Xiang 2021), peer-to-peer (P2P) lending (Gonzalez 2022), or holding cryptocurrencies (Fujiki 2020). While some studies argue that individuals with higher financial literacy might recognize the risks they may incur when using fintech services, they are less likely to adopt fintech services (Li et al. 2020; Chen and Xiang 2021). In contrast, most studies show a positive correlation between financial literacy and the adoption and usage of fintech services. Based on the above analysis, we propose the following hypothesis:

H1
Individuals with a higher level of financial literacy are more likely to adopt and use fintech services than those with a lower level of financial literacy.

Financial literacy, behavioral traits, and ePayment adoption and use
The literature shows that behavioral traits play an important role in individuals’ financial decisions. Risk attitude is an essential factor influencing various personal financial decisions (Snelbecker et al. 1990). Risk attitudes are important in financial planning models and consumer decision-making frameworks (Han et al. 2019). Ajzen and Fishbein (1975) posit that risk attitudes affect individuals’ beliefs and, ultimately, influence their decisions. Therefore, individuals with different levels of risk aversion may exhibit different financial investment behaviors. Previous studies have shown that risk preferences can influence risky financial decisions, such as participation in stock markets or holdings of risky assets (Badarinza et al. 2016; Barberis et al. 2006). Therefore, risk preferences may explain the differences in the uptake of fintech services among individuals. Lin et al. (2013) examine the online P2P lending market and find a significant positive correlation between risk-loving attitudes and Internet financing volume. Similarly, Han et al. (2019) show that financial knowledge and risk attitudes are strongly associated with participation in P2P lending in China.

Among risk-averse individuals, those with a higher level of financial literacy may not be as affected by risk aversion in their decision to adopt fintech services compared with those with a lower level of financial literacy. Thus, the effect of financial literacy may be greater among risk-averse individuals. Morgan et al. (2019) argue that fintech services entail numerous risks that are more diverse and harder to spot than those associated with traditional financial products and services. Studies have shown that education may encourage risk-averse individuals to take risks (Dohmen et al. 2005; Hryshko et al. 2011). Moreover, Jung (2015) argues that higher education might diminish risk aversion through a better understanding of how to deal with risks. Therefore, we propose the following four hypotheses.

H2a
The effect of financial literacy on the adoption of e-money is higher for risk-averse people.

H2b
The effect of financial literacy on the adoption of mobile payment apps is higher for risk-averse people.

H2c
The effect of financial literacy on the usage frequency of e-money is higher for risk-averse people.

H2d
The effect of financial literacy on the usage frequency of mobile payment apps is higher for risk-averse people.

Herd behavior is another trait widely studied in financial markets, especially in stock markets. Zhang and Chen (2017) define herding as “individuals doing what other individuals are doing, even when their information suggests them to do something different from the others.” Individuals with herd behavior may ignore their viewpoints and expertise, regardless of whether they are valid, to make decisions consistent with the herd (Devenow and Welch 1996). In such cases, failure is the result of a mistake made by the herd rather than one of the herd members (Ahmad and Mahmood 2020). The literature shows herd effects are strong determinants of portfolio choice and stock market participation (Hong et al. 2004; Brown et al. 2008; Van Rooij et al. 2011). Since individuals with herd behavior tend to ignore their expertise and judgment and follow the herd in making decisions, the effect of financial literacy on the adoption and usage of fintech services may differ for individuals with herd behavior, depending on the herd’s decisions. Widely used fintech services (e.g., e-money) may be adopted by individuals with herd behavior to the same extent as those without herd behavior. Early adopters can obtain tips, advice, or information about fintech services. Thus, the effect of financial literacy may not differ between individuals with and without herd behavior. However, for new services only adopted by a limited number of people (e.g., mobile payment apps),Footnote6 the adoption is likely to be lower for those with herd behavior. Regarding usage frequency, whether the effect of financial literacy differs between individuals with and without herd behavior may depend on the herd size. If a herd is large, financial literacy may not vary significantly. Therefore, we propose the following hypotheses:

H3a
The effect of financial literacy on the adoption of e-money is not different between individuals with and without herd behavior.

H3b
The effect of financial literacy on the adoption of mobile payment apps is lower among individuals with herd behavior.

H3c
The effect of financial literacy on the usage frequency of e-money is not different among individuals with herd behavior.

H3d
The effect of financial literacy on the usage frequency of mobile payment apps is lower among individuals with herd behavior.

Empirical approach
Data source
The Bank of Japan’s Financial Literacy Survey is an online questionnaire survey conducted to understand the current state of financial literacy, that is, the financial knowledge and financial decision-making skills of individuals aged 18–79 years in Japan, chosen in proportion to Japan’s current demographic structure (CCFSI 2016, 2019). The first survey was conducted in 2011 by the CCFSI, followed by the second and third rounds in 2016 and 2019, respectively. Twenty-five thousand individuals participated in the 2016 and 2019 surveys. This study uses data from the 2019 survey because information on fintech usage was available only in this survey.

Questions on financial literacy include true/false questions on “financial knowledge and financial decision-making skills” and “characteristics of behavior and attitude.” Approximately half of the questions are similar to those in the surveys conducted by the U.S. Financial Industry Regulatory Authority Investor Education Foundation and the Organization for Economic Co-operation and Development (OECD; CCFSI 2016, 2019). Information on sex, age, place of residence, occupation, annual income, and experience of participating in financial education is collected. Finally, information on using fintech services and products is also collected.

Empirical approach
To quantify the effect of financial literacy on the decision to adopt ePayment services, we estimate the following equation:

FTi=β0+β1FLi+β2Behavei+β3Xi+ηi,
(1)
where the dependent variable FTi indicates whether individual i uses an ePayment service (e-money or mobile payment app). FLi is the financial literacy index value of individual i; Behavei is a set of two behavioral traits (risk-averse or herd behavior) of individual i; Xi is a vector of the control variables; and ηi is the error term. We use linear probability regression to estimate Eq. (1) because we use the IV approach, and the linear probability regression allows us to test the validity of the IVs in a straightforward manner.

We also analyze how financial literacy affects the usage frequency of e-money or mobile payment apps. As described below, because our dependent variable depicts the ordering of the usage frequency, we estimate an ordered probit model in which the dependent variable takes one of the four intensity levels. We assume the existence of a latent continuous exact variable FT∗i that determines the order of the intensity of using ePayment services. The following equation characterizes the underlying:

FT∗i=α0+α1FLi+α2Behavei+α3Xi+ϵi,
(2)
where FT∗i is the observed category of a response corresponding to the ith order of ePayment usage intensity, FLi is the financial literacy index, Behavei is a set of two behavioral traits of individual i, andXi is a vector of the control variables. All the independent variables in this equation are the same as those in Eq. (1).

We further examine whether financial literacy mitigates behavioral traits by augmenting Eqs. (1) and (2) as follows:

FTi=γ0+γ1FLi+γ2Behavei+γ3Behavei∗FLi+γ4Xi+εi
(3)
and

FT∗i=θ0+θ1FLi+θ2Behavei+θ3Behavei∗FLi+θ4Xi+μi.
(4)
Our coefficients of interest in these two equations are γ3 and θ3 that indicate how the financial literacy score could change the effects of behavioral traits on the adoption and usage frequency of ePayment services FTi and FT∗i, respectively. All variables are measured as previously described.

The coefficient estimates of the financial literacy variable may be biased due to reverse causality (i.e., using ePayment services may help adopters improve their financial literacy) and omitted variables, such as unobserved motivations and abilities, which affect both the adoption of ePayment services and financial literacy. To address these potential endogeneity problems, we use two IVs. First, following Fernandes et al. (2014) and Murendo and Mutsonziwa (2017), we use the mean financial literacy score at the prefectural level as the first instrument for individual financial literacy. Second, we use information from the question “Are you aware of the lowering of the age of adulthood?” In 2018, the Japanese government approved an amendment to the country’s civil code, which reduced the age of legal adulthood by two years to 18, effective April 1, 2022. This change is essential for two reasons. First, it significantly affected youths, their guardians, and those who cared for them. Second, it changed a law that had been in effect for more than 140 years. The rationale for using this instrument is that those aware of this change are more likely to be interested in accumulating socio-economic knowledge. As improving financial literacy is a form of human capital accumulation (Lusardi and Mitchell 2014), awareness of changes in the age of adulthood is expected to correlate positively with financial literacy. Simultaneously, this may not necessarily correlate with adopting and using fintech services. Although this variable has not been used in other studies, it aligns with those that instrument financial literacy by acquiring economic and financial knowledge (Hsiao and Tsai 2018).

Variable construction
There are two independent variables in our analysis. We construct our dependent variables using information from these questions: “How often do you use e-money?” and “How often do you use [mobile] payment apps?” The answers to these two questions are “Almost every day,” “About once a week,” “About once a month,” “I hardly ever use,” and “I don’t use.” The first dependent variable is related to the adoption of ePayment services. This binary variable takes a value of one if a person uses either e-money or mobile payment apps at least once a month and zero otherwise. The second dependent variable is the usage frequency of e-money or mobile payment apps. Using the same information, we construct the following four levels of intensity of use for each product: (i) daily, (ii) once a week, (iii) once a month, and (iv) no use.

We use a set of 25 questions from the survey to calculate a financial literacy index. This set consists of 18 questions on financial transactions; economic and financial knowledge; and knowledge of wealth-building, insurance, lending, and borrowing. In addition, there are seven questions on financial decision-making skills, such as household budget management, life planning skills, and outside expertise.Footnote7 The financial literacy score is calculated as the number of correct answers, ranging from 0 to 25. For ease of interpretation, we calculate the z-score of the financial literacy score.

We examine two types of behavioral traits: risk aversion and herd behavior. For risk aversion, we use information from the question “Suppose you invested 100,000 yen, there is an equal probability that you would either gain 20,000 yen or lose 10,000 yen. What would you do?” People are viewed as risk-averse if they answer “No, I would not invest.” For herd behavior, we use information from the question “How much do you agree or disagree that the statement ‘When there are several similar products, I tend to buy what is recommended as the highest-selling product rather than what I actually think is a good product’ applies to you personally?” People are considered to have herd behavior if they answer “Very much agree” or “Somewhat agree.” If they answer “Neutral,” “Somewhat disagree,” or “Disagree,” they are considered to not have herd behavior.

Xi is a vector of control variables that may influence the adoption of ePayment services. The first set of control variables is derived from the implications of the UTAUT2 Model. We control for individuals’ perception of the availability of stores that accept ePayment services (as a proxy for the facilitation condition), the time to process payment (as another proxy for the facilitation condition), the perception regarding managing private information and tools to control overpayment (as proxies for perceived risks), and the perception of cash usage (as a proxy for performance expectancy) based on their responses to the relevant questions in the survey. We also include individual characteristics such as age group, sex, level of general education (dummies), financial education, income (dummies), occupation, and frequency of reading financial and economic news.Footnote8

Descriptive analysis
Table 1 presents the descriptive statistics of the sample. Owing to missing information, our analysis is based on 24,516 observations. The average financial literacy score is 14.1 (standard deviation: 6.9). Approximately 35.8% and 8.0% of the respondents in our sample use e-money and mobile payment apps, respectively, and 38.0% use at least one ePayment service. Regarding behavioral traits, 77.3% and 16.7% of the respondents are classified as risk-averse and exhibiting herd behavior, respectively. Surprisingly, nearly 25% of the respondents are fully satisfied with cash payments. Only 18.5% think that private information is not secure, and 10% state that tools to prevent the overuse of ePayment services are lacking. More than 23% of the respondents report that a limited number of stores accept ePayment services.

Reference: https://jfin-swufe.springeropen.com/articles/10.1186/s40854-023-00504-3

Leave a Comment

Your email address will not be published. Required fields are marked *