“Do socio-economic factors impede the engagement in online banking transactions? Evidence from Ghana”

Researchers have long pondered on the online banking transaction adoption. Some of these studies focus primarily on the motivating factors that affect customers’ intention to adopt/accept these services (technologies). However, research into the constraining factors, in particular socio-economic factors, barely exist in the literature, especially in the context of sub-Saharan Africa. Against this background, the paper seeks to fill in this gap by: (1) assessing the socio-economic factors impeding the engagement of e-banking transactions among retail bank customers in Ghana, and (2) examining the moderating effect of ‘customer experience of Internet’ on the identified factors that inhibit the engagement in online banking in Ghana. The paper used a quantitative research approach to obtain data from two leading Ghanaian banks. Out of the 450 questionnaires distributed, 393 were valid for analysis. Data were analyzed with the aid of PLS-SEM (partial least squares and structural equation modeling). Findings revealed that perceived knowledge gap and the price of digital devices were directly important to the intention to disembark on e-banking transactions among Ghanaian bank customers. Whilst customer experience (frequent use of the Internet), as a mod- erator variable, has a significant effect on the interaction between perceived knowledge gap and the intent to disembark on e-banking transactions; and finance charges and the intent to disembark on e-banking transactions. Study implications and directions for future research are discussed in the paper.


INTRODUCTION
The advent of Financial Technology (FinTech) in the 21st century is rapidly shifting the overall pattern of businesses (Nyangosi, Sharma, 2020). Whilst the society and its socio-economic environment of individuals continue to be a key determinant of the development of the Internet usage (Saleem & Higuchi, 2014) and for that matter, ICT. FinTech has come to stay. Owing to that, the advent of e-commerce is one of the utmost profitable changes that has transformed the trading of products and services over the Internet, especially within the banking sector (Boateng, Adam, Okoe, & Anning-Dorson, 2016; Miniwatts Marketing Group, 2014). As a matter of fact, research on Internet banking has gained the attention of many notable stakeholders, including bankers, customers, financial regulators among others, so as to examine the repercussions of engaging in such kinds of transactions. This paper defined online banking as a way/process of banking in which all financial transactions are performed electronically over the Internet, either from the banking hall or at the comfort zone of a user. In other words, online banking (e-banking) enables banks, customers and other parties to manage accounts and undertake/make account transactions directly/indirectly with the bank in question via the Internet, this, according to Oertzen and Odekerken-Schröder (2019), Ofori, Boateng, Okoe, and Gvozdanovic (2017), is described as Internet banking. However, unwillingness on the part of a bank customer to engage in the Internet banking platform for a successful and easier financial transaction remains a major concern amongst several financial institutions, especially in developing countries (Nasri, 2011;Yiu, Grant, & Edgar, 2007).
Research has shown that online transaction progress in the third world countries, in particular, in the sub-Saharan Africa, is comparatively very low (Jibril, Kwarteng, Pilik, Botha, & Osakwe, 2020; Sharma, 2011). Therefore, this study spans socio-economic factors that influence the unwillingness to engage in e-banking transactions in a developing country, particularly in Ghana. The study argues that, given the rate of Internet penetration and diffusion in the market arena, the effect of customer initiation in adopting online transactions over the web may be significant to the banking sector and that some aversion to adopting the technology may, therefore, offer and explain how customers perceive online banking in a lower penetration Internet community where this study barely exists.
Until recently, studies on e-banking or online banking transactions have focused more on the motivating factors that drive the intent to adopt e-banking transactions, rather than the de-motivating factors that propel the action (see Rahi, Ghani, & Ngah, 2019;Singh & Malhotra, 1970; Tarhini, El-Masri, Ali, & Serrano, 2016). However, in under-developed countries like Sub-Sahara economies, the factors associated with the former have not been explored and are still rudimentary in the literature in terms of both the engagement and disengagement situations. In this light, this study describes socioeconomic factors as the distinction between societal problems and economical situations. While societal factors here are associated with the societal impediment in line with the tendency to adopt online banking transactions (income, education, community safety, and social support), economic conditions refer to "situations associated with the wellbeing of the retail customer averting him/her to adopt an online banking transaction (e.g., income, employment, etc.)" (Nabareseh, Osakwe, Klímek, & Chova, 2014). Therefore, to examine the effects of both societal problems and economic scenarios averting a customer to engage in online banking transactions in a developing country while ensuring the willingness to make banking transactions via online, this research (1) investigates the socio-economic factors obstructing the engagement of online banking transactions based on the literature related to digital device prices, infrastructure, perceived financial cost/charge, and perceived knowledge gap and customer experience, and (2) examines the moderating effect of 'customer experience of Internet' on the identified factors that inhibit the engagement of online banking in Ghana.
The study informs e-banking practitioners (about the constraints) that potentially frustrate customers about what online banking entails and its relevance to banking institutions. Hence, this study propels financial institutions to strategically strengthen the usage and continuity of technological advancement as socio-economic factors (constraints) in the retail banking sector are concerned.

Infrastructure
Facilitating condition, such as investment of infrastructure, is commonly noted as an important key driver of socio-economic development of every nation (Shankar & Meyer, 2009), while studies have shown that the "quality, quantity, and accessibility of economic infrastructure in developing countries lag considerably behind" (Boateng et al., 2008) as compared to the developed world (see Jibril et al., 2020;Nwaiwu et al., 2020). In light of this, scaling up infrastructure investment or support systems is largely described as a key driver to speed up socio-economic growth and development, especially in the undeveloped economies (Awh & Waters, 1974). In fact, many developing countries have been scaling up an infrastructure support system, mostly through public spending, as well as growing participation of the private sector (Nabareseh et al., 2014). Therefore, infrastructure gaps are still large in pursuing sustainable technological development, such as smart cities, e-payment, e-revenue mobilization, among others, and filling these gaps would help in the medium to longterm sustainability of innovation diffusion and successful technology adoption by the citizenry.

Price of devices
This refers to the unit price of the digital device in the electronic market. In other words, it means the value of the amount paid to acquire any electronic device (Esteve & Machin, 2007) in order to have access to online transactions. In the developing world, since many citizens are leaving below the poverty line (Ali, 2016) or within the low-income bracket (Ali & Langendoen, 2007), buying sophisticated devices (such as smartphones, tablets, notebook, etc.) at higher prices makes it difficult to access such income brackets. It is also important to acknowledge the concept of 'consumer rationality' in the field of consumer behavior. This concept reiterates the principle of demand, which suggests that low-income earners demand less of a product (good) when the price is high, and vice versa.

Perceived financial cost
Financial cost towards the adoption/engagement in online transactions refers to expenses associated with the online transaction. Thus, these are expenses incurred in buying or selling a good or service or even using an online platform for other intended purpose (Stein et al., 2005; Wu et al., 2014) such as money transfer and interbank transfer/deposit. Yu (2012), in his work, argued that potential users of electronic banking systems/transactions are quite skeptical with regard to the potential charges that may be attached due to services rendered by a bank. Nonetheless, since first-timers anticipate that they would be deducted from their deposit or whichever possible, this perceived transaction cost deters potential users from engaging in 'loos game transaction', especially in a low-income bracket group (Amegbe & Osakwe, 2018).
As a matter of fact, bank customers (both online and off-line) are mostly not satisfied when it comes to a levy on transaction-related activities, especially when the perceived cost does not correspond to the expected service provided. However, as the new technology (e-banking transaction) comes with an associated cost, hence, they (banks) transfer the service charge/ cost on to the users (Sathye, 1999;Yiu et al., 2007), since the cost is a shared responsibility to both a bank and a customer in question (Thambiah et al., 2010). This suggests that a shift in charges to bank clients can potentially discourage them to engage in an online banking transaction.

Perceived knowledge gap and customer experience
The knowledge gap indicates the view that a person with a higher status in an economy/society is privy to information offered by the mass media, and vice versa (Wei & Zhang, 2008). As a result, this variation in the society leads to an increased gap in knowledge between these two segments (higher and lower socioeconomic status).

AIMS
The purpose of this paper seeks to fill in the gap by: 1) assessing the socio-economic factors impeding the engagement in e-banking transactions among retail bank customers in Ghana; and 2) examining the moderating effect of 'customer experience of the Internet' on the identified factors that inhibit the engagement in online banking in Ghana.
To achieve the above goals, the paper puts forward the following hypotheses: H1: Lack of infrastructure support in a developing country positively predicts the unwillingness of bank customers to engage in online banking transactions.
H2: High price of digital devices (tablet, smartphone, etc.) in a developing country with respect to low income earners positively predict the unwillingness of consumers to engage in online banking transactions.
H3: Financial (operating) costs charged by banks toward online banking transactions potentially trigger customers' unwillingness to engage in/adopt online banking transactions.
H4: Perceived knowledge gap in the Internetusage environment directly predicts customers' unwillingness to engage (adopt) in online banking transactions.
H5a: Customers' experience in Internet usage significantly moderates the relationship between perceived Internet knowledge gap and unwillingness to engage in online banking transactions.
H5b: Customers' experience in Internet usage significantly moderates the relationship between perceived financial cost/charge and unwillingness to engage in online banking transactions.
To conclude, the paper has summarized the literature review in a proposed research model (see Figure 1) for further investigation and validation.

Measurement of constructs
It is important to note that the measures of the constructs in this paper were adapted from the existing literature. Again, it is expedient to note that the indicators (a clear qualitative statement) were measured using a five-point Likert scale, with 1 = completely disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = completely agree.  (Ajzen, 1991;Ali, 2016). Also, see Table 2 for the definition of the variable for the respective constructs.

Data collection and analytic technique
This paper used both random and non-randomized sampling techniques for data extraction. First, a non-randomized sampling technique (non-probability) was adopted to select target banks (unit of analysis lected for the study irrespective of whether he/she is an off-line or on-line bank customer.
Data was finally collected using a structured questionnaire (both the intercept approach and online survey. With the intercept, it means the customers were trapped at the bank's premises, while the online survey link was sent to participants who requested it mainly because they were not ready at the time the field officials intercepted them at the designated points. Out of the 450 questionnaires distributed, 393 were valid for analysis. Though study subjects were pre-qualified to ensure their knowledge of online banking services with respect to their level of understanding of socio-economic factors affecting the engagement in Internet banking transactions. Data collection was carried out between November 2019 and February 2020. On average, it took seven minutes to fill the questionnaire. Table 1 shows a socio-demographic profile of the respondents to this study. To analyze data, the partial least squares (PLS) approach is used since it focuses on the ex-

Test of the common method bias (CMB)
Following the earlier works by P. M. Podsakoff et al. (2003), the paper emphasizes in the current survey questionnaire, especially in the header section (at the beginning of the ques-tionnaire), that 'there are no right or wrong answers to the questions asked'. The researchers proceeded to assure the respondents of their anonymity and were asked to freely quit from answering the questionnaire at their discretion.
Regarding the qualitative measures used to address potential concerns about common method bias (CMB), the paper took inspiration from the recent suggestions in the PLS-SEM literature from Kock and Hadaya, (2018), hence, the research employed a full collinearity approach, in particular, a variance inflation factor (VIF), to examine the evidence of CMV. The results of this post-hoc estimate show that common method bias (CMB) is not a major concern since the computed VIFs (see Table 2) are less than three  (3) taking into account the maximum threshold of ten (10). Therefore, in this analysis, the concerns about CMB are minimal, hence the potential CMB concerns are low.

Model assessment
According to the relevance and key recommendations in the PLS literature by pioneering scholars (see Hair et al., 2014;Hair et al., 2017), the paper first assessed the convergent validity and subsequently the discriminant validity of the outer model using the SmartPLS 3.2.9 version of statistical software. According to the estimations from the rigorous analysis, the construct indicators' loadings are relatively larger than 0.7, considering the threshold of 0.6 (see Bagozzi & Yi, 1988). Table 2 has the constructs exhibiting high composite reliability scores, with average variance extracted (AVE) scores of constructs exceeding 0.5, whiles the Cronbach alpha of the research constructs exceeded 0.7. Hence, the study constructs have exhibited convergent validity (see Table 2).
Moreover, to establish the discriminant validity, on the other hand, Fornell and Larcker's (1981) criterion was used to assess the presence of discriminant validity among the research constructs (Henseler et al., 2015). Findings from Fornell-Lacker's criterion indicated that constructs satisfy both basic and stringent assumptions and this, therefore, establishes discriminant validity. In particular, the values in the diagonal (in bold) of the Fornell-Lacker's table (see Table 3) indicate AVEs of the measured constructs and met the threshold of 0.5 (Fornell & Larcker, 1981).

Structural model
After assessing the initial step in the PLS analysis, the structural parameters of the fitted model were evaluated. This study used different quality criteria to assess the model, namely adjusted R 2 magnitude (strength) and statistical significance of path loadings. Meanwhile, far from the basic requirement in the current PLS reporting (Hair, Risher, Sarstedt, & Ringle, 2019), the standardized root mean squared residual (SRMR) = 0.087, and is acceptable if < 0.1; this indicates a good fit between the data and the research model. Concerning the direct hypothetical relationships, the price of a digital device (PDD) and perceived knowledge gap (PKG), thus, confirming H3 and H4, respectively, has a significant relationship affecting the engagement of e-banking transactions among the sampled respondents. Whilst infrastructure support and perceived financial charges/costs (PFC) were not directly significant towards the intention to disengage in e-banking transactions, which disproved H1 and H4, respectively (see Table 4).
However, the moderation analysis of the structural model shows a significant relationship. Importantly, the paper considered customers' experience, in particular, their 'frequent use of the Internet' as a moderating variable, to determine the moderating effects on the interactions between PKG → INTENT and PFC → INTENT. Interestingly, there was a significant relationship between the two interactions, thus, confirming H5a and H5b, respectively, at a bootstrapping sample size of 999 with a T-test greater than 1.96 or P-value less than 0.05.
In a nutshell, Table 4 presents the results of the structural model and, accordingly, the research findings. R 2 shows that about 45% of variance in the dependent variable (intention) is explained by the independent variables in the research model.

DISCUSSION
Based on the results, the paper concludes that there is a significant impact of socio-economic factors on the advent of Fintech, in particular, e-banking transactions. It worth noting that FinTech would persistently emerge as a robust and global e-commerce platform/App that facilitates banks and other service providers, especially in the financial industry. This will help to initiate and offer users alternative consumption of innovative products and services (Denny, 1970 where Ghana cannot be isolated so far as the study of technology adoption in a developing country is concerned. Also, Wei and Zhang (2008) argue that as per the rate of low-income earners and a high poverty rate among people from the less developed countries (LDC's), their purchasing power regarding the demand for electronic goods is relatively low, thus, affecting their inclination for digital devices such as smartphones, notebooks, etc. This assertion is in tandem with the current evidence from Ghanaian bank customers. Hence, the findings of this research largely reinforce the impression that these constructs potentially influence the engagement/disengagement in e-banking transactions, especially among low Internet penetration settings .
Interestingly, INFTRAS and PFC (H1 and H3 respectively) showed an insignificant determinant towards customers' inability to hook up on e-banking transactions. Consequently, the study suggests that the customers (thus, immediate beneficiaries) of the online banking system believe that there is an expected cost to be incurred while using such a service. Therefore, to their (customers) best of knowledge, the said operating charges are not a significant impediment in their quest to engage/disengage in e-banking transactions. This research is in contrast to the works of Awh and Waters (1974), Nabareseh et al. (2014), and Shankar and Meyer (2009). As per the analysis, respondents believe that the lack of an infrastructural support system plays a crucial role in the individuals' willingness to engage in online banking transactions.
Regarding the moderation analysis (H5a and H5b), the result prove that there is a significant moderating effect on the interaction between PKG and INTENT to disembark on e-banking transactions given customer experience (frequency of the Internet use) as a moderating variable (Mattila & Wirtz, 2001). Again, holding 'frequency use of Internet' as a moderator, there was a significant relationship between the interaction of PFC and INTENT to disembark on e-banking transactions among bank customers in Ghana. This sug- In theory, this research developed a framework to augment the extant literature by offering some empirical evidence of socio-economic factors such as perceived knowledge gap and price of digital devices as key important indicators influencing e-banking transactions among bank customers of a sub-Saharan African region. Besides, the study further deepens the growing discussion of Internet banking adoption and retention across the globe. The results are expected to provide a theoretical contribution to the area of retail banking and to understanding consumer behavior in the Ghanaian financial services industry. This paper also offers several avenues to the academic community by inviting other scholars to consider hidden variables that might have been overlooked. Notwithstanding, this study model (see Figure 1) is first of its kind that only concentrates on socio-economic factors that impede the engagement in e-banking transactions from a developing country perspective. Hence, it is a novel approach for the interested scholars to gauge the reapplication of the model in a similar academic context to ensure optimum use of the study model.

CONCLUSION
Several studies have extensively discussed the adoption of online banking transactions, whereas this study dwelled on the socio-economic factors that impede customers' intention to make use of online banking transactions in a developing country. The study used a quantitative approach to obtain data from two leading Ghanaian banks. Data were analyzed using PLS-SEM. Findings revealed that the perceived knowledge gap and еру price of digital devices were directly relevant to the intention to disembark on e-banking transactions among sampled customers of Ghanaian banks. Whilst customer experience (frequency of the Internet use), as a moderating variable, had a significant indirect effect on the relationship between perceived knowledge gap and the intent to disembark on e-banking transactions, as well as between finance charges and the intent to disembark on e-banking transactions.
The research finally suggested that banks and Fintech players use strategic mechanisms that could profitably influence customers to use their electronic financial services while limiting the potential rise of the above-mentioned constraints. However, the paper admonished that banks and government institutions could ensure massive infrastructural development in order to resolve the examined constraints so as to achieve the objective/purpose of e-banking transactions in the financial industry. Interestingly, and more importantly, this study is first of its kind in assessing socio-economic constraints, in particular, in Ghana, where internet penetration remains low relative to global average rate.
The study is not without limitations. First, the research considered only responses from the bank customers without considering the organizational perspective. Second, the study filtered out only five factors (constructs) from the literature, that is, perceived knowledge gap, price of digital device, perceived financial charges, customer experience, and infrastructure support system were identified as socio-economic factors impeding the engagement in e-banking transactions in Ghana. Third, the sample size of the study is relatively low in terms of the number of retail banks in Ghana, thereby affecting the generalization of the work. Last but not least is a single-based geographical study.
Therefore, future scholars are encouraged to consider research that integrates perspectives from both individual and organizational levels. Additional constructs that are considered as relevant socio-economic factors are also welcomed in future studies. This study also recommends future researchers to expand the sample size and the scope of the study, such as a comparative study between two sub-Sahara African countries. This will rather increase the reliability and validity of the research model.