Trust in Government,Electronic Taxation,and the Adoption of Fintech

Abstract

Mobile money (MoMo) services have become an integral part of the banking system in developing countries. Research is needed to understand the implications of taxing this financial technology sector and inform public policy. The current MoMo tax literature is limited by its focus on personal factors while neglecting institutional factors. The present study develops an integrated model of individuals’ attitudes toward MoMo tax payments and their intentions to use MoMo services while paying related MoMo taxes. To test the model, we analyze survey data collected from 892 participants across 16 regions of Ghana using structural equation modeling. Our findings indicate that perceived trust in government and perceived tax burden are significant determinants of attitudes toward MoMo tax payment, while perceived tax burden is influenced by perceived trust in government. Similarly, perceived tax burden mediates the effect of perceived trust in government on attitudes toward MoMo tax payments. Furthermore, the perceived usefulness of MoMo services is determined by perceived ease of use of MoMo services, whereas MoMo services use/tax payment intentions are determined by perceived trust in government, attitudes toward MoMo tax payments, and perceived usefulness of MoMo services. Finally, perceived usefulness of MoMo services mediates the effects of perceived ease of use on MoMo services and MoMo service use/tax payment intentions. We discuss explanations for these findings and provide practical, theoretical, and research implications.

Keywords

trust in government electronic taxation mobile money fintech tax burden financial inclusion

Introduction

Mobile Money (MoMo) services offer convenient and secure alternatives to traditional banking and have transformed financial transactions in regions with limited access to formal financial institutions (Koomson et al., 2021; Pobee et al., 2023). In recent years, several governments have introduced legislation to tax MoMo transactions. In Ghana, the Payment Systems and Services Act, 2019 (Act 987), established a regulatory framework for the fintech industry to facilitate the digital delivery of MoMo and other financial services. This legislation was followed by the Electronic Transfer Levy Act (Act 1075) in March 2022, which imposed a 1.5% E-levy (MoMo tax hereinafter) on the total daily amount of electronic transactions exceeding GH¢100.00 (roughly USD8), whether for value-added services, product sales, or simple money transfers to family and friends. The decision to tax MoMo transactions with no additional monetary value, such as those supporting family and friends, is surprising because it conflicts with MoMo’s core utility as a solution to financial exclusion (Arestoff & Venet, 2017; Donovan, 2012; Mbiti & Weil, 2015; Morawczynski & Pickens, 2009). Immediately after the introduction of the MoMo tax in Ghana in 2022, monthly MoMo volume sank to its lowest value ever (¢71.4 billion) and has yet to surpass the highest monthly volume (¢90.5 billion) prior to MoMo tax introduction.¹ This decline in MoMo use raises the question of whether the MoMo tax provides alternative funding to the government at the expense of financial inclusion.

To address this question, this paper investigates the factors influencing attitudes toward MoMo tax payments and intentions to use MoMo services in Ghana. Identifying the factors that shape individuals’ attitudes and intentions toward MoMo usage and related tax payments is crucial to determine whether the MoMo tax impedes the adoption of MoMo services. In Ghana, the MoMo tax is automatically deducted once a MoMo transaction is authorized; hence, the behavioral attitudes and intention to use MoMo services are directly tied to the behavioral attitudes and intentions to pay the MoMo tax, as the user has no control over the decision to pay the tax. Although the greater financial inclusivity and ease of use of MoMo services compared to traditional banking systems positively influence their adoption (Akinyemi & Mushunje, 2020; Hornuf et al., 2024; Narteh et al., 2017; Osei-Assibey, 2015; Tobbin, 2010; Tobbin & Kuwornu, 2011), attitudes about compliance with the MoMo tax may negatively affect the uptake of MoMo services. Tax compliance is influenced not only by trust in the tax recipient, as validated by the stewardship and governance literature (Taing & Chang, 2021), but also by the perceived burden of existing taxes (Ali et al., 2001; Yamamura, 2014). Consequently, individuals who have lower trust in government or a stronger perception that current taxes are burdensome may seek to avoid MoMo tax payments by limiting their use of MoMo services.

Furthermore, the literature on the dynamics of trust in MoMo services has focused on trust in MoMo service providers (Baganzi & Lau, 2017; Bongomin et al., 2019; Narteh et al., 2017; Osei-Assibey, 2015; Tobbin, 2010; Tobbin & Kuwornu, 2011) and MoMo agents (Johnen et al., 2023), largely omitting empirical analyses of the impact of trust in the MoMo tax recipient (i.e., the government). Exceptions are studies by Amoah et al. (2023) and Anyidoho et al. (2023), who showed that trust in government promotes MoMo adoption. This scarcity of relevant work prevents a complete understanding of the impact of tax payments on MoMo adoption. Moreover, studies of the influence of MoMo taxes on individual usage intentions have predominantly used the MoMo tax rate as the independent variable (Anyidoho et al., 2023; Mensah et al., 2022; Mpofu, 2022; Pobee et al., 2023). This stream of MoMo tax literature confirms that the likelihood of tax compliance decreases when current MoMo tax is considered burdensome. However, the influence of an individual’s perception of their existing tax burden prior to MoMo tax introduction (which is unlikely to be affected by the MoMo tax rate) on MoMo adoption (and hence MoMo tax compliance) has not been examined.

Another unaddressed issue is whether attitudes about MoMo tax payment drive MoMo usage intentions after MoMo tax introduction. Because MoMo services improve the overall well-being of underprivileged groups (such as those with low education levels or incomes, women, older people, and the unbanked) and also enhances ease of financial transactions with less cost among the already financially included groups such as people with high education or levels of incomes, men and the banked (Amoah et al., 2020; Della Peruta, 2018; Gai et al., 2018; Koomson et al., 2021; Narteh et al., 2017; Pobee et al., 2023; Senyo et al., 2022), these groups may develop favorable attitudes and intentions towards MoMo usage. However, empirical evidence on whether these groups may develop similar favorable attitudes towards MoMo tax payments and MoMo service use/tax payment intentions post MoMo tax introduction is not available.

These gaps in the literature give rise to three key research questions; First, do trust in government and current tax burden influence attitudes toward MoMo tax payment and the intentions toward MoMo usage and payment of related taxes? Relatedly, does trust in government influence the perception of current tax burden? Second, do perceptions of MoMo services’ usefulness and ease of use influence intentions toward MoMo usage and payment of related taxes? And third, do attitudes toward MoMo tax payment lead to intention to use MoMo services and payment of related taxes?

We answer these questions by integrating the core tenets of the technology acceptance model (TAM), namely, the perceived usefulness and ease of use of MoMo services with tax burden and government trust frameworks. We capture trust in government by assessing individuals’ perceptions of the government’s competence, integrity, and benevolence, while tax burden is measured by individuals’ perceptions of how burdensome their current tax obligations are. We conduct our analyses using survey data collected from 892 respondents in Ghana and co-variance based structural equation modeling (CB–SEM).

To the best of our knowledge, this study is the first to integrate the elements of technology adoption, tax burden, and trust in government in an empirical examination of MoMo services and their promises of enhancing financial inclusion in the post-MoMo tax environment. In doing so, our study addresses a limitation of recent MoMo studies, which predominantly use pre-tax data (Tetteh et al., 2023, p. 9), and provides an alternative framework for examining MoMo tax attitudes and adoption intentions. Additionally, we contribute to the literature by shedding light on the intricate relationships between trust in government, tax burden, attitudes toward MoMo tax and MoMo adoption while highlighting certain nuances addressing barriers to financial inclusion based on our model’s addition of financial inclusion controls. Our findings inform MoMo policymakers and service providers in designing interventions and strategies as related to t MoMo tax introduction.

Theoretical Background and Literature Review

MoMo Adoption in Africa and Ghana

MoMo services are mobile phone-based services that facilitate electronic money transfers through telecommunication networks to network subscribers. To receive or transfer money, a mobile phone user must register with an agent. The user deposits money onto his or her SIM card, which then shows as “e-money” in the user’s “e-wallet” (Baganzi & Lau, 2017) and can also be withdrawn in cash by the depositor. While accessible to all mobile users, MoMo services are primarily important for unbanked individuals (Della Peruta, 2018) and accelerate money transfers among marginalized populations in developing economies (Narteh et al., 2017). MoMo services hold significant potential for financial inclusion in Africa, where most of the population is unbanked. However, MoMo technology adoption remains low in Africa as a whole (Akinyemi & Mushunje, 2020; Kiconco et al., 2019; Kikulwe et al., 2014).

In Ghana, MoMo technology adoption has been relatively high in both the central regions (where conventional banking services are fully available) and the peripheral regions of the country (where conventional banking services are mostly absent). It is estimated that approximately 60% of residents of Ghana have MoMo accounts,² providing strong evidence of the positive effects of MoMo availability on financial inclusion (Senyo et al., 2022). At the individual and business levels, MoMo services facilitate distance payments and direct borrowing of money from the user’s phone without traditional intermediaries. At the government level, state institutions are using MoMo platforms to transfer allowances and support to individual citizens (WorldBank, 2020). However, similar to other countries with high rates of MoMo adoption, Ghana’s introduction of regulations, fees, and taxes, even on non-business-oriented MoMo transactions, is impeding the rate of adoption in new contexts and discouraging existing users from expanding their use of MoMo services (Anyidoho et al., 2023; Mensah et al., 2022; Ofosu-Ampong, 2024; Takyi, 2024; Tetteh et al., 2023).

Theoretical Perspectives on MoMo Adoption

Understanding people’s decisions to use MoMo services is critical for the success of MoMo platforms and their services (Johnen et al., 2023; Meli et al., 2022; Ofosu-Ampong, 2024; Takyi, 2024). MoMo adoption has been examined through the lens of information systems theories, such as technology adoption theories and models. These studies have primarily drawn on TAM, the unified theory of acceptance and use of technology (UTAUT), innovation diffusion theories, and trust frameworks (Mensah et al., 2022; Ofosu-Ampong, 2024).

Following the logic of the TAM, elements of various MoMo platforms, such as usefulness and ease of use, have been used to explain the adoption of MoMo services (Akinyemi & Mushunje, 2020; Narteh et al., 2017; Osei-Assibey, 2015; Tobbin, 2010; Tobbin & Kuwornu, 2011). A meta-analysis by Hornuf et al. (2024) indicated that TAM-related determinants are the most important drivers of mobile fintech adoption in Sub-Saharan Africa. They found that the higher perceptions of usefulness and ease of use of MoMo services are associated with a greater likelihood of MoMo adoption.

Closely related to this stream of the literature is research drawing on the UTAUT (Baganzi & Lau, 2017; Malinga & Maiga, 2020; Mensah et al., 2022; Pobee et al., 2023), which focuses on effort expectancy, performance expectancy, facilitating conditions, and social influence. The last three factors are expected to positively influence the likelihood of MoMo adoption, whereas effort expectancy is expected to have the opposite effect. Hornuf et al.’s (2024) meta-analysis indicated that UTAUT-related determinants greatly impact MoMo adoption.

A third stream of research theoretically integrates constructs from both the TAM and the UTAUT but is mainly anchored in innovation diffusion theories and trust constructs. Studies focused on innovation diffusion expect the likelihood of MoMo adoption to be high if the MoMo technology has a relative cost advantage and is compatible, observable, uncomplex and available for trial (Della Peruta, 2018; Narteh et al., 2017; Ofosu-Ampong, 2024; Osei-Assibey, 2015; Tobbin, 2010; Tobbin & Kuwornu, 2011). Research based on trust constructs anticipates that the likelihood of MoMo adoption will be high if the MoMo technology is trustworthy and perceived to be riskless (Baganzi & Lau, 2017; Narteh et al., 2017; Okello Candiya Bongomin & Ntayi, 2020; Tobbin & Kuwornu, 2011).

Although the above theoretical perspectives provide interesting and valid explanations for MoMo adoption, several aspects of current trends in the MoMo industry have received little or no attention in the literature. First, with the exception of recent work by Ofosu-Ampong (2024) and Takyi (2024), all studies of MoMo adoption were conducted prior to the introduction of important government policy initiatives that dramatically altered the MoMo industry, most notably the MoMo tax. Taxation influences several aspects of individual behavior. Because MoMo taxes automatically apply even to transactions intended to support family and friends, it is critical to examine the antecedents of people’s attitudes and intentions to continue using MoMo services and payment of related taxes.

Second, studies of the impact of trust in MoMo services have not considered the role of trust in the government that collects taxes on MoMo transactions. Despite the work of Tetteh et al. (2023) and Anyidoho et al. (2023), trust in government remains empirically uncharted in MoMo research, which has mainly focused on trust in MoMo service providers (Baganzi & Lau, 2017; Bongomin et al., 2019; Narteh et al., 2017; Osei-Assibey, 2015; Tobbin, 2010; Tobbin & Kuwornu, 2011) and MoMo agents (Johnen et al., 2023). This is a potentially critical factor because individual trust in government is relatively low in the developing economies where MoMo services are particularly vibrant.³

Model Development and Hypotheses

To answer the research questions and further elucidate the effects of relevant factors that have already been established in the MoMo literature, we propose an integrated model explaining individual MoMo usage and its relationships with tax payment attitudes and intentions. We develop the model by identifying critical factors from the tax burden literature, the government trust literature, and the TAM and include control variables for financial inclusion—age, sex, income, education, and unbanked.

First, we argue that because the MoMo tax is automatically deducted at the source (i.e., once a MoMo transaction is authorized, the MoMo user has no control over whether to pay the MoMo tax), behavioral attitudes and intentions to use MoMo services are linked to the behavioral attitudes and intentions to pay the MoMo tax. Hence, our model predicts both MoMo usage and MoMo tax payment simultaneously.

Second, as confirmed in the trust and tax compliance literature, trust in government significantly increases individuals’ likelihood of paying taxes (Murphy, 2004; Scholz & Lubell, 1998; Taing & Chang, 2021) while reducing their perception of the tax burden (Yamamura, 2014). Third, due to the relative novelty of the MoMo tax and its financial implications for MoMo users, we anchor our work in the theory of planned behavior (TPB) (Ajzen, 1991). Specifically, we argue that individuals may not develop the intention to use MoMo services amid tax introduction unless they have already preliminarily formed a favorable attitude toward MoMo tax payment. Fourth, since MoMo services are an innovative technology, we also incorporate elements from the TAM (Davis et al., 1989) and argue that individuals’ use of MoMo services may be predicted by their perceptions of the ability of MoMo services to enhance their existing financial transactions and the efforts needed to perform such transactions.

In summary, the proposed model integrates the existing tax burden perceptions of MoMo users and their perceived trust in government to predict their attitudes toward MoMo tax payment and intentions to use/pay MoMo tax. The model also integrates TAM constructs capturing MoMo users’ perceptions of the usefulness, ease of use and attitudes toward MoMo tax influence on MoMo usage intentions.

Perceived Usefulness and Ease of Use

Perceived usefulness and ease of use are the most important factors predicting mobile fintech adoption in Sub-Saharan Africa (Hornuf et al. (2024). These two variables have been explored by drawing on the TAM framework to explain users’ behavioral intentions in technology adoption (Baah-Peprah, 2023; Davis et al., 1989; Venkatesh & Davis, 2000). Perceived usefulness is the extent to which an individual thinks that using a system will enhance their performance, whereas perceived ease of use is the amount of effort required to use a system (Venkatesh & Davis, 2000). Accordingly, we define the perceived usefulness of MoMo services as the extent to which MoMo users think that MoMo services enhance their financial transactions. Similarly, we define perceived ease of use of MoMo services as the degree to which MoMo users believe that all processes involved in conducting a MoMo transaction require little effort.

In line with Davis et al. (1989) and Venkatesh and Davis (2000), various studies have confirmed that higher levels of perceived usefulness and ease of use increase an individual’s intention to use MoMo services (Akinyemi & Mushunje, 2020; Narteh et al., 2017; Ofosu-Ampong, 2024; Tobbin, 2010; Tobbin & Kuwornu, 2011). Moreover, TAM posits that the effect of perceived ease of use on intentions is mediated by the perceived usefulness of the system, and technology adoption and behavioral studies have confirmed this effect (Baah-Peprah, 2023; Gefen & Straub, 2000; Venkatesh & Davis, 2000). However, this mediation effect has not been explored in the MoMo context. We therefore hypothesize the following:

The perceived ease of use of MoMo services positively influences (a) the perceived usefulness of MoMo services and (b) intentions to use MoMo services.

The perceived usefulness of MoMo services (a) positively influences intentions to use MoMo services and (b) mediates the relationship between perceived ease of use of MoMo services and intentions to use MoMo services.

MoMo Tax Attitudes

The association between attitude and intention is relevant for explaining tax-related compliance intentions, and this knowledge can be used to improve tax-related systems in which a positive tax attitude influences tax behavioral intentions (Taing & Chang, 2021). Because MoMo services enhance the overall well-being of underprivileged groups, including the less educated, women, the elderly, the unbanked, and low-income individuals (Amoah et al., 2020; Della Peruta, 2018; Gai et al., 2018; Koomson et al., 2021; Narteh et al., 2017; Pobee et al., 2023; Senyo et al., 2022), it is reasonable to assume that individuals in these groups developed positive attitudes toward MoMo services prior to the introduction of the MoMo tax. However, there is no empirical evidence on whether these attitudes translate into intentions to conduct MoMo transactions that incur MoMo taxes.

For less-privileged individuals, such as unbanked individuals, MoMo services may provide the only means of conducting financial transactions. However, even privileged groups, such as banked individuals, may be negatively impacted by effects of the MoMo tax, such as higher MoMo transaction costs and tax burdens. It is unlikely that an individual will engage in MoMo services without at least some preliminary consideration of attitude judgements. Regardless of an individual’s identification as privileged or less privileged, a favorable attitude toward MoMo tax payments may lead to intentions to use MoMo services and pay the related tax.

Thus far, we have theorized that intentions to use MoMo services and pay its related taxes are planned behaviors that are influenced by favorable attitudes toward MoMo tax payments, as posited by TPB (Ajzen, 1991). In addition, behavioral research on the adoption of mobile services reveals that attitude significantly influences mobile service adoption behaviors (Gopi & Ramayah, 2007), and this effect has been validated for multiple alternative finance platforms, including crowdfunding (Baah-Peprah, 2023), mobile banking services (Amoah et al., 2020), and MoMo services (Ofosu-Ampong, 2024). Although taxation was not considered, Ofosu-Ampong (2024) confirmed that attitudes toward MoMo services positively influence MoMo adoption.

In line with TPB, we define attitudes toward MoMo tax payments as a MoMo user’s overall evaluation of how favorable MoMo tax payments are. We argue that the extent to which an individual uses MoMo services is contingent upon how favorably they view the MoMo tax and have positive expectations about it, giving rise to the following hypothesis:

A favorable attitude toward the MoMo tax positively influences intentions to use MoMo services while paying the MoMo tax.

Government Trust

Trust is an important determinant of MoMo adoption. Prior to the introduction of the MoMo tax, studies of MoMo adoption focused on users’ trust in the MoMo service provider and in MoMo agents (Baganzi & Lau, 2017; Narteh et al., 2017; Okello Candiya Bongomin & Ntayi, 2020; Tobbin & Kuwornu, 2011). With the introduction of the MoMo tax, another MoMo stakeholder—the government—enters the discussion of trust. Among the few empirical studies of trust in government and its implications for MoMo adoption, Tetteh et al. (2023), Amoah et al. (2023), and Anyidoho et al. (2023) have shown that individuals’ trust in government positively influences their usage of MoMo services.

Analyses of the role of MoMo users’ trust in government are anchored in the assumptions of stewardship theory, with government functioning as the steward of collected tax revenues. As trust is a significant factor influencing their attitudes and behavioral intentions toward online financial transactions in general (Baah-Peprah & Shneor, 2022), individuals’ trust in government acts as an antecedent of their financial commitment to the government’s policy (Scholz & Lubell, 1998). Here, trust significantly influences the likelihood of paying taxes (Murphy, 2004; Scholz & Lubell, 1998; Taing & Chang, 2021). This likelihood depends on positive attitudes and intentions toward tax compliance (Taing & Chang, 2021) and the assumption that the government will exercise competence, integrity, and benevolence in its accountability for the tax revenues collected. Accordingly, we hypothesize the following:

Trust in government positively influences (a) attitudes toward MoMo tax payments and (b) intentions to use MoMo services while paying the MoMo tax.

Tax Burden

Taxing electronic commerce is challenging. In their quest to improve their revenue base by imposing new taxes, governments in developing countries risk restricting the development of new e-markets and the involvement of the community in the evolution and growth of these e-markets (Bongomin et al., 2019; Jones & Basu, 2002). These unintended consequences of taxation may shrink the tax base, leading to an uneven effect and, in the long run, jeopardizing the already fragile economies of these countries (Jones & Basu, 2002). Such consequences are usually magnified when the community participating in the e-market perceives their existing tax obligations as burdensome (Yamamura, 2014).

The long-term detrimental consequences of general taxation of e-markets should serve as a warning for implementing MoMo taxes. The governments of developing economies view taxation of MoMo transactions as a game changer for revenue inflows (Mpofu, 2022). However, reducing or eliminating taxes enhances financial inclusion and provides individuals with basic financial services. Here, Muthiora and Raithatha (2017) and Anyidoho et al. (2023) have argued that exempting MoMo services from taxation is necessary to promote financial inclusion and to facilitate payments for goods and services, which may indicate prices that already include taxes such as VAT.

The literature confirms that MoMo taxes negatively moderate the effect of intention on actual usage of MoMo innovation (Pobee et al., 2023), and MoMo tax exemption positively moderates the relationship between MoMo adoption and financial inclusion (Bongomin et al., 2019). Conversely, MoMo taxation significantly influences MoMo adoption (Anyidoho et al., 2023; Ofosu-Ampong, 2024; Takyi, 2024), and removing MoMo taxes positively influences MoMo users’ actual behavior toward MoMo services (Mensah et al., 2022; Ndung’u, 2019). Rojas-Suarez and Pacheco (2017) argued that MoMo users may try to avoid paying the MoMo tax by making fewer MoMo transactions. However, Mpofu’s (2022) comprehensive critical review found mixed results on the impact of MoMo taxes on MoMo services adoption.

Bongomin et al. (2019) and (Rojas-Suarez & Pacheco (2017) found that taxes levied on MoMo transactions that are above standard rates significantly reduce the usage of MoMo services. However, a tax rate that is considered high in one country might not be considered high in another. MoMo tax rates vary greatly even in regions with similar economies, including in Sub-Saharan Africa. For example, the MoMo tax rate is 0.02% in Cameroon, 0.05% in Uganda, 2% in Zimbabwe, and 1.5% in Ghana. Though heterogeneity in MoMo tax rates between countries may explain Mpofu’s (2022) mixed findings on the impact of MoMo taxes on MoMo adoption, examining MoMo users’ perceptions of their existing tax burden prior to MoMo tax introduction might eliminate the confounding effect of tax rates and enable a critical assessment of the effect of MoMo taxation that can be generalized across contexts.

The general tax literature confirms that the likelihood of tax compliance is negatively related to the perceived burden of taxation (Ali et al., 2001; Yamamura, 2014). Thus, we argue that MoMo users who believe their tax burden is relatively high will have less favorable attitudes toward MoMo tax payments and may avoid paying the MoMo tax by limiting their use of MoMo services. This argument is underpinned by two theoretical foundations. First, TPB (Ajzen, 1991) suggests that an individual is affected by their belief in the ease of completing a task, that is, perceived behavioral control. Therefore, we assume that when MoMo users perceive paying taxes as easy and not burdensome, they will exhibit greater tax compliance. Second, self-efficacy theory, which is part of TPB, and social cognitive theory (Heffernan, 1988) predict that an individual’s belief and self-confidence in their ability to complete a task enhance their intentions to do so. Accordingly, MoMo users who perceive that they can pay MoMo taxes are likely to use MoMo services. In other words, MoMo users who consider their current tax obligations to be burdensome will develop strategies to reduce their taxes, while the opposite will be true for MoMo users who do not consider their current tax obligations to be burdensome. Thus, perceived tax burdens will affect individuals’ attitudes toward MoMo tax payments and intentions to use MoMo services/pay the MoMo tax. Accordingly, we hypothesize the following:

Perceived current tax burden negatively influences (a) attitudes toward MoMo tax payments and (b) intentions to use MoMo services while paying the MoMo tax.

Furthermore, the degree to which an individual considers a tax to be burdensome is contingent on several perceptions, including trust. Notably, the tax literature confirms that people perceive their tax obligation as burdensome if they perceive that the taxes are not appropriate for the benefit of society (Yamamura, 2014). That is, taxpayers may consider the costs of paying taxes higher than the benefits they receive. On the contrary, taxpayers perceive taxes to be less burdensome when they trust that the government will efficiently use the taxes to maximize societal welfare (Yamamura, 2014). Accordingly, we hypothesize the following:

Perceived trust in government (a) negatively influences perceived current tax burden, (b) mediates the relationship between perceived trust in government and attitudes toward MoMo tax payments, and (c) mediates the relationship between perceived trust in government and intentions to use MoMo services while paying the MoMo tax.

Figure 1 summarizes the hypotheses in this study and control variables for financial inclusion: education level, gender, age, access to banking, and income (Amoah et al., 2020; Della Peruta, 2018; Gai et al., 2018; Koomson et al., 2021; Narteh et al., 2017; Pobee et al., 2023; Senyo et al., 2022).

Figure 1.

Summary of hypotheses.

Methods

Study Context

This study uses survey data collected from Ghana, a country coping with financial inclusion challenges but a fintech-ready context, nonetheless. Ghana is a unique institutional setting that differs from the often studied ‘Western, educated, industrialized, rich and democratic—WEIRED context’ (Henrich et al., 2010). Prior to the introduction of MoMo services in Ghana, remittances were mostly performed through bus drivers. Senders would visit a bus station and negotiate with a bus driver to take the money and deliver it to the recipient (for a fee). The amount would then be delivered to the recipient, who might wait for several hours due to the difficulty of predicting the arrival time of the bus driver. The introduction of MoMo services significantly eased these remittance hustles in Ghana. Accordingly, Statista (2023)⁴ identified Ghana as a notable example of the increase in MoMo adoption in Sub-Saharan Africa. Furthermore, the Global Findex Report (2021)⁵ indicated that MoMo has significantly enhanced financial inclusion in Ghana, as evident in the percentage of financially included individuals which increased from 41% in 2014 to 58% in 2017. Ghana has also demonstrated remarkable growth in MoMo transaction volumes, from approximately USD69 million in 2022 to USD123 million in 2023.⁶ In terms of taxation, Ghana is among the pioneers in MoMo taxation, which was introduced by the government in the form of the ‘Electronic Transfer Levy Act’ (Act 1075). Hence our choice of Ghana as a study context allows us to observe MoMo use intentions and behavior, while exploring the implications of MoMo taxation.

Measurement Items and Scale Generation

Our research model relies on multi-item measurements. Accordingly, previously validated measurement items from the literature were adapted with minor wording changes to better fit the MoMo context as detailed in Table 3. Measurements were conducted on seven-point Likert scales, where 1 indicated strong disagreement with the statement and 7 indicated strong agreement with the statement. The items were revised in rounds of review, and based on feedback from a pilot study of potential respondents.

Data Collection and Sampling

We data collected via an online survey of Ghanaian adults in 2023, while using the SurveyXact software. Participants responded to a questionnaire based on their experience of using MoMo services. Local collaborators distributed the survey via online platforms to the general population in all 16 regions of Ghana, as well as engaged in follow-ups and reminders. Because the survey included a long list of questions in English, respondents were offered an internet data token with a value of 1 euro. The survey was conducted following best practice recommendations as outlined by several researchers (Cook et al., 2000; Porter & Whitcomb, 2003; Roberts & Allen, 2015).

The survey was distributed to 2,848 individuals across the country; 1,047 responded, corresponding to a 37% response rate. After removing entries with missing data, those suspected of monotonous responses (straight lining of same answers), and responses that were provided within less than the minimum time recorded during the pilot survey (10 minutes), we were left with 892 responses, corresponding to an 85% retention rate. Accordingly, this sample size meets the requirements for multivariate data analysis according to Hair et al. (2010) and, specifically, SEM (Barrett, 2007; Mulaik, 2007; VanVoorhis & Morgan, 2007). The demographics characteristics of the respondents are reported in Table 1.

Table 1.

Demographic Characteristics of the Respondents.⁷

Demographics	Characteristic	Frequency	Percentage (%)
Sex	Male	505	56.66
Sex	Female	387	43.34
Age	18–24	245	27.40
	25–34	465	52.01
	35–44	112	15.90
	45 and above	49	5.49
Education	Up to high school	173	6.30
	Bachelor	531	19.37
	Master’s or above	189	59.43
Employment	Unemployed	359	40.20
Employment	Employed	534	59.80
Mobile money usage (reason)	Business	118	13.21
Mobile money usage (reason)	Family and friends	775	86.79
Mobile money usage (model)	Send money/pay bills	229	25.64
Mobile money usage (model)	Receive money/bills	664	74.36
Mobile money service provider	MTN	325	36.39
	Vodafone	459	51.43
	AirtelTigo	89	9.97
	Other	19	2.21
Financial inclusion–banked (percentage of region’s population with a bank account)	5 (minimum)–19	101	11.32
	20–30	387	43.39
	30–40 (maximum)	404	45.29

Non-response Bias Check

Non-response bias occurs when respondents have significantly different characteristics than survey nonrespondents. Percent to address this concern, we tested for the possibility of non-response bias in our data via a wave analysis, as recommended by Armstrong and Overton (1977) and employed by survey studies in various fields. Accordingly, we split our sample into two groups representing the first and last halves of the respondents. These two groups were considered respondents and non-respondents, respectively, because of the delay in participation between them (Karahanna & Straub, 1999). The t test of the two samples indicated the absence of significant differences in gender, age, education, employment, income, and MoMo transactions paid and received (both quantity and amount). Hence, there was no evidence of severe non-response bias in our data (Table 2).

Table 2.

Non-response Bias Check.

Demographics	Mean of First Responders	Mean of Last Responders	T	df	p-Value
Gender	0.455	0.413	1.2834	889.88	0.200
Age	29.545	29.374	0.3611	888.34	0.718
Education	3.036	3.022	0.2942	889.64	0.769
Employment	0.603	0.594	0.2729	889.99	0.785
Income	1515.952	1442.765	0.5946	872.26	0.552
No. of MoMo transactions (paid)	282.780	218.038	1.2356	842.99	0.217
No. of MoMo transactions (received)	307.595	266.054	0.7570	879.77	0.449
Amount of MoMo transactions (paid)	1365.241	1385.615	−0.1124	889.54	0.911
Amount of MoMo transactions (received)	1403.192	1273.799	0.8046	889.64	0.421

Normality Check

A normal data distribution test is required for SEM estimations. We therefore evaluated the normality of our data. We used the Mardia (1970) test to examine multivariate normality and found that the data were non-normally distributed. For robustness, we further assessed the univariate normality of each measurement item using the Shapiro and Wilk (1965) test and found the absence of univariate normality, with all p-values below the .05 significance threshold. Because none of the variables exhibited characteristics of a normal distribution, following the recommendations of Rosseel (2012), we employed the Satorra-Bentler rescaling method, also known as robust maximum likelihood, for all SEM estimations. For CB–SEM, the lavaan package in the R software suite was used.

Data Analysis and Empirical Results

We created a structural model using CB–SEM to analyze the measurement and structural models. CB-SEM was chosen over partial least squares SEM (PLS–SEM) based on the purpose of the study, model complexity, and characteristics of the data. CB–SEM works well with complex models and large sample sizes (Rosseel, 2012). In addition, CB–SEM is more effective than PLS–SEM when the purpose of the research is theoretical integration of several constructs comprising multiple measurements (Deng et al., 2018; Henseler et al., 2015; Rosseel, 2012).

Measurement Model Assessment

We first confirmed the reliability and validity of the measurement model by performing exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). We also assessed the potential for multicollinearity among the constructs because SEM considers random measurement error, which increases the risk of collinearity among latent constructs and less stable parameter estimates (Grapentine, 2000).

Exploratory and Confirmatory Factor Analyses

We first conducted an EFA of our measurement model, which resulted in the removal of some items that either exhibited cross-loadings or low levels of loading below 0.4 (Hair et al., 2010) (Table 3). We followed this with a CFA of all items that were deemed valid in the EFA. The fit indices of the CFA (Table 4) confirmed that for all items, the ratio of chi-square to degrees of freedom was below the upper threshold of 3, the comparative fit index (CFI) and Tucker–Lewis index (TLI) both exceeded the minimum threshold of 0.90, and the root mean square error of approximation (RMSEA) values and standardized mean square root (SRMR) values were below the 0.08 maximum threshold. These goodness-of-fit indices meet the required thresholds recommended by best practice (Barrett, 2007; Hair et al., 2010; Rosseel, 2012) and suggest a good fit of our measurement model.

Table 3.

Exploratory Factor Analysis and Measurement Items.

Latent constructs	Measurement Items		Factor Loadings	Source
Perceived government competence (PGC)	PGC1	Overall, I find the government to be capable	0.835 ***	PGC1–4 were adapted from the “dimensions of trust in government and online transactions” literature (Bouckaert & Van de Walle, 2003; Gefen, 2002; Van de Walle, 2017)
	PGC2	Overall, I find the government to be effective	0.919 ***
	PGC3	Overall, I find the government to be skillful	0.915 ***
	PGC4	Overall, I find the government to be professional	0.919 ***
Perceived government integrity (PGI)	PGI1	Overall, I find the government to be sincere	0.965 ***	PGI1–3 were adapted from the “dimensions of trust in government and online transactions” literature (Bouckaert & Van de Walle, 2003; Gefen, 2002; Van de Walle, 2017)
	PGI2	Overall, I find the government to be truthful	0.940 ***
	PGI3	Overall, I find the government to be promise fulfilling	Removed
Perceived government benevolence (PGB)	PGB1	Overall, I find the government to care about the welfare of the public	0.943 ***	PGB1–3 were adapted from “dimensions of trust in government and online transactions” literature (Bouckaert & Van de Walle, 2003; Gefen, 2002; Scholz & Lubell, 1998; Song & Lee, 2016; Van de Walle, 2017)
	PGB2	Overall, I find the government to be motivated to act in the public interest	0.932 ***
	PGB3	Generally, how much of the time do you think you can trust the government?	0.869 ***
Perceived current Tax burden (PTB)	PTB1	To what degree do you think your current tax is a burden?	0.729 ***	PTB1–5 were generated specifically for this study based on the “perceived tax burden”-related literature
	PTB2	I think the amount of tax I pay currently is too high	0.891 ***
	PTB3	I think I am currently paying too much tax	0.945 ***
	PTB4	I think there are too many taxes to be paid by me	0.787 ***
	PTB5	I pay too much tax	0.866 ***
Attitudes toward MoMo Tax payments (ATT)	ATT1	I think I would like to pay the e-Levy	Removed	ATT1–6 were adapted from the “individuals’ attitudes about online platforms” literature (Baah-Peprah, 2023; C.-L. Hsu & Lin, 2008; M.-H. Hsu et al., 2006)
	ATT2	I am likely to feel good about paying the e-Levy	0.888 ***
	ATT3	I think that paying the e-Levy is good	0.946 ***
	ATT4	I think that paying the e-Levy is appropriate	0.937 ***
	ATT5	I think that paying the e-Levy is beneficial	0.901 ***
	ATT6	I have a positive opinion about the e-Levy	Removed
	ATT7	I have positive feelings about the e-Levy	Removed
MoMo usage intentions (MUI)	MUI1	When needed, I intend to use mobile money for my payment transactions	Removed	MUI1–5 were adapted from the “individual financial transaction intentions” literature (Baah-Peprah, 2023; Shneor & Munim, 2019)
	MUI2	When needed, I predict that I will use mobile payments for my payment transactions in the future	0.795 ***
	MUI3	It is likely that I will use mobile payments for my payment transactions in the future	0.865 ***
	MUI4	I have the intention to use mobile payments for my payment transactions	0.886 ***
	MUI5	I intend to actively use mobile payments for my payment transactions	0.817 ***
Perceived usefulness of MoMo (PUF)	PUF1	Overall, I find the mobile money payment system(s) (I usually use) to be useful	Removed	PUF1–7 were adapted from the “individuals’ perceived usefulness of online-payment-mediated platforms” literature (Baah-Peprah, 2023; Shneor & Munim, 2019)
	PUF2	I find the mobile money payment system(s) I usually use to be valuable to me	0.727 ***
	PUF3	The content of the mobile money payment system(s) I usually use is useful to me	0.755 ***
	PUF4	The mobile money payment system(s) I use is functional	0.700 ***
	PUF5	The mobile money payment system(s) (I usually use) helps me to make or pay for necessary and useful transactions	0.701 ***
	PUF6	The mobile money payment system(s) (I usually use) helps me to support people and causes more efficiently	Removed
	PUF7	In general, the mobile money payment system(s) (I usually use) is useful for helping people that need support	Removed
Perceived ease of-use of MoMo (PEU)	PEU1	The mobile money payment system(s) I usually use is simple to use, even when using it for the first time	0.705 ***	PEU1–7 were adapted from the “individuals’ perceived ease-of-use online- payment-mediated platforms” literature (Baah-Peprah, 2023; Shneor & Munim, 2019)
	PEU2	In the mobile money payment system(s) I usually use, everything is easy to find	0.714 ***
	PEU3	The structure and content of the mobile money payment system(s) I usually use is easy to understand	Removed
	PEU4	It is easy to navigate and move within the mobile money payment system(s) I usually use	0.714 ***
	PEU5	Learning to use the mobile money payment system(s) (I usually use) is easy	0.713 ***

Table 4.

Measurement Model Reliability, Validity and Multicollinearity.

	Reliability	AVE	VIF	MPI	PEU	PUF	ATT	PGC	PGI	PGB	PTB
MUI	0.91	0.709	1.727	1	(0.181)	(0.328)	(0.040)	(0.091)	(0.055)	(0.037)	(0.001)
PEU	0.78	0.665	2.232	0.425	1	(0.438)	(0.000)	(0.021)	(0.006)	(0.000)	(0.006)
PUF	0.81	0.511	2.949	0.573	0.662	1	(0.004)	(0.039)	(0.018)	(0.006)	(0.000)
ATT	0.96	0.843	1.838	0.201	0.002	0.060	1	(0.334)	(0.295)	(0.369)	(0.108)
PGC	0.94	0.805	6.173	0.302	0.145	0.197	0.578	1	(0.817)	(0.671)	(0.034)
PGI	0.95	0.908	9.091	0.234	0.078	0.134	0.543	0.904	1	(0.794)	(0.029)
PGB	0.94	0.838	5.556	0.191	0.006	0.075	0.607	0.819	0.891	1	(0.033)
PTB	0.93	0.718	1.311	−0.023	0.077	0.012	−0.329	−0.183	−0.172	−0.182	1

Note. The values below and above the diagonal (in parentheses) are the correlations of the constructs and squared correlations of the constructs, respectively. See Table 3 for definitions of abbreviations. Reliability was measured by Cronbach’s alpha. PGC, PGI and PGB are hierarchically reflective constructs of perceived government trust hence their high VIF and construct correlation values.

Common Method Bias

The use of the same measurement scale throughout the survey may raise concerns regarding common method bias. Accordingly, we followed the procedures suggested by Podsakoff et al. (2003) to examine whether our data suffered from such bias. We used Harman’s single-factor approach and created a single factor with all measurement indicators in the EFA without rotation. The proportion of the data explained by the single factor was 24%, far less than the maximum threshold of 50%, eliminating concerns about common method bias in our data.

Reliability, Validity and Multicollinearity

Reliability is defined as the extent to which individual items provide the estimated value of the latent construct (Gefen et al., 2000; Hair et al., 2014). To confirm the reliability of the constructs, we assessed the composite reliability (see Table 4) of all latent variables, resulting in Cronbach’s alphas that were all well above 0.70 (Cronbach, 1951).

To confirm that all latent constructs were distinguishable from each other (discriminately valid), we employed the divergent validity criteria proposed by Fornell and Larcker (1981) and performed a robust assessment of the SEM constructs. As shown in Table 4, the divergent validity of the constructs in our model was confirmed, as the squared correlation value of each construct was less than its average variance extracted (AVE) value. In addition, all AVE values were above the threshold of 0.50 (Bagozzi & Yi, 1988; Fornell & Larcker, 1981).

The problem of multicollinearity is closely related to discriminant validity concerns. If constructs are too highly correlated and are not distinguishable from each other, then they lack discriminant validity (Grewal et al., 2004; Mason & Perreault Jr., 1991). Constructs that pass both reliability and validity tests are unlikely to be highly correlated (Anderson & Narus, 1984; Grewal et al., 2004). Although our model passed both reliability and validity tests, hence indicating no serious multicollinearity issues, for robustness we further tested the multicollinearity of our model. The variance inflation factor (VIF) values were less than 3 for all constructs except those estimated to reflect and load as second-order constructs (see Table 4). These results indicate the absence of multicollinearity issues in our model (Hair et al., 2019). Additionally, we computed the heterotrait–monotrait (HTMT) ratios of the model’s latent variables (Table 5). All ratios were less than 1, further confirming that our model’s constructs represent different theoretical concepts and are statistically different (Henseler et al., 2015).

Table 5.

Heterotrait–Monotrait (HTMT) Ratios.

	ATT	MUI	PEU	PGB	PGC	PGI	PTB	PUF
ATT	-
MUI	0.202
PEU	0.031	0.433
PGB	0.632	0.197	0.027
PGC	0.587	0.317	0.152	0.822
PGI	0.556	0.239	0.074	0.903	0.896
PTB	0.328	0.043	0.083	0.186	0.180	0.168
PUF	0.057	0.580	0.692	0.068	0.197	0.132	0.031	-

Note. See Table 3 for definitions of abbreviations.

Second-order Construct Assessment

As shown in Table 4, there are high correlations between perceived government competence, integrity, and benevolence, in line with trust theory (Gefen, 2002). Perceived government competence, integrity, and benevolence also have high VIF values >3 (0.6173, 9.091 and 5.556) and high HTMT ratios (0.903, 0.822 and 0.896), indicating excessive multicollinearity (García et al., 2015; Henseler et al., 2015; O’brien, 2007). These results confirm that perceived government competence, integrity and benevolence are three underlying dimensions of trust (Gefen, 2002; Hardin, 1991; Mayer et al., 1995) that need to be combined into a second-order construct (Bradley & Henseler, 2007; Chin et al., 2003), following best practices (García et al., 2015; O’brien, 2007).

To assess trust in government as a second-order construct in our model, we followed the recommendations of Bradley and Henseler (2007) and employed the repeated indicator approach, also called hierarchical model assessment (Chin et al., 2003). The results are presented in Table 6. The significant weights (standard estimates, z and p values) of the first-order constructs—government competence, integrity, and benevolence—on the designated second-order construct, trust in government, and the AVE and CR values confirm that trust in government is a multidimensional construct with three ontological attributes: government competence, integrity, and benevolence. Accordingly, our second-order construct of trust in government is confirmed in line with trust theory and backed statistically by our multicollinearity checks and hierarchical model assessment.

Table 6.

Weights of the First-order Constructs on the Designated Second-order Construct.

Second-order Construct	First-order Constructs	Standard Estimate	z-Value	p-value
Trust in government (AVE = 0.570)(CR = 0.970)	Competence	0.691***	7.386	.000
	Integrity	0.660***	7.397	.000
	Benevolence	0.493***	7.387	.000

Structural Model Assessment and Hypothesis Testing

We assessed our model’s inner structure by evaluating the goodness-of-fit indices from the SEM analysis (Hair et al., 2019). As shown in Table 7, the values of CFI, TLI, RMSEA, SRMR, and chi-square (ratio of t-statistics to degree of freedom) all meet the required criteria for goodness of fit. The predictive powers of the model’s outer structure were assessed by the value of R-squared (R²), which indicates the proportion of the dependent variables explained by the model. The results of the hypothesis testing are schematically illustrated in Figure 2 and summarized in Table 7. We estimated two different SEM models in which the hypothesized effects were tested without controls (Model 1) and with controls (Model 2) and obtained consistent findings across the models.

Table 7.

Summary of Hypotheses Testing and Estimation Results.

Hypothesized Effect	Path	Model 1 (Without Controls)	Model (2) (With Controls)	Decision
H1a	PEU → PUF	0.702 (0.088) ***	0.711 (0.089) ***	Supported
H1b	PEU → MUI	0.056 (0.092)	0.049 (0.097)	Not supported
H2a	PUF → MUI	0.510 (0.106) ***	0.522 (0.110) ***	Supported
H2b	PEU → PUF → MUI	0.358 (0.083) ***	0.71 (0.088) ***	Supported
H3	ATT → MUI	0.105 (0.027) **	0.106 (0.027) **	Supported
H4a	PTG → ATT	0.561(0.040) ***	0.561 (0.040) ***	Supported
H4b	PTG → MUI	0.131 (0.029) **	0.126 (0.030) **	Supported
H5a	PTB → ATT	−0.223 (0.041) ***	−0.222 (0.040)***	Supported
H5b	PTB → MUI	0.023 (0.031)	0.024 (0.031)	Not supported
H6a	PTG → PTB	−0.188 (0.034) ***	−0.182 (0.034) ***	Supported
H6b	PTG → PTB → ATT	0.042 (0.011) ***	0.041 (0.011) ***	Supported
H6c	PTG → PTB → MUI	−0.004 (0.005)	0.004 (0.005)	Not supported
	PTG → ATT → MUI	0.059 (0.016) **	0.060 (0.016) **	Supported
Unhypothesized mediations	PTB → ATT → MUI	−0.024 (0.008) *	−0.024 (0.008) *	Supported
	Sex → MUI		0.013 (0.075)	Non-significant
	Sex → ATT		−0.088 (0.092) **	Significant
	Sex → PUF		−0.072 (0.058) *	Significant
	Sex → PTB		−0.004 (0.101)	Non-significant
	Age → MUI		−0.010 (0.007)	Non-significant
	Age → ATT		0.031 (0.008)	Non-significant
	Age → PUF		0.058 (0.006)	Non-significant
	Age → PTB		−0.012 (0.009)	Non-significant
	Income _log → MUI		0.024 (0.029)	Non-significant
Financial inclusion controls	Income _log → ATT		−0.028 (0.034)	Non-significant
	Income _log → PUF		−0.015 (0.023)	Non-significant
	Income _log → PTB		0.041 (0.038)	Non-significant
	Education → MUI		−0.028 (0.064)	Non-significant
	Education → ATT		−0.016 (0.077)	Non-significant
	Education → PUF		−0.011 (0.064)	Non-significant
	Education → PTB		0.053 (0.082)	Non-significant
	Banked→ MUI		0.013 (0.057)	Non-significant
	Banked → ATT		0.083 (0.065) **	Significant
	Banked → PUF		−0.094 (0.092) **	Significant
	Banked → PTB		−0.021 (0.070)	Non-significant
	CFI >0.90	0.921	0.919
	TLI >0.90	0.913	0.909
Goodness-of-fit indices	RMSEA <0.08	0.061	0.056
	SRMR <0.08	0.043	0.042
	χ2 (t-statistic/df) < 3	(1696.115/695) = 2.44	(1969.139/725) = 2.72

	R² ATT	41.2%	42.1%
Explanatory power	R² MUI	36.3%	46.7%
	R² PUF	49.2%	52.9%
	R² PTB	3.5%	3.9%
Number of observations		892	892

Note. Standard errors in parentheses; *p < .05, **p < .01, ***p < .001. Abbreviations: PTG, Perceived trust in government; PTB, Perceived tax burden; ATT, Attitudes towards MoMo tax payment; MUI, MoMo usage intentions; PUF, Perceived usefulness of MoMo; PEU, Perceived ease-of-use of MoMo.

Figure 2.

Summary of overall results of structural model assessment.

First, our model explains 42.1% of attitudes toward MoMo tax payments; perceived trust in government has a positive and significant effect on attitudes toward MoMo tax payments, whereas perceived tax burden has a negative and a significant effect (supporting Hypotheses H4a and H5a, respectively). In addition, perceived tax burden significantly mediates the relationship between perceived trust in government and attitudes toward MoMo tax payments (supporting Hypothesis 6b).

Second, our model explains 46.7% of MoMo service use/tax payment intentions; perceived usefulness of MoMo services, attitude toward MoMo tax payments, and perceived trust in government all have positive and significant effects on this variable (supporting Hypotheses H2a, H3, and H4b, respectively). By contrast, perceived ease of use of MoMo services and perceived tax burden have no significant effects on MoMo service use/tax payment intentions (failing to support Hypotheses H1b and H5b, respectively). Nonetheless, perceived usefulness of MoMo services fully mediates the effect of perceived ease of use of MoMo services on MoMo services use/tax payment intentions (supporting Hypothesis H2b). We also find that perceived tax burden does not mediate the relationship between perceived trust in government and MoMo services use/tax payment intentions (failing to support Hypothesis H6c).

Third, our model explains 52.9% of perceived usefulness of MoMo services, which is positively and significantly affected by perceived ease of use of MoMo services (supporting Hypothesis H1a). Fourth, our model explains 3.9% of perceived tax burden, with a significant and negative effect of perceived trust in government (supporting Hypothesis 6a).

Finally, when examining the effect of financial inclusion controls, we find that the perceived usefulness of MoMo services and consequently attitudes toward MoMo tax payments are more positive among women than among men. In addition, respondents in regions with higher proportions of the banked population have lower perceptions of the usefulness of MoMo services but more favorable attitudes toward MoMo tax payments. The effects of the other controls are not significant.

Discussion

Overall, our findings capture the antecedents of attitudes toward MoMo tax payments and MoMo services use/tax payment intentions. We draw on TAM and extend it by examining the roles of perceived usefulness, perceived ease of use, attitude toward MoMo tax payments, trust in government and tax burden as instrumental factors predicting attitudes and intentions to use MoMo services amid the introduction of the MoMo tax.

First, among the two predictors of behavioral intentions from TAM (Davis et al., 1989), only perceived usefulness of MoMo services has a significant positive influence on MoMo service use/tax payment intentions. Perceived ease of use of MoMo services only influences such intentions indirectly through perceived usefulness of MoMo services. However, we find that the perceived usefulness of MoMo services is predicted by the perceived ease of use of these services. These findings align with those of prior studies (Akinyemi & Mushunje, 2020; Ofosu-Ampong, 2024) but also contradict prior findings on the relationship between ease of use and intentions reported by Tobbin and Kuwornu (2011) and Narteh et al. (2017). These studies found direct positive effects of ease of use on both usefulness and MoMo adoption intentions in Ghana (Tobbin & Kuwornu, 2011) and other selected African countries (Narteh et al., 2017). A meta-analysis of TAM (Yousafzai et al., 2007a, 2007b) and other studies outside the context of MoMo research (Baah-Peprah, 2023; Gefen & Straub, 2000; Keil et al., 1995) also confirm our findings.

The discrepancies between our findings and those of Tobbin and Kuwornu (2011) may be explained by changes in industry mechanisms since 2011, when there were limited (if any) MoMo usage experiences and the MoMo tax had not yet been introduced. Moreover, more than a decade of experience with MoMo technology may have dampened the importance of ease of use. Additionally, our sample is three times larger and more diverse than the sample in Tobbin and Kuwornu (2011)’s study, which targeted respondents at shopping malls (which are mostly visited by affluent people in Ghana). Our findings also contradict the recent work of Ofosu-Ampong (2024) possibly because our study sample is substantially larger than his and has wider national coverage across 16 regions and all walks of life. By contrast, Ofosu-Ampong sampled professionals (accounting and IT professionals), a group of people who may have little or no difficulty conducting MoMo transactions. Although Narteh et al.’s (2017) study also had a large sample, it was conducted before MoMo tax introduction. Here, unlike these prior studies (Narteh et al., 2017; Ofosu-Ampong, 2024; Tobbin & Kuwornu, 2011), use of MoMo in our study includes more than one task, where individuals need to consider intentions for both MoMo use and MoMo tax payment at the same time. This level of complexity may account for the opposing findings.

Second, the MoMo literature has underscored the relevance of MoMo services for enhancing the financial inclusion of underprivileged individuals, such as the less educated, women, older people, the unbanked and those with low incomes (Amoah et al., 2020; Della Peruta, 2018; Gai et al., 2018; Koomson et al., 2021; Narteh et al., 2017; Pobee et al., 2023; Senyo et al., 2022). The proportions of these groups that use MoMo services are significantly higher than those of their privileged counterparts, for example, the banked population.⁸ Consequently, MoMo tax obligations may obscure the intentions of underprivileged groups to use MoMo services even if they have favorable MoMo tax payment attitudes. Nevertheless, this obscuration may not be strong enough to influence the effect of attitude on intentions (Ajzen, 1991). Here, we confirm that favorable attitudes toward MoMo tax payments is positively associated with MoMo services use/tax payment intentions. This finding is in line with the established predictions of TPB in explaining attitude and intention relationships in, for example, mobile service adoption behaviors (Gopi & Ramayah, 2007). It is also consistent with the literature on technological innovation adoption in other fields, including alternative finance technologies (Baah-Peprah et al., 2024; Shneor & Munim, 2019; Shneor et al., 2021), e-commerce technologies (Herrero Crespo & Rodríguez del Bosque, 2008; Pavlou, 2002), e-health treatment technologies (Rich et al., 2015), and e-tax systems (Taing & Chang, 2021).

Third, while trust in MoMo service providers and MoMo agents is critical in MoMo adoption intentions, the introduction of MoMo taxes has heightened the relevance of trust in another key stakeholder: the government. Although empirical explorations of this topic remain limited, the works of Tetteh et al. (2023), Amoah et al. (2023), and Anyidoho et al. (2023) confirmed a positive effect of government trust on MoMo adoption but measured trust using binary conditions of “yes or no” to a single statement such as “trust the government” (Amoah et al., 2023; Tetteh et al., 2023) or “misuse of collected taxes” (Anyidoho et al., 2023). Our study employed a multidimensional and hierarchical construct approach, which captures individuals’ trust in government with three ontological attributes of government: competence, integrity, and benevolence. Our empirical results confirm that trust in government positively impacts attitudes toward MoMo tax payments and MoMo services use/tax payment intentions. We also find that perceived tax burden significantly positively mediates the effect of trust in government on attitudes toward MoMo tax payments but not MoMo services use/tax payment intentions. These findings further strengthen earlier arguments that trust in government is a critical factor in MoMo usage (Amoah et al., 2023; Anyidoho et al., 2023; Tetteh et al., 2023) and favorable attitudes toward MoMo tax payments.

Finally, while some studies have reported a negative effect of Ghana’s MoMo tax on MoMo adoption (Anyidoho et al., 2023; Ofosu-Ampong, 2024; Takyi, 2024), a meta-analysis by Mpofu (2022) provided mixed evidence. These inconsistencies may be partially attributable to a stronger emphasis of these studies on the MoMo tax rate than on MoMo users’ perceptions of their tax burden. The current study finds that MoMo users’ perceptions of their tax burden are crucial for examining the implications of the MoMo tax, especially since individual tax responsibilities and MoMo taxes differ significantly across countries. Our model shows that individuals’ perceived current tax burden negatively affects their attitudes toward MoMo tax payments and influences the effects of these attitudes on their intentions to use MoMo services/pay the MoMo tax. These novel insights and the suggestion of items for measuring perceived tax burden are important contributions of this work and a divergence from prior work, which has solely examined the presence of the MoMo tax (Anyidoho et al., 2023; Bongomin et al., 2019; Mensah et al., 2022; Ndung’u, 2019; Pobee et al., 2023; Rojas-Suarez & Pacheco, 2017; Takyi, 2024).

Overall, our findings suggest that MoMo taxation represents a financial inclusion paradox. Specifically, even though the perceived current tax burden does not directly influence MoMo services use/tax payment intentions, it may jeopardize the intended benefits of MoMo services, such as financial inclusion, and its claim as a means of alternative funding, by negatively influencing users’ attitudes toward MoMo services. These attitudes may in turn negatively influence usage intentions, hence exerting an indirect rather than direct negative effect on intentions.

Finally, it may be worthwhile mentioning some insights emerging from effects of control variables as related to the role of MoMo in financial inclusion. Compared with men, women have higher perceptions of MoMo usefulness but lower attitudes toward MoMo tax payments. These findings indicate that MoMo services enhance the financial inclusion of females, who traditionally dominate financially excluded groups, especially in developing economies. However, taxing MoMo transactions negatively affects females’ attitudes toward using MoMo services. Similarly, although respondents from banked regions exhibit higher MoMo tax attitudes than respondents from unbanked regions, the latter have higher perceptions of the usefulness of MoMo services than the former. These findings further support our argument that taxing MoMo transactions is an alternative funding paradox or financial inclusion paradox. Females continue to face greater financial exclusion than males,⁹ and our findings could be used to guide policymakers in addressing barriers to females’ financial inclusion.

Conclusion

MoMo is becoming an integral part of the banking system in developing countries. To understand the implications of taxing Momo transactions and inform public policy, research is needed to examines personal, behavioral and psychological factors in tandem with institutional factors, such as citizens’ trust in government. Such an approach provides a more comprehensive framework for understanding MoMo usage, extending insights from earlier literature that has focused solely on personal or behavioral factors while neglecting perceptions of relevant institutions.

This study investigates MoMo usage intentions amid the introduction of the MoMo tax in Ghana and the attitudes toward MoMo taxation. This is achieved through the outlining and testing of a conceptual framework integrating technology acceptance, perceived trust in government, and tax burden perceptions. By applying SEM analytical techniques to survey data from 892 individuals from across all 16 regions of Ghana, we analyzed five factors relevant to MoMo adoption and tax payment: perceived trust in government, perceived tax burden, attitudes toward MoMo taxation, and the two main behavioral components of the TAM (perceived usefulness and ease of use).

Our empirical results reveal two significant determinants of attitudes toward MoMo tax payments (i.e., perceived trust in government and perceived tax burden), one significant determinant of perceived tax burden (i.e., perceived trust in government), one significant determinant of perceived usefulness of MoMo services (i.e., perceived ease of use), and three significant determinants of MoMo service use/tax payment intentions (i.e., perceived trust in government, attitudes toward MoMo taxation, and perceived usefulness of MoMo services).

Practical and Policy Implications

Our findings provide insights that can inform the formulation of tax policies, especially for taxing transactions via alternative financing services that play a role in enhancing financial inclusion. Inappropriate timing of tax introduction might jeopardize the relevance of these tools for financial inclusion. Taxes should be introduced when trust in government seems positive, thus improving compliance attitudes and intentions while reducing perceived tax burden. From a national policy standpoint, a country-level assessment of individuals’ trust in government may be relevant before making decisions on tax introduction. Here, to help mitigate some public concerns, governments may publish the achievements made thanks to taxes collected, such information may also be possible to convey digitally to digital taxpayers, which may both enhance trust in government and positive attitudes towards tax payment.

In addition, the findings on perceived usefulness and ease of use of MoMo services suggest that MoMo service providers should pay attention to improving the relevance of their technology and reducing user-related efforts. Including additional services such as loan facilities and credit scores while reducing the user effort needed to complete transactions is key to ensuring individuals’ continued use of MoMo services amid tax obligations. The addition of such services may enhance financial inclusion by acting as points of entry into the mainstream financing sector. For example, credit scores could be used to assess an individual’s creditworthiness when seeking traditional sources of funding, such as bank loans, rather than imposing collateral requirements. Given the effects of perceived tax burden, ease of use and usefulness observed in this study, it would be beneficial if tax policies on MoMo services allow deduction of interest paid on loans (loans taken from MoMo service providers by individuals) from MoMo tax obligations.

Theoretical Implications

This study has several theoretical implications. First, our integrated model extends the determinants of attitudes toward MoMo taxation and MoMo services use/tax payment intentions by including multidimensional government trust components. In doing so, this study provides a comprehensive understanding of the nuanced relationships between governance tax systems and fintech adoption broadly, and MoMo technology more specifically. Although prior work has established the relevance of individual theories for MoMo adoption, this study is the first to suggest an integration of several theories.

Second, we confirm that although taxing transactions via fintech services that promote financial inclusion, such as MoMo services, may represent an attractive source of government revenue (Mpofu, 2022), such taxation also represents a paradox thanks to its adverse effects on financial inclusion. Here, such taxation may jeopardize the benefits of MoMo services for financial inclusion, particularly among females, who exhibit higher perceptions of MoMo usefulness but less favorable attitudes towards MoMo tax payments than males. These findings contribute to the ongoing theoretical discussion on the implications of MoMo taxation broadly, and with respect to gender differences in financial inclusion. Furthermore, our findings that tax burden directly influences attitudes toward MoMo tax payments and indirectly influences MoMo services use/tax payment intentions are novel and further enhances the understanding of digital taxation payment intentions.

Third, this study contributes to an emerging stream of the literature that focuses on the impact of trust in government on FinTech services’ adoption, and MoMo service adoption as part of it. Trust is both a highly subjective psychological construct and ontologically attributable to one’s perceptions of the competence, integrity, and benevolence of others (Gefen, 2002; Mayer et al., 1995). Despite this complexity, the prior literature has examined MoMo users’ trust using binary measurements (Amoah et al., 2023; Anyidoho et al., 2023; Tetteh et al., 2023). The present study is the first to examine MoMo users’ perceived trust in government while fully anchoring its measurement in all the ontological attributes underlying the concept as suggested by trust theory.

Limitations and Directions for Research

While this study presents interesting insights and complements the theoretical arsenal for explaining attitudes and behavioral intentions in fintech adoption, it also has certain limitations that provide directions for future research.

First, although our study is based on a large sample covering all regions of the study context (Ghana), the generalizability of the findings beyond Ghana is limited. Studies in different national contexts are needed to validate the applicability of our theoretically integrated and extended model in explaining MoMo tax attitudes and MoMo services use/tax payment intentions. Comparative analyses of MoMo tax regimes that differ in terms of MoMo infrastructural environments, cultural contexts, and government trust levels would be particularly valuable. Similarly, future studies could compare countries with different levels of economic development while capturing various levels of channels of access to finance to determine whether MoMo is an “improvement-driven” fintech tool or a “necessity-driven” fintech tool.

Second, research on the adoption of mobile services has revealed that incorporating behavioral theories (such as TPB) in addition to traditional information system theories significantly enhances the understanding of mobile service adoption behaviors (Gopi & Ramayah, 2007). In addition, various versions of information systems theories, such as TAM2 (Venkatesh & Davis, 2000), propose relevant relationships between specific TAM and TPB variables, which have been validated for other alternative finance platforms (Baah-Peprah, 2023) and mobile banking services (Amoah et al., 2020). TPB has proved to be a strong theoretical lens for explaining tax-related compliance intentions and can be used to improve tax-related systems (Taing & Chang, 2021). It would be valuable to employ the full TPB as an alternative theoretical framework while complementing the effects of attitudes with those of subjective norms and perceived behavior control as prescribed by the theory.

Third, although our survey data were collected immediately after the introduction of the MoMo tax, and hence represents data from a unique point in time for assessing attitudes toward MoMo taxation and intentions toward continued MoMo usage, future studies may present longitudinal data based on changes in relevant policies, and following longer experience with such taxation.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Prince Baah-Peprah

Samuel Anokye Nyarko

Notes

References

Ajzen

(1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T

Akinyemi

B. E.

Mushunje

(2020). Determinants of mobile money technology adoption in rural areas of Africa. Cogent Social Sciences, 6(1), Article 1815963. https://doi.org/10.1080/23311886.2020.1815963

Ali

M. M.

Cecil

H. W.

Knoblett

J. A.

(2001). The effects of tax rates and enforcement policies on taxpayer compliance: A study of self-employed taxpayers. Atlantic Economic Journal, 29(2), 186–202. https://doi.org/10.1007/BF02299137

Amoah

Korle

Asiama

R. K.

(2020). Mobile money as a financial inclusion instrument: What are the determinants? International Journal of Social Economics, 47(10), 1283–1297. https://doi.org/10.1108/IJSE-05-2020-0271

Amoah

Kwablah

Dzeha

G. C.

(2023). Trust in government and electronic levy payment decisions in Ghana. Urban Governance, 3(4), 252–258. https://doi.org/10.1016/j.ugj.2023.08.004

Anderson

J. C.

Narus

J. A.

(1984). A model of the distributor’s perspective of distributor-manufacturer working relationships. Journal of Marketing, 48(4), 62–74. https://www.jstor.org/stable/1251511

Anyidoho

N. A.

Gallien

Rogan

van den Boogaard

(2023). Mobile money taxation and informal workers: Evidence from Ghana’s E-levy. Development Policy Review, 41(5), Article e12704. https://doi.org/10.1111/dpr.12704

Arestoff

Venet

(2017). Learning to walk before you run: Financial Behavior and mobile banking in Madagascar. https://hal.science/hal-01491217

Armstrong

S. J.

Overton

T. S.

(1977). Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14(3), 396–402. https://doi.org/10.1177/002224377701400320

10.

Baah-Peprah

(2023). Explaining reward crowdfunding backers’ intentions and behavior. Baltic Journal of Management, 18(2), 262–281. https://doi.org/10.1108/BJM-07-2022-0268

11.

Baah-Peprah

Shneor

(2022). A trust-based crowdfunding campaign marketing framework: Theoretical underpinnings and big-data analytics practice. International Journal of Big Data Management, 2(1), 1–24. https://doi.org/10.1504/IJBDM.2022.119453

12.

Baah-Peprah

Shneor

Munim

Z. H.

(2024). “In this together”: On the antecedents and implications of crowdfunding community identification and trust. Venture Capital, 1–31. https://doi.org/10.1080/13691066.2024.2310232

13.

Baganzi

Lau

A. K.

(2017). Examining trust and risk in mobile money acceptance in Uganda. Sustainability, 9(12), 2233. https://doi.org/10.3390/su9122233

14.

Bagozzi

R. P.

(1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94. https://doi.org/10.1007/BF02723327

15.

Barrett

(2007). Structural equation modelling: Adjudging model fit. Personality and Individual Differences, 42(5), 815–824. https://doi.org/10.1016/j.paid.2006.09.018

16.

Bongomin

G. O. C.

Yourougou

Munene

J. C.

(2019). Digital financial innovations in the twenty-first century. Journal of Economic and Administrative Sciences, 36(3), 185–203. https://doi.org/10.1108/JEAS-01-2019-0007

17.

Bouckaert

Van de Walle

(2003). Comparing measures of citizen trust and user satisfaction as indicators of ‘good governance’: Difficulties in linking trust and satisfaction indicators. International Review of Administrative Sciences, 69(3), 329–343. https://doi.org/10.1177/0020852303693003

18.

Bradley

Henseler

(2007). Modeling reflective higher-order constructs using three approaches with PLS path modeling: A Monte Carlo comparison.

19.

Chin

W. W.

Marcolin

B. L.

Newsted

P. R.

(2003). A partial least squares latent variable modeling approach for measuring interaction effects: Results from a Monte Carlo simulation study and an electronic-mail emotion/adoption study. Information Systems Research, 14(2), 189–217. https://doi.org/10.1287/isre.14.2.189.16018

20.

Cook

Heath

Thompson

R. L.

(2000). A meta-analysis of response rates in web-or internet-based surveys. Educational and Psychological Measurement, 60(6), 821–836. https://doi.org/10.1177/00131640021970934

21.

Cronbach

L. J.

(1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297–334. https://doi.org/10.1007/BF02310555

22.

Davis

F. D.

Bagozzi

R. P.

Warshaw

P. R.

(1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982–1003. https://doi.org/10.1287/mnsc.35.8.982

23.

Della Peruta

(2018). Adoption of mobile money and financial inclusion: A macroeconomic approach through cluster analysis. Economics of Innovation and New Technology, 27(2), 154–173. https://doi.org/10.1080/10438599.2017.1322234

24.

Deng

Yang

Marcoulides

K. M.

(2018). Structural equation modeling with many variables: A systematic review of issues and developments. Frontiers in Psychology, 9(1), 580. https://doi.org/10.3389/fpsyg.2018.00580

25.

Donovan

(2012). Mobile money for financial inclusion. Information and Communications for Development, 61(1), 61–73. https://doi.org/10.1596/978-0-8213-8991-1

26.

Fornell

Larcker

D. F.

(1981). Structural equation models with unobservable variables and measurement error: Algebra and statistics. Journal of Marketing Research, 18(3), 382–388. https://doi.org/10.2307/3150980

27.

Gai

Qiu

Sun

(2018). A survey on FinTech. Journal of Network and Computer Applications, 103(3), 262–273. https://doi.org/10.1016/j.jnca.2017.10.011

28.

García

López Martín

Salmerón

(2015). Collinearity: Revisiting the variance inflation factor in ridge regression. Journal of Applied Statistics, 42(3), 648–661. https://doi.org/10.1080/02664763.2014.980789

29.

Gefen

(2002). Reflections on the dimensions of trust and trustworthiness among online consumers. ACM SIGMIS - Data Base: The DATABASE for Advances in Information Systems, 33(3), 38–53. https://doi.org/10.1145/569905.569910

30.

Gefen

Straub

Boudreau

M.-C.

(2000). Structural equation modeling and regression: Guidelines for research practice. Communications of the Association for Information Systems, 4(1), 7. https://doi.org/10.17705/1CAIS.00407

31.

Gefen

Straub

D. W.

(2000). The relative importance of perceived ease of use in IS adoption: A study of e-commerce adoption. Journal of the Association for Information Systems, 1(1), 8. https://doi.org/10.17705/1jais.00008

32.

Gopi

Ramayah

(2007). Applicability of theory of planned behavior in predicting intention to trade online: Some evidence from a developing country. International Journal of Emerging Markets, 2(4), 348–360. https://doi.org/10.1108/17468800710824509

33.

Grapentine

(2000). Path analysis vs. structural equation modeling. Marketing Research, 12(3), 12.

34.

Grewal

Cote

J. A.

Baumgartner

(2004). Multicollinearity and measurement error in structural equation models: Implications for theory testing. Marketing Science, 23(4), 519–529. https://doi.org/10.1287/mksc.1040.0070

35.

Hair

J. F.

Jr. Sarstedt

Hopkins

G Kuppelwieser

(2014). Partial least squares structural equation modeling (PLS-SEM): An emerging tool in business research. European Business Review, 26(2), 106–121. https://doi.org/10.1108/EBR-10-2013-0128

36.

Hair

J. F.

Risher

J. J.

Sarstedt

Ringle

C. M.

(2019). When to use and how to report the results of PLS-SEM. European Business Review, 31(1), 2–24. https://doi.org/10.1108/EBR-11-2018-0203

37.

Hair

J. F. J.

Black

W. C.

Babin

B. J.

Anderson

R. E.

(2010). Multivariate data analysis (7th ed.). Pearson.

38.

Hardin

(1991). Trusting persons, trusting institutions. Strategy and Choice, 185, 185–209.

39.

Heffernan

C. J.

(1988). Social foundations of thought and action: A social cognitive theory, Albert Bandura Englewood Cliffs, New Jersey: Prentice Hall. Behaviour Change, 5(1), 37–38. https://doi.org/10.1017/S0813483900008238

40.

Henrich

Heine

S. J.

Norenzayan

(2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2–3), 61–83. https://doi.org/10.1017/S0140525X0999152X

41.

Henseler

Ringle

C. M.

Sarstedt

(2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115–135. https://doi.org/10.1007/s11747-014-0403-8

42.

Herrero Crespo

Á.

Rodríguez del Bosque

(2008). The effect of innovativeness on the adoption of B2C e-commerce: A model based on the theory of planned behaviour. Computers in Human Behavior, 24(6), 2830–2847. https://doi.org/10.1016/j.chb.2008.04.008

43.

Hornuf

Safari

Voshaar

(2024). Mobile fintech adoption in Sub-Saharan Africa: A systematic literature review and meta-analysis. SSRN Electronic Journal, 1–72. https://ssrn.com/abstract=

44.

Hsu

C.-L.

Lin

J. C.-C.

(2008). Acceptance of blog usage: The roles of technology acceptance, social influence and knowledge sharing motivation. Information & Management, 45(1), 65–74. https://doi.org/10.1016/j.im.2007.11.001

45.

Hsu

M.-H.

Yen

C.-H.

Chiu

C.-M.

Chang

C.-M.

(2006). A longitudinal investigation of continued online shopping behavior: An extension of the theory of planned behavior. International Journal of Human-Computer Studies, 64(9), 889–904. https://doi.org/10.1016/j.ijhcs.2006.04.004

46.

Johnen

Parlasca

Mußhoff

(2023). Mobile money adoption in Kenya: The role of mobile money agents. Technological Forecasting and Social Change, 191(7), Article 122503. https://doi.org/10.1016/j.techfore.2023.122503

47.

Jones

Basu

(2002). Taxation of electronic commerce: A developing problem. International Review of Law, Computers & Technology, 16(1), 35–51. https://doi.org/10.1080/13600860220136093

48.

Karahanna

Straub

D. W.

(1999). The psychological origins of perceived usefulness and ease-of-use. Information & Management, 35(4), 237–250. https://doi.org/10.1016/S0378-7206(98)00096-2

49.

Keil

Beranek

P. M.

Konsynski

B. R.

(1995). Usefulness and ease of use: Field study evidence regarding task considerations. Decision Support Systems, 13(1), 75–91. https://doi.org/10.1016/0167-9236(94)E0032-M

50.

Kiconco

R. I.

Rooks

Solano

Matzat

(2019). A skills perspective on the adoption and use of mobile money services in Uganda. Information Development, 35(5), 724–738. https://doi.org/10.1177/0266666918788908

51.

Kikulwe

E. M.

Fischer

Qaim

(2014). Mobile money, smallholder farmers, and household welfare in Kenya. PLoS One, 9(10), Article e109804. https://doi.org/10.1371/journal.pone.0109804

52.

Koomson

Bukari

Villano

R. A.

(2021). Mobile money adoption and response to idiosyncratic shocks: Empirics from five selected countries in sub-Saharan Africa. Technological Forecasting and Social Change, 167, Article 120728. https://doi.org/10.1016/j.techfore.2021.120728

53.

Malinga

R. B.

Maiga

(2020). A model for mobile money services adoption by traders in Uganda. The Electronic Journal on Information Systems in Developing Countries, 86(2), Article e12117. https://doi.org/10.1002/isd2.12117

54.

Mardia

K. V.

(1970). Measures of multivariate skewness and kurtosis with applications. Biometrika, 57(3), 519–530. https://doi.org/10.1093/biomet/57.3.519

55.

Mason

C. H.

Perreault Jr

W. D.

(1991). Collinearity, power, and interpretation of multiple regression analysis. Journal of Marketing Research, 28(3), 268–280. https://www.jstor.org/stable/3172863

56.

Mayer

R. C.

Davis

J. H.

Schoorman

F. D.

(1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://www.jstor.org/stable/258792

57.

Mbiti

Weil

D. N.

(2015). Mobile banking: The impact of M-Pesa in Kenya. In Edwards

Johnson

Weil

D. N.

(Eds), African successes, volume III: Modernization and development (pp. 247–293). University of Chicago Press.

58.

Meli

S. D.

Djoumessi

Y. F.

Djiogap

C. F.

(2022). Analysis of the socio-economic determinants of mobile money adoption and use in Cameroon. Telecommunications Policy, 46(9), Article 102412. https://doi.org/10.1016/j.telpol.2022.102412

59.

Mensah

I. K.

Mwakapesa

D. S.

Mensah

P. A.

Mensah

J. K.

(2022). The impact of government electronic transaction levy on the continuous adoption of mobile money transactions in Ghana. In Proceedings of the 4th Africa-Asia dialogue network (AADN) International conference on advances in business management and electronic commerce research, Ganzhou, Jiangxi, China, 25–27 November, 2022, pp. 17–23.

60.

Morawczynski

Pickens

(2009). Poor people using mobile financial services: Observations on customer usage and impact from M-PESA. World Bank. https://www.sidalc.net/search/Record/dig-okr-109869492

61.

Mpofu

F. Y.

(2022). Industry 4.0 in financial services: Mobile money taxes, revenue mobilisation, financial inclusion, and the realisation of sustainable development goals (SDGs) in Africa. Sustainability, 14(14), 8667. https://doi.org/10.3390/economies10090219

62.

Mulaik

(2007). There is a place for approximate fit in structural equation modelling. Personality and Individual Differences, 42(5), 883–891. https://doi.org/10.1016/j.paid.2006.10.024

63.

Murphy

(2004). The role of trust in nurturing compliance: A study of accused tax avoiders. Law and Human Behavior, 28(2), 187–209. https://doi.org/10.1023/B:LAHU.0000022322.94776.ca

64.

Muthiora

Raithatha

(2017). Rethinking mobile money taxation: Taxation and mobile money in Tanzania. GSMA.

65.

Narteh

Mahmoud

M. A.

Amoh

(2017). Customer behavioural intentions towards mobile money services adoption in Ghana. Service Industries Journal, 37(7–8), 426–447. https://doi.org/10.1080/02642069.2017.1331435

66.

Ndung’u

(2019). Could taxation of mobile banking in Africa stall financial inclusion? Africa in focus, Forsight Africa 2019 , (assessed January 2024).

67.

O’brien

R. M.

(2007). A caution regarding rules of thumb for variance inflation factors. Quality and Quantity, 41(5), 673–690.

68.

Ofosu‐Ampong

(2024). New policies, new behaviors: How digital taxation shapes mobile money use in Ghana. Engineering Reports, Article e12860. https://doi.org/10.1002/eng2.12860

69.

Okello Candiya Bongomin

Ntayi

(2020). Trust: Mediator between mobile money adoption and usage and financial inclusion. Social Responsibility Journal, 16(8), 1215–1237. https://doi.org/10.1108/SRJ-01-2019-0011

70.

Osei-Assibey

(2015). What drives behavioral intention of mobile money adoption? The case of ancient susu saving operations in Ghana. International Journal of Social Economics, 42(11), 962–979. https://doi.org/10.1108/IJSE-09-2013-0198

71.

Pavlou

P. A.

(2002). What drives electronic commerce: A theory of planned behavior perspective. Academy of Management Proceedings, 2002(1), A1–A6. https://doi.org/10.5465/APBPP.2002.7517579

72.

Pobee

Jibril

A. B.

Owusu-Oware

(2023). Does taxation of digital financial services adversely affect the financial inclusion agenda? Lessons from a developing country. Digital Business, 3(2), Article 100066. https://doi.org/10.1016/j.digbus.2023.100066

73.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J.-Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. https://doi.org/10.1037/0021-9010.88.5.879

74.

Porter

S. R.

Whitcomb

M. E.

(2003). The impact of lottery incentives on student survey response rates. Research in Higher Education, 44(4), 389–407. https://www.jstor.org/stable/40197312

75.

Rich

Brandes

Mullan

Hagger

M. S.

(2015). Theory of planned behavior and adherence in chronic illness: A meta-analysis. Journal of Behavioral Medicine, 38(4), 673–688. https://rdcu.be/dL4uY

76.

Roberts

L. D.

Allen

P. J.

(2015). Exploring ethical issues associated with using online surveys in educational research. Educational Research and Evaluation, 21(2), 95–108. https://doi.org/10.1080/13803611.2015.1024421

77.

Rojas-Suarez

Pacheco

(2017). An index of regulatory practices for financial inclusion in Latin America: Enablers, promoters and preventers Working paper (August 11, 2017). https://www.cgdev.org/publication/index-regulatory-practices-financial-inclusion-latin-america

78.

Rosseel

(2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02

79.

Scholz

J. T.

Lubell

(1998). Trust and taxpaying: Testing the heuristic approach to collective action. American Journal of Political Science, 42(3), 398–417. https://www.jstor.org/stable/2991764.

80.

Senyo

Karanasios

Gozman

Baba

(2022). FinTech ecosystem practices shaping financial inclusion: The case of mobile money in Ghana. European Journal of Information Systems, 31(1), 112–127. https://doi.org/10.1080/0960085X.2021.1978342

81.

Shapiro

S. S.

Wilk

M. B.

(1965). An analysis of variance test for normality (complete samples). Biometrika, 52(3/4), 591–611. https://doi.org/10.2307/2333709

82.

Shneor

Munim

Z. H.

(2019). Reward crowdfunding contribution as planned behaviour: An extended framework. Journal of Business Research, 103(2), 56–70. https://doi.org/10.1016/j.jbusres.2019.06.013

83.

Shneor

Munim

Z. H.

Zhu

Alon

(2021). Individualism, collectivism and reward crowdfunding contribution intention and behavior. Electronic Commerce Research and Applications, 47(3), Article 101045. https://doi.org/10.1016/j.elerap.2021.101045

84.

Song

Lee

(2016). Citizens’ use of social media in government, perceived transparency, and trust in government. Public Performance and Management Review, 39(2), 430–453. https://doi.org/10.1080/15309576.2015.1108798

85.

Taing

H. B.

Chang

(2021). Determinants of tax compliance intention: Focus on the theory of planned behavior. International Journal of Public Administration, 44(1), 62–73. https://doi.org/10.1080/01900692.2020.1728313

86.

Takyi

P. O.

(2024). The impact of electronic levy on mobile money transactions: Lessons from Ghana. Policy Studies, 1–17. https://doi.org/10.1080/01442872.2024.2303348

87.

Tetteh

C. K.

Amoah

Kwablah

Asiama

R. K.

Ahiabor

(2023). A test of behavioural changes to electronic levy: Evidence from mobile money transactions in a developing country. Cogent Economics & Finance, 11(1), Article 2202963. https://doi.org/10.1080/23322039.2023.2202963

88.

Tobbin

Kuwornu

(2011). Adoption of mobile money transfer technology: Structural equation modeling approach. European Journal of Business and Management, 3(7), 59–77.

89.

Tobbin

P. E.

(2010). Modeling adoption of mobile money transfer: A consumer behaviour analysis. In Paper presented at The 2nd International Conference on Mobile Communication Technology for Development, Kampala, Uganda, 2010. https://vbn.aau.dk/en/publications/modeling-adoption-of-mobile-money-transfer-a-consumer-behaviour-a

90.

Van de Walle

(2017). Trust in public administration and public services. Trust at risk: Implications for EU, 118, 118–128. https://doi.org/10.2777/364327

91.

VanVoorhis

C. W.

Morgan

B. L.

(2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology, 3(2), 43–50. https://doi.org/10.20982/tqmp.03.2.p043

92.

Venkatesh

Davis

F. D.

(2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186–204. https://doi.org/10.1287/mnsc.46.2.186.11926

93.

WorldBank . (2020). Assessing the economic impact of COVID-19 and policy responses in sub-Saharan Africa. An analysis of issues shapping Africa. Africa Pulse .

94.

Yamamura

(2014). Trust in government and its effect on preferences for income redistribution and perceived tax burden. Economics of Governance, 15(1), 71–100.

95.

Yousafzai

S. Y.

Foxall

G. R.

Pallister

J. G.

(2007a). Technology acceptance: A meta‐analysis of the TAM: Part 1. Journal of Modelling in Management, 2(3), 251–280. https://doi.org/10.1108/17465660710834453

96.

Yousafzai

S. Y.

Foxall

G. R.

Pallister

J. G.

(2007b). Technology acceptance: A meta‐analysis of the TAM: Part 2. Journal of Modelling in Management, 2(3), 281–304. https://doi.org/10.1108/17465660710834462