The financial sector is evolving at a rapid pace in this technological era. The deployment of financial technology has reshaping financial services. Digital financial services have the potential to lower costs by maximising economies of scale and increasing the speed of financial transactions (The World Bank, 2020). Digital inclusive finance plays significant role in poverty reduction (Zhou & Wang, 2021). Due to their lower economic status, the people with low incomes are more likely to be the underserved population. Digital financial services promote digital financial inclusion with more tailored financial services for low-income households.

Digital financial inclusion is a process of including those underserved and financially marginalised populations to access financial services through digital means at an affordable cost. Digital financial services have become the key driver and top priority in many countries’ national agendas in achieving financial inclusion (Khera et al., 2021). The successful story of digital money M-PESA in Kenya which accelerated financial inclusion (Liu et al., 2021), shows the significant contribution of digital financial services in improving financial inclusion.

Despite Malaysia being a middle-to-high income economy that has experienced significant economic growth, poverty and income inequality remain persistent concerns for the Malaysian government (The World Bank, 2024). The latest 2022 poverty statistics reveal a contrast in poverty rates across states. Some states experiencing high poverty rates are Sabah (19.7%), Kelantan (13.2%), and Sarawak (10.8%). In contrast, more urbanised states such as Pulau Pinang (2%), Selangor (1.5%), Kuala Lumpur (1.4%), and Putrajaya (0.1%) have a low poverty rate. Other states, including Kedah (9%), Perak (7.5%), Pahang (6.3%), Terengganu (6.2%), Johor (4.5%), Negeri Sembilan (4.4%), Melaka (4.2%), Perlis (4%), and Labuan (2.5%) reported moderate poverty rates (DOSM, 28 July 2023). These disparities show uneven development, which leads to economic imbalances and income inequality.

The Malaysian government has formulated strategic development frameworks such as the Twelfth Malaysia Plan (RMK-12) 2021-2025 and the Malaysia financial sector blueprint 2022-2026 to address the issues. These strategic development frameworks aim to transform Malaysia into a prosperous, inclusive and sustainable country with improved financial well-being of households and advanced digitalisation of the financial sector. Hence, it is crucial to significantly change their lives, especially for low-income households whose income falls below the average national poverty line (PLI). This group, classified as absolute poverty in Malaysia (DOSM, 2023), can benefits from inclusion in the financial system, which provides them with tools, resources and opportunities to pursue financial security and a more promising future.

Although digital financial services offer significant benefits for achieving financial inclusion, many countries still face considerable challenges in using digital financial services to tackle poverty (Ozili, 2018). The reasons highlighted are the lack of trust and confidence (Malady, 2016), connectivity, financial literacy, and social awareness of digital financial services (Aziz & Naima, 2021).

While Malaysia has made significant progress in promoting digital financial services, low-income households continue to experience barriers, particularly in financial literacy, geographical accessibility, and conducting digital transactions (BNM, 2023). Malaysia lags in specific areas compared to neighbouring countries like Singapore and Thailand. For example, a report by OECD (2023) shows that Malaysia has lower fixed (12.4 subscriptions per 100 inhabitants) and mobile (127.4 subscriptions per 100 inhabitants) broadband penetration rate than Singapore (27.4 & 169.6 subscriptions) and Thailand (18.5 & 121.8 subscriptions), which may limit access to digital financial services. Additionally, a report on digital financial services for financial inclusion in Southeast Asia by Kim et al. (2022) highlights that Malaysia lacks a comprehensive cybersecurity law, unlike Singapore and Thailand, which have enacted the Cybersecurity Act, which may affect the trust and the security of digital financial transactions.

Additionally, the Hand Phone Users Survey 2021 (MCMC, 2021) shows that Malaysia has a high penetration rate of smartphones with more than 90% of individuals with income levels below RM3000 owning smartphones. However, accessing banking activities is among the lowest Internet activities. Furthermore, the survey also shows that 50% of the smartphone users in Malaysia do not use mobile payment apps. Hence, suggesting an opportunity to expand digital financial inclusion and empower low-income households with digital financial services.

Previous studies examined financial inclusion with traditional indicators such as outreach, penetration, availability, accessibility and technology (Bongomin & Ntayi, 2020; Morgan & Long, 2020; Vaid et al., 2020) and psychological factors such as financial literacy, trust and financial behaviour (Vaid et al., 2020; Toronto Centre, 2022). Another strand of the literature uses macroeconomic variables like the number of ATMs, the number of depositors, bank branches, deposits per capita, the age dependency ratio and inflation (Omar & Inaba, 2020; Nsiah et al., 2021) to explain financial inclusion. Other studies have also investigated the interaction between financial inclusion, income inequalities (Kling et al., 2020), and financial well-being (Nandru & Rentala, 2019).

Despite the notable advancements in the financial industry, the literature on digital financial inclusion is still developing. Ozili (2018) presented a conceptual framework on digital financial inclusion, highlighting the role of government and financial service providers. Other studies focus on the issues such as adoption of digital financial services (Hossain et al., 2020; Yan et al., 2021), poverty reduction (Zhou & Wang, 2021) and financial inequalities (Demir et al., 2022).

This study aims to expand the existing literature by filling in the gaps related to digital financial inclusion among low-income households who earned below the poverty line (PLI). This study examines the determinants of digital financial inclusion among the low-income households in Malaysia. The investigation on this target population is important because being included in the financial system provides them with accessibility to financial tools that could help them better manage their finances and thus contribute to poverty alleviation in Malaysia. Furthermore, due to the development disparities between Peninsular and East Malaysia, there are significant differences in their economic development. Households experiencing absolute poverty were more prevalent in East Malaysia, whereas Peninsular Malaysia had comparatively lower incidences (DOSM, 2023). East Malaysia has lower access to basic facilities than Peninsular Malaysia (Wan Usamah, 2024). This study suggests that the two regions face different challenges in achieving digital financial inclusion. Hence, a multigroup analysis is conducted to provide a comprehensive view of digital financial inclusion in Malaysia. The study samples are segregated into two groups, Peninsular Malaysia and East Malaysia. It is important to understand the challenges and opportunities for digital financial inclusion in each region and to bridge the development disparities between these regions. This investigation is crucial for narrowing income gaps and ensuring equitable access to financial services across these two regions. The findings provide valuable insights to practitioners, particularly policymakers, in formulating targeted financial inclusion policies and infrastructure development tailored to regional needs. Moreover, understanding the determinants of digital financial inclusion allows financial institutions to design more accessible and inclusive financial services, ensuring the participation of underserved populations in the digital economy. This reduces inequalities, fosters balanced economic development, and creates a more resilient, inclusive financial landscape in Malaysia.

This study makes several contributions. Unlike past studies, this study adopts a broader perspective by examining individual-level factors (digital literacy, financial literacy, and trust) and supporting factors (financial service providers and infrastructure) that facilitate digital financial inclusion. This study collects data from the demand side of the economy, particularly low-income households, to reveal their accessibility to digital financial inclusion and develop an evidence-based model that sheds light on the barriers they face. Furthermore, related studies on financial inclusion, such as Mahdzan et al. (2023) on financial well-being, examined across high-, middle-, and low-income groups, and Rahman et al. (2021) on the B40 group. This study extends the literature by focusing on a subset of the B40 group whose household income falls below the poverty line. Focusing on this group is essential, given their unique challenges in achieving financial stability. A multigroup analysis is conducted to further deepen the analysis by comparing the findings between Peninsular and East Malaysia to provide a comprehensive understanding of regional disparities in digital financial inclusion.

The concept of digital financial inclusion is mainly based on the theories of financial inclusion. Kling et al. (2020) explain that financial exclusion can lead to persistent inequality from an economic theory perspective. The low initial wealth of the poor group faces barriers in the human and physical investment that limit their opportunity for social mobility and perpetuates inequality (Mehrotra & Yetman, 2014). This aligns with the financial inclusion theory that posits that access to affordable financial services is crucial to disadvantaged and low-income groups (Mahendra Dev, 2006). This is supported by Nsiah et al. (2021), who argued that financial inclusion is one of the panaceas to reduce poverty. Financial inclusion provides opportunities for people to benefit from financial services and thus is a key element in ending social exclusion (Mubiru, 2012). The financial sector plays a vital role in improving financial inclusion by providing financial services such as savings, credit, and insurance, thus narrowing income inequality (Park & Mercado, 2018). The financial inclusion indicators, such as the number of bank branches, deposit accounts, outstanding loans, and credit, are important in reducing income inequality over the long run (Verma & Giri, 2024). Omar and Inaba (2020) further highlight that the role of the financial sector, from the dimensions such as penetration, availability, and usage of financial services, influences the financial inclusion level. They found that financial inclusion significantly reduces poverty rates and income inequality, particularly in developing countries, by empowering low-income groups to build wealth and improve their financial stability.

The integration of digital technology into the financial sector enables poor groups to benefit from the formal financial services, contributing to the reduction of poverty (Lumsden, 2018) and income inequality (Yu & Wang, 2021). A study by Aziz and Naima (2021) demonstrated that digital financial inclusion is the interaction between financial inclusion (financial access and literacy), digital inclusion (accessibility, affordability and ability), and social inclusion (social networks and social capital). Ozili

(2018) also highlighted that to achieve the objective of financial inclusion and poverty reduction through digital financial services, a reliable digital financial service infrastructures and financial service providers who improve the delivery of the existing financial services are crucial elements.

The innovation diffusion theory developed by Rogers in 1962 provides an understanding of how an individual adopts new ideas and technology. Five constructs were established to explain the innovation adoption: relative advantage, compatibility, complexity, trialability, and observability (Rogers, 1962). The advantages of digital financial services, such as lower cost and broader outreach, compared to traditional financial services, foster digital financial inclusion. This is supported by the financial-led growth theory, which emphasizes the role of financial development in economic growth (Schumpeter, 1911). Schumpeter argued that an efficient financial system promotes growth by promoting efficient resource allocation and supporting key functions of financial services (mobilizing savings, managing risks and facilitating transactions). These are essential drivers of technological innovation and economic development. The financial development with digital innovation enhances financial accessibility. Studies such as Goldsmith (1969), McKinnon (1973), and Shaw (1973) found the positive impact of financial development on economic growth. Financial development, including the growth of financial institutions, expansion of financial services, and the advancements of financial infrastructure, are key to foster digital financial inclusion, as they encourage the widespread adoption of digital financial services.

To find out how to increase digital financial inclusion for households who earned below PLI in Malaysia, this study proposed five determinants (i.e., digital literacy, financial literacy, trust, digital financial service infrastructure, and the role of financial service providers) as explained.

service platform. (2017) I can make investment through digital financial DFIC5 service platform. DFIC6 I use digital financial service at least once a month. DFIC7 Digital financial service platform is user friendly. Aware about the security of a website (https sites, DIGL DIGL1 safety logo or certificates) before making digital 0.818 financial transactions. Use smart phone to send or receive calls, sends or DIGL2 receives text messages, sends or receives photos. Use smart phone browses or uses the Internet, DIGL3 downloads music, videos or games, makes financial transactions, or access social networking site. Kass-Hanna Know how to open menu, finds the particular menu et al. (2021), DIGL4 options, initiates a transaction, and complete digital Kumar et al. financial transactions successfully. (2022) Know how to correct an error or reverses or cancels DIGL5 a transaction successfully when doing transactions on digital financial service platform. Have an experience to transfer/receive money using DIGL6 digital financial service platform. Have an experience to check my account balance DIGL7 using digital financial service platform. FINL FINL1 I know the importance to save money for the future. 0.818 OECD (2022) I know the importance of financial planning for my FINL2 future. (continued)

Cronbach’s Variables Items Descriptions Sources Alpha FINL3 I know the importance of recording my expenses. FINL4 I think carefully before deciding to buy something. I have adequate awareness about the financial FINL5 products and services offered. I have adequate awareness about the interest rates FINL6 charged on financial products and services. Financial institutions are reliable for availing digital TRU TRU1 0.855 financial services. TRU2 Feel safe when using digital financial services. Bongomin & TRU3 Digital financial service platform is trustworthy. Ntayi (2020) I am confident in using digital financial service TRU4 platform. I have a device (laptop, smartphone, tablet) to access Vaid et al. INFRA INFRA1 0.868 to the digital financial service. (2020) INFRA2 My living area has stable internet connection. INFRA3 I can afford to buy data or subscribe internet line. Opening financial accounts through digital financial INFRA4 service platform is easy. Applying a loan through digital financial service INFRA5 platform is easy. Transferring/receiving money through digital INFRA6 financial service platform is easy. Doing investment and getting wealth INFRA7 management/financial advisory services through digital financial service platform is easy. Purchasing/renewing life/health/general insurance INFRA8 through digital financial service platform is easy. FaceID/TouchID/digital authentication that linked to INFRA9 financial services accounts for transaction authentication is convenient. It is easy to reach financial institutions through call INFRA10 center or LiveChat on digital financial service platform. Financial service providers provide necessary Vaid et al. FSP FSP1 information about digital financial products and 0.751 (2020) services. Financial service providers offer digital financial FSP2 products and services that meet my needs. Formal financial service providers charge lower FSP3 lending rate than informal financial services (main kutu/Ah Long/Biaohui/illegal pawnshop). Financial service providers charges lower lending FSP4 rate than other moneylenders (e-wallet, pawnshop, credit community etc.). FSP5 It is convenient to access digital financial service. Using digital financial services save my effort of FSP6 going to bank/physical store. Notes. DFIC=Digital financial inclusion; DIGL=Digital literacy; FINL=Financial literacy; TRU=Trust; INFRA=Digital financial service infrastructure; FSP=Financial service providers.

This study uses partial least squares-structural equation modelling (PLS-SEM) for hypothesis testing. PLS-SEM is suitable for explanatory studies (Sarstedt et al., 2017), which matches this study. It also does not require normally distributed data and excels in handling models with multiple interrelated latent constructs (Hair & Alamer, 2022). A two-step approach of PLS-SEM was executed, where the measurement model was used to assess the validity and reliability of the constructs prior to the structural model that was used to assess the relationship between the variables (Anderson & Gerbing, 1988). Additionally, multigroup analysis (MGA) techniques were employed to compare the data between Peninsular and East Malaysia. The sample data was segregated into two groups (i.e., Peninsular and East Malaysia). The three-step process of measurement invariance of composites (MICOM) was conducted to determine measurement invariance before the multigroup analysis examining the variations between the two regions (Cheah et al., 2023).

The following model is designed to examine the objective of this study.

Where DFIC (digital financial inclusion) is the dependent variable, and a set of independent variables include DIGL (Digital Literacy), FINL (Financial Literacy), TRU (Trust), INFRA (Infrastructure) and FSP (Financial Service Providers).

Respondent’s Demographic Analysis

Table 2 presents the demographic profile of the respondents. Most respondents were females (80.44%) and Malays (83.43%). About 67.04% of the respondents were married, completed secondary school (53.2%), aged 25-34 years (27.84%), lived in rural areas (68.74%), and earned RM1,501 to RM2,208 (52.69%). The majority of the respondents were from Sabah (15.54%), followed by Sarawak (13.41%), Kedah (12.55%), Perak (11.53%), Kelantan (9.22%), Perlis (8.28%), and Terengganu (6.75%). This data represents the population of households whose income falls below the average national poverty line (PLI).

The results confirm the hypothesis developed in this study. TRU is the most significant factor in determining the DFIC among low-income households in Malaysia. This finding is corroborated by Bongomin and Ntayi (2020) and Setiawan et al. (2024). TRU is an essential factor affecting DFIC. When conducting digital financial transactions, users are exposed to financial risks such as scams and identity theft. A higher level of TRU in digital financial services can lead to greater utilization of these services and, therefore, enhance greater DFIC. Next, DIGL was found to influence DFIC positively. This is supported by previous studies of Jack and Suri (2011), Mujeri and Azam (2018), Lyons et al. (2021) and Lyons and Kass-Hanna (2021) who demonstrated that DIGL increases the usage of digital financial services. When low-income households are equipped with DIGL, they can access digital financial services effectively and are able to take preventive steps against online financial fraud, which boosts DFIC. Furthermore, INFRA is found to be positively associated with DFIC. Studies by Demirgüc- Kunt et al. (2017), D’Silva et al. (2019) and Demir et al. (2022) demonstrated consistent findings where a well- established INFRA increases financial inclusion. The INFRA, which provides reliable internet connectivity and interoperability of digital financial services platforms, facilitates low-income households in conducting digital financial transactions. Meanwhile, FSP only positively affected the DFIC for the complete and Peninsular samples. Demirgüc-Kunt et al. (2017) and Thathsarani et al. (2021) found that FSP that provides innovative digital financial services that meets low-income households’ needs pave the way to DFIC. However, FINL is not found to be associated with DFIC. Bongomin et al. (2016) argued that FINL indirectly influences DFIC. They explained that the absence of social capital hinders the effectiveness of FINL initiatives of poor households in Uganda to access financial services and thus fails to enhance DFIC.

Lastly, the multigroup analysis revealed that DIGL level and a well-established INFRA are essential to boosting the DFIC in Peninsular Malaysia at the same time, the enhancement of the role played by the FSP and increasing the TRU in utilizing digital financial service can efficiently increase the DFIC in East Malaysia. Although DIGL and INFRA are important to both Peninsular and East Malaysia in achieving DFIC, they faced different challenges due to disparities in development. The economy is generally more developed and dynamic, and there are more digital financial activities in Peninsular Malaysia. As a result, low-income households in Peninsular Malaysia are exposed to various forms of digital crimes. Well-equipped with DIGL and a well-established INFRA are crucial for the low-income households in Peninsular Malaysia to protect their online personal and financial information and conduct digital financial transactions effectively and efficiently. On the other hand, the geographical factors of East Malaysia, particularly in remote areas, could present barriers to DFIC among low- income households. FSP may not complete financial services and customer service support in certain areas, which lowers the confidence in utilising digital financial services. The role of FSP is prominent in providing digital financial services with an innovative approach and increasing their TRU to be included in the financial system so that they can fully benefit from the digital financial services.

This study enhances the theoretical understanding of digital financial inclusion by integrating individual-level factors (digital literacy, financial literacy, and trust) and supporting-level factors (digital financial service infrastructure and financial service providers) in predicting digital financial inclusion. This is supported by theories of financial inclusion, such as vulnerable group theory, system theory, and financial literacy theory, as discussed by Ozili (2018) as foundations for understanding financial inclusion. Additionally, few studies in the field compare the differences between Peninsular and East Malaysia. This study reveals significant regional differences between Peninsular and East Malaysia, which are crucial for policymakers and practitioners in developing strategies to improve digital financial inclusion in Malaysia.

The implications are derived from the findings of this study. Enhancing digital financial inclusion among low- income households in Malaysia requires a focus on building trust in digital financial services. Digital financial services expose users to cybersecurity risks such as fraud and data breaches, potentially leading to financial loss. Practitioners must establish cybersecurity and financial regulations and measures to protect users from cyber threats. When users have confidence in the security and reliability of digital financial services, it encourages wider adoption among low-income households, thereby increasing digital financial inclusion.

Furthermore, individuals with digital literacy use digital financial services more effectively. Digital literacy provides people with the knowledge and skills to operate digital devices and engage in secure digital practices. They become aware of cyber threats, understand how to protect their information, and conduct digital financial transactions securely. To enhance digital literacy, practitioners should implement a digital literacy campaign and program to increase awareness and provide resources to enhance digital literacy. Platforms such as broadcast media and social media can effectively spread information and improve digital literacy.

Moreover, a well-established digital financial service infrastructure is crucial to digital financial inclusion. Key components of digital financial service infrastructure include a user-friendly interface and stable, reliable internet connectivity, which are prerequisites for effectively and efficiently accessing and conducting digital financial transactions. Practitioners like the government and telecommunication companies should work together to develop and provide affordable and reliable internet connectivity. Meanwhile, financial service providers can enhance the features of their digital financial service platform to ensure they are user-friendly, secure and accessible. These efforts would contribute to increasing digital financial inclusion.

DFIC is important for economic development, poverty reduction, and eliminating inequality in a nation. This study aims to examine factors determining DFIC among low-income households in Malaysia. Our first findings show that, of the five factors, TRU, DIGL and INFRA affected DFIC among low-income households in Malaysia for the total sample. However, FSP only significantly impacts DFIC for the complete and Peninsular Malaysia samples. Interestingly, FINL is the only construct that is not related to DFIC. Instead, FINL might influence DFIC indirectly, particularly low-income households, as they are more likely to rely on informal networks, such as family or friends, to meet their financial needs rather than engage with the formal financial system, even if they possess financial knowledge.

Moreover, the PLS-MGA analysis demonstrated that all the constructs, except FINL, were significantly different between Peninsular and East Malaysia. We found that DIGL and INFRA are more important in increasing the DFIC among low-income households in Peninsular Malaysia, while the role of FSP and TRU is more effective in increasing DFIC in East Malaysia.

Several limitations are present in this study. This study examined the direct relationship between the variables without considering other factors such as financial behaviour and financial attitude that might mediate or moderate the relationships. This could lead to an incomplete understanding of how other factors interact with these relationships. Furthermore, this study focused only on households that earned below the poverty line in Malaysia. The investigation on different income level groups could offer a more comprehensive view of how digital financial inclusion in Malaysia varies across different socio-economic statuses.