Financial Inclusion in Rural South Africa: A Qualitative Approach

Abstract

Financial inclusion efforts have resulted in a rapid increase in access to financial services. However, the usage of these financial services has not expanded at the same pace, especially in rural areas. The paper explores the factors that have caused usage to lag behind access using a qualitative approach. Data is collected from two predominantly rural provinces in South Africa using focus group discussions. While supply-side factors of distance and transaction costs are important, demand-side factors, including lack of employment, low and irregular incomes, financial illiteracy, and risk and trust perceptions, play a more significant role. We suggest that creating an enabling environment for the development of mobile money could overcome proximity barriers and result in better inclusion of rural communities. There is a need to invest in technology to improve network and Internet reception in rural areas. In addition, the government needs to reconsider the exclusive issuance of e-money by banks. Partnerships with supermarket money markets also have the potential to expand financial inclusion. Moreover, post-adoption financial education should complement efforts to expand financial inclusion. Simplified and transparent cost structures could help resolve the mistrust of banks.

1. Introduction

Financial inclusion has been recognized as a key enabler in poverty reduction. It is an important tool in alleviating poverty and enabling inclusive growth. Through access to financial services, even the very poor can save. For instance, Chakrabarti and Sanyal (2016) show that even those who live on less than a dollar a day seek to manage their money by saving when they can and borrowing when needed. Similarly, access to microcredit has been confirmed to benefit the poor. The literature shows that access to credit can improve the living standards of the poor (Demirguc-Kunt et al. 2017; Ozili 2021). It can allow them to have the capital to start new businesses, raise household incomes, reduce food insecurity, and improve access to services such as health (Rosenberg 2010).

In rural areas, access to credit has been shown to improve access to farming implements and increase agricultural output (Abraham 2018). The resulting increases in income can allow households to smooth consumption and make investments in education which can help break the poverty cycle (Garikipati et al. 2016; Zulfiqar 2016; Kandulu et al. 2020). Moreover, access to credit and insurance services can help the poor mitigate risk and protect themselves against shocks to income. Related, access to different payment mechanisms has increased the efficiency and volume of remittances central to the livelihoods of the poor, especially in rural areas. Therefore, including the poor in all these services could have a very significant effect on poverty alleviation.

The definition of financial inclusion takes various forms, making measuring it somewhat unclear. Most definitions, especially in the early literature, confined the definition of financial inclusion access. Empirical measurements, for example, have leaned towards the possession of a deposit or transaction account, access to credit and other access facilities such as the number of ATMs per given population. Discussions within global organizations have also focused mainly on access (AFI 2016; World Bank 2014). The Maya declaration, a treaty that binds almost all developing countries through a commitment to 80 measurable targets to increase financial inclusion in its initiation in 2011, almost totally focused on access.1

The literature, however, has increasingly recognized that providing access to financial services does not necessarily have the expected impacts on those who are supposed to use the services but do not use them. The (World Economic Forum 2018) argues that sustainable uptake and usage of financial services, which positively affect welfare, requires a greater understanding of the factors affecting the demand and supply side of financial inclusion.

Evidence shows that many people, especially in developing countries, do not use financial services even when they have access to them. For instance, Dupas et al. (2016) conducted an experiment in three countries, including Uganda, Malawi, and Chile. They removed the cost barriers of opening and maintaining a basic savings account. All the operating and account maintenance fees were paid. They found that after two years, of the individuals who opened accounts, only 17% in Uganda, 10% in Malawi, and 3% in Chile were active users. Active users are individuals who made at least five deposits in two years. Similar results are reported by (Prina 2015) for Nepal. The South African government responded by partnering with the country’s biggest four banks and developing an entry-level bank account called the Mzansi account. By 2008, there was an uptake of more than 6 million accounts. Of the 6 million accounts opened, only 3.2 million have been used (Hanouch 2012; Shipalana 2019).

Financial inclusion can be voluntary or involuntary. Voluntary exclusion exists when individuals choose not to use the services, perhaps because they do not need such services or may be able to access the services through someone else. Therefore, the focus of financial inclusion efforts is to reduce the number of people who are involuntarily excluded from financial services. Several factors can explain involuntary exclusion. Firstly, it may be that financial services are costly. Associated costs include monetary costs such as withdrawal fees, bank charges, interest rates, and travel to service points (Karlan et al. 2014). Other factors may include non-monetary barriers such as the lack of trust in financial institutions and the lack of adequate product knowledge. Product knowledge can be due to a lack of literacy on the user’s part but can also result from complex product information making it very difficult for users to understand (Karpowicz 2016; Abel et al. 2018). In addition, regulation can deter the use of financial services. One example is compliance with documentation which some consumers in marginalized populations may not have. For instance, the literature shows that the lack of national identity cards can prevent accessing financial services (Karpowicz 2016; Abel et al. 2018).

Given this, the paper examines the factors that influence the use of financial services in rural South Africa. The South African financial sector is highly developed and almost at the same level as the financial markets of highly developed economies. Formal inclusion in the country is high when measured using access. About 67% of adults had accounts at a formal financial institution in 2017 compared to the sub-Saharan African average of 39% (Demirguc-Kunt et al. 2018). In addition, significant efforts have been made to provide banking services in rural areas through the Mzansi account initiative (James 2014; National Treasury 2020). Nevertheless, the usage of the services is very low, especially in rural areas.

The paper makes two main contributions. Firstly, the literature on factors affecting usage of financial services is relatively nascent compared to that on access to financial services. For this reason, we use a narrative approach to identify the factors and understand the whys associated with those factors. Moreover, most existing empirical literature focuses on supply-side factors (Yangdol and Sarma 2019). Although much has been done to remove supply-side barriers, usage of financial services in rural areas remains low. Therefore, the study contributes to a general understanding of factors affecting the usage of financial services and specifically addresses the impact of policy initiatives like the Mzansi initiative. Secondly, the paper addresses the usage of financial services in a rural setting. There is very scant literature on financial inclusion in rural Africa. Perhaps more importantly, looking at financial inclusion in a rural area in a country like South Africa can highlight the detail that gets lost when the focus is on national data, which ignores the access and usage disparities between rural and urban areas.

The paper proceeds next with a discussion of financial inclusion in South Africa. That discussion is followed by a discussion of the methodology used. The results are then presented, followed by a discussion and conclusion.

2. Brief Literature Review on Determinants of Financial Inclusion

Financial inclusion has been cited as a key driver of inclusive growth and rural development. However, rural areas still experience barriers to financial inclusion, which are higher than those experienced in urban areas. The barriers to financial inclusion can be classified into supply-side and demand eligibility factors. We will discuss each of these below.

2.1. Supply-Side Factors

Supply-side factors arise from the way financial institutions operate or offer services. The literature identifies three main supply-side factors. These include eligibility (such as documentation and inappropriate screening criteria), affordability (including bank charges, minimum balances and interest rates), and availability or proximity factors (including distance to service points and poor infrastructure).

2.1.1. Eligibility Factors

Eligibility factors include documentation and screening criteria. Insufficient and unclear documentation has been cited as an essential exclusion factor, especially among young adults in rural areas (Demirgüç-Kunt and Klapper 2012; Demirguc-Kunt et al. 2018). This barrier can exhibit itself in the form of high document requirements or simply the requirement of documents that excluded populations cannot provide, such as payslips and other income flow documentation. In addition, the documentation that banks need to comply with national regulations for onboarding clients by financial services providers can be very onerous, leading to financial exclusion (National Treasury 2020). Demirguc-Kunt et al. (2018) find that 20% of adults without an account cited lack of documentation as the main reason. This rate was higher in some countries, such as Zimbabwe, where 49% of adults reported not opening an account due to a lack of required documentation. Similar results are found by (Abel et al. 2018).

Selection criteria by financial service providers also present an obstacle to financial inclusion. Financial markets are fraught with information and symmetries which make that assessment of potential customers. In addition to the documentation barrier mentioned above, rural populations are less likely to have regular incomes, credit histories, or collateral to provide financial services providers with the relevant information or mitigation options for assessing and managing risk. Traditional risk assessment results in high levels of exclusion from financial services (Dlamini and Simatele 2021; Prabhakar and Weber 2020; Chen and Yuan 2021). The provision of inappropriate products can also lead to exclusion. More customizable and specialized rules for excluded populations will likely promote financial inclusion (Varghese et al. 2018; Aduda and Kalunda 2012).

2.1.2. Affordability

When financial services are available at prices that potential users cannot afford, they choose not to use them. Associated costs include account maintenance fees, account minimum balances, processing fees, interest rates, and remittance fees. The evidence shows that the poor and those who live in rural areas are more likely to be excluded from financial markets because of cost (Zins and Weill 2016). Demirguc-Kunt et al. (2018) show that 60% of adults without accounts indicated that they did not have accounts due to cost-related barriers. One contributing factor to this high cost is the high cost of serving low-income and rural populations (Aduda and Kalunda 2012). Developments in technology, however, can reduce these costs resulting in much lower costs of providing financial services to rural and low-income consumers (Manyika et al. 2016; Dube et al. 2021).

The related literature shows that digital finance, such as mobile money, has the potential to increase financial inclusion (Suri 2017; Demirguc-Kunt et al. 2018). Nevertheless, this seems to be only for specific services. The challenges of designing digital financial services that match the quality and benefits offered by over-the-counter services and products are still significant (Wang and He 2020; Demirguc-Kunt et al. 2018; Zhao 2016). In addition, the literature shows that the provision of digital finance to rural areas is faced with challenges in unequal and insufficient infrastructure as well as significant digital divides from the urban areas (Salemink et al. 2017; Liu et al. 2021). Moreover, low and irregular incomes also lead to financial exclusion (Park and Mercado 2018; Abel et al. 2018; Nkuna et al. 2018). The World Bank survey shows that nearly two-thirds of those without bank accounts indicated they could not open them because of insufficient and irregular incomes. A fifth of all respondents cited lack of income as the only reason they were not opening a bank account (Demirguc-Kunt et al. 2018).

2.1.3. Availability and Proximity

The literature shows a positive correlation between financial services access points and financial inclusion. Therefore, travel to the nearest service point is likely to harm financial inclusion. Users must travel long distances to points of service, which is very costly, especially given low incomes (Abel et al. 2018; Nkuna et al. 2018). Demirguc-Kunt et al. (2018) show that 22% of adults without a bank account do not have one because the service points are too far away. In some countries such as Brazil, Indonesia and Kenya, this rate is higher at 41%. Technological developments which have led to the introduction of ATMs, debit and credit cards, mobile banking, and mobile money have effectively reduced the distance between two service points. For instance, Bachas et al. (2018) show that using debit cards reduces the median road distance to the service point by 70%. Mobile money is also emerging as the most effective way to bridge the distance to the point of service in geographically excluded areas due to the high level of mobile phone ownership. However, the adoption of financial technology is limited amongst those who could benefit from it. For example, ATMs, the primary vehicle for reducing distance through card use, are mainly an urban phenomenon. In addition, access to financial technology in areas where it could benefit many has been hampered by inadequate infrastructure and a lack of relevant regulation (Liu et al. 2021; Simatele and Mbedzi 2021).

2.2. Demand-Side Factors

Demand-side factors are related to consumer constraints. These include financial literacy, financial capability, and psychological and cultural barriers.

2.2.1. Financial Literacy

The literature indicates a positive relationship between financial literacy and financial inclusion (Goyal and Kumar 2021; Grohmann et al. 2018). Here, we define financial literacy as the ability to understand and apply financial concepts and risks and the skill, motivation, and confidence to apply this knowledge when making decisions in different contexts (OECD 2014). Moreover, financial literacy affects financial behavior and improves decision making resulting in better risk management (Paiella 2016; Goyal and Kumar 2021; Allgood and Walstad 2016). For example, evidence shows that financial literacy is positively correlated with savings behavior and portfolio choice. Individuals with higher levels of financial literacy are more likely to make better decisions about debt (Lusardi and Tufano 2015) and retirement planning (Gallego-Losada et al. 2022; Niu et al. 2020). The study by (Allgood and Walstad 2016) finds that the effect of perceived (self-assessed) literacy has greater efficacy on financial behavior than measured financial literacy. The development of this ability and skill can be aided through financial education and ensuring that customer information about concepts and products is unambiguous and in the appropriate language (Agwu 2020). Triki and Faye (2013) indicate that financial literacy is one of two key barriers to financial inclusion in Africa. Hasan et al. (2021) show that the effect of financial literacy is more pronounced in rural areas.

The growing use of digital finance has heightened the need for financial literacy. Morgan et al. (2019) argue that a greater level of financial literacy is needed to benefit from financial services via fintech fully, as well as to minimize financial fraud and the loss of personal information. While digital finance is potentially very beneficial, digital literacy in many areas is low and can work as a barrier to using digital financial services (Nkuna et al. 2018; Ebong and George 2021).

2.2.2. Financial Capability

Although financial literacy and capability are similar and often overlap, they are not the same. Financial literacy focuses on knowledge and skills (OECD 2014), while financial capability is broader and encompasses attitudes and behaviors related to finance (World Bank 2014) or the motivation and ability to plan financial matters (Shankar 2013). The literature on financial capability is still nascent. However, evidence is growing supporting a positive relationship between financial inclusion and financial capability. Potocki and Cierpiał-Wolan (2019) find a positive association between financial capability and financial inclusion for rural populations in Poland.

2.2.3. Psychological and Cultural Factors

Psychological factors such as trust can have a significant effect on financial behavior. Evidence shows that trust has a positive effect on financial inclusion. Trust lowers the perceived risk of transactions and financial contracts (Xu 2020; Li et al. 2019). As a result, individuals are more willing to use financial services. For example, Galiani et al. (2020) show that increasing trust significantly increases savings. Ghosh (2021) shows that it significantly affects both the ownership and use of bank accounts. Moreover, trust plays a significant role in the adoption and use of digital finance, which is a crucial factor in expanding financial inclusion (Arif et al. 2016; Simatele, forthcoming). Increasing evidence shows that low-income consumers do not open bank accounts because of a lack of trust in banks (Xu 2020; Barajas et al. 2020).

3. Financial Inclusion in South Africa

South African financial markets are very well developed, with many services and service points. The National Treasury (2020) has argued that in South Africa, acquiring a basic bank account is the gateway to full financial inclusion. Therefore, government efforts aim to increase the number of people with bank accounts. This thinking underpins the Mzansi accounts program, developed in 2004 as a partnership between the government, the four big banks and the South African post bank. This account was developed as a basic account for low-income consumers. The flagship aspects of the Mzansi account were offering an account with a debit card, basic transactions facilities at a capped cost and no monthly or management fees. Interoperability was also included so individuals could withdraw money from any Postbank nationwide.

The second push was the introduction of the South African Social Security Agency (SASSA) MasterCard in 2012.2 These cards allow recipients to withdraw cash at ATMs and points of sale without incurring any charges. The two programs resulted in significant uptake of bank accounts, as shown in Figure 1. The percentage of formally included individuals increased from 50% in 2004 to 93% in 2018. The SASSA card drive seems to have had a more significant impact, as shown in

Table 1. Inclusion of Grant recipients.

	2004	2007	2012	2016	2018	2019
South Africa Total		2004	2007	2012	2016	2018
Banked		46	60	67	77	80
Other formal		4	4	6	8
Only informal		12	11	8	3
Excluded		38	25	19	11
Grant/SASSA Recipient
Banked		34	54	76	100
Other formal		6	7	6	4
Only informal		26	22	9	6
Excluded		34	17	9	8

Source of data: (FinMark Trust 2018, 2020).

. By 2016, all SASSA recipients had a bank account.

While the figures show that South Africa has achieved very high levels of financial inclusion, Figure 2 shows a very low usage level. By 2018, about 40% of the accounts were only used as mailboxes or were completely dormant. A mailbox account is one which the owner merely uses to receive salaries and government benefits. Once the salary or grant is deposited, the account owner withdraws all of it and does not use the account until the next deposit is made. On the other hand, a dormant account refers to an account with no financial activity for an extended period, excluding the posting of interest. The number of mailbox accounts is much higher when you look at individuals on the SASSA account, which is used to receive government grants. In 2019, 69% of all accounts held by SASSA recipients were used only as mailboxes. This proportion is much higher than the national proportion of 50%. Therefore, low-income individuals are not engaging with the transactions accounts in a meaningful way. The observed high level of financial inclusion is not accompanied by appropriate use, therefore, not translating into improved quality of life as expected.

The motivation behind expanding bank account ownership was to enable low-income individuals to have a safe place for savings and to enable them to have access to other services that come with account ownership. In addition to using bank accounts mainly as mailboxes, we note that many individuals still prefer to use cash payment despite developments in mobile money. Figure 3 shows the percentage of adults who used various digital financial services in 2017. Only 14% of the 40% poorest account owners used the phone or internet to check their accounts. The ownership of mobile money accounts is also very low, at a national average of 31%. The trends in the other statistics suggest that this figure is likely to reflect higher-income users.

Related, 42% of grant recipients withdraw their money immediately after it has been deposited. About 69% of them use cash payments, underling how cash reliant they are (see Figure 4). Digital finance, especially mobile money, has been applauded for its potential benefits, especially for low-income households. Increased use could help poor households like those on grants cut down on transaction costs associated with using cash.

The lack of usage of services available through transaction accounts is also reflected in other services. The number of people who are saving in their bank accounts is low. As can be seen from Figure 5, the number of people reporting having savings at a formal institution stagnated on average between 2014 and 2019. From a marginal increase of 20% in 2014 to 23% in 2017, the number of people formally saving dropped to 20% in 2019. Informal savings are increasing more rapidly than formal savings. For instance, informal savings aggregated with saving at home increased from 19% in 2014 to 39% in 2019.

A similar pattern is observed in the use of credit. Figure 6 shows the level of credit in the country. Access and usage of credit are minimal. The national average access is 51%, but that rate is higher for rural areas at 63%, as shown in Figure 6.

On Figure 7, this study has identified various reasons that are challenging the par-ticipants. Further discussion about these challenges is made on the data analysis.

Race plays a crucial role in South Africa’s social-economic demographics.

Table 2. Borrowing by demographics 2016.

	Banked	Formally Served	Informally Served	Family/Friends	Not Served
Male	14	33	2	2	49
Female	11	32	2	2	53
Black	8	33	2	2	55
Coloured	11	34	1	1	53
Asian	33	28			39
White	43	29		1	27
Metro	19	36	2	2	41
Small	11	30	2	2	55
Rural	4	30	1	2	63

Source of data: (FinMark Trust 2018).

shows that access to bank credit amongst whites is at least five times as much as access amongst blacks. Similarly, only 27% of whites are not served compared to 55% of blacks. About 53% of colored adults do not have any form of financial services. Consequently, financial inclusion efforts must redress historical imbalances in the financial service sector, especially those in rural communities.

Financial inclusion in insurance is still exclusive. Figure 8 shows the types of insurance used. Most low-income households do not own vehicles and are unlikely to have life insurance. Given that they mostly use public health services, it is safe to conclude that middle- and high-income households mainly use most insurance services shown in Figure 8. However, funeral insurance is quite popular and is the most used insurance service. Berg (2011) argues that this is probably due to the prevalence of informal funeral cover.

4. Methods and Data

4.1. Data

The study was exploratory. For that reason, qualitative data was preferred to get some sense of the perspectives of rural communities about financial services. Qualitative data was collected through focus group discussions. A total of 20 focus groups were selected from the rural communities in the Eastern Cape and KwaZulu Natal provinces. The focus group discussions centered on understanding the different types of financial services used by the participants. In addition, questions were asked about the available services not used by participants within the groups to elicit reasons why these services were not being used.

The participants were purposively selected according to their socioeconomic characteristics. The selection of villages ensured that they were at different distances from the central business district. Care was taken to include a balance across gender and age in inviting people to be part of the focus groups. However, more females were willing to take part in the focus groups. The focus groups ranged between 5 to 12 people. The data was collected over one month in 2018 for both provinces. The demographic profile is given in

Table 3. Participant characteristics.

	Eastern Cape Province			Distance *	KwaZulu Natal Province			Distance *
EC Villages	Males	Females	Total		Males	Females	Total
FG 1	4	4	8	35 km	5	4	9	54 km
FG 2	4	6	10	37 km	8	1	9	20 km
FG3	1	9	10	47 km	5	5	10	80 km
FG4	0	10	10	56 km	5	4	9	79 km
FG 5	4	4	8	70 km	0	10	10	70 km
FG 6	0	10	10	20 km	4	4	8	60 km
FG 7	4	6	10	25 km	2	5	7	50 km
FG 8	4	6	10	65 km	5	1	6	10 km
FG 9,	3	4	7	52 km	2	6	8	10 km
FG10	4	6	10	50 km	6	3	9	15 km
	Age group by gender
Age Group	Males			Females	Total	Respondents (%)
20–35	28			27	55	31%
36–59	25			60	85	47%
60+	17			21	38	22%
Total	70			108	178	100%

* distance to nearest financial service point.

. A total of 178 participants took part in the focus groups, and most of the participants were women.

4.2. Data Analysis

The discussions were conducted in Xhosa and Zulu, the main languages spoken in the rural areas of the two provinces. The discussions were recorded, transcribed, and translated into English. The analysis of the text was done based on the translated transcripts. Because we were looking to identify the reasons for the use or non-use of various financial services and the reasons behind their use or non-use, deductive coding was used. Figure 1 shows the process. Accordingly, the predetermined codes used were services use and reasons for use. The first stage of the analysis was to search the body of text in the translated scripts for the different services used by the focus group participants. Seven types of services were identified, as discussed below. The next step was to identify why each service was used or not used.

This study has highlighted various reasons that are challenging the participants. Some of the reasons given affected the use of more than one service. For instance, high bank charges discouraged the use of bank accounts while at the same time encouraging the use of informal savings. Similarly, the use of informal credit and supermarket money markets were partly explained by irregular incomes. As discussed above, the literature attributes financial exclusion to proximity, price of products, employment and income factors, financial literacy, and capability related factors. In line with the deductive approach, these were chosen as the predestined sub-themes. The reasons identified for use were then grouped into these sub-themes. Not all factors identified in the literature were present. For instance, no cultural factors were identified. We also searched the text for indications of the approximate level of use. The sub-themes were then grouped into two themes per literature: demand-side and supply-side factors. The discussion of the results is based on this classification.

5. Results

5.1. Services Used

Eight different types of services are identified. The use of informal financial services dominates. There is moderate use of bank accounts, although, as indicated below, these are mainly used to receive income of various types. Over 60% of the participants had opened a bank account at least once. However, less than 50% of these were active accounts. Less than 10% of the participants who used bank accounts indicated using services beyond receipt of salaries and grants. The least used forms of services are formal credit and digital finance.

The principal reason for opening accounts was to use them as a mailbox for salaries, grants, and other related transfers, reflecting what the discussion in Section 2 indicated. Once the transfers and salaries had been deposited, the account holders would withdraw the money and wait for further deposits. Only 4 out of the 20 focus groups had participants who indicated having funeral cover. Funeral cover was the only insurance service that participants used. Almost all the groups indicated that supermarkets play an important role in enabling access and usage of financial services. Over 70% indicated using supermarkets to withdraw money either through the money markets or as cash back at the point of sale. However, using cash back at the point of sale is premised on a debit or credit card. Nevertheless, most participants did not realize this was a way of using their bank accounts. The money market emerged as one of the most prominent services.

5.2. Factors Affecting the Use of Financial Services

5.2.1. Supply-Side Factors

Supply-side factors arise from the way financial institutions operate or offer services. The literature identifies three main supply-side factors that affect financial inclusion. These include eligibility (such as documentation and inappropriate screening criteria), affordability (including bank charges, minimum balances and interest rates), and availability or proximity factors (including distance to service points and poor infrastructure) (Ramlee and Berma 2013; Yangdol and Sarma 2019). Such factors typically affect services’ availability and are more likely to affect access, as documented in the literature. We found that affordability and proximity factors also affect usage. Figure 7 shows the associated codes and illustrative narratives.

5.2.2. Proximity Factors

Distances from service points can lead to financial exclusion due to the scarcity of services. Most focus groups indicated that travelling to a bank was very costly. Some communities live as far as 80 km from the nearest service provider. Several participants indicated that they had opened bank accounts but were only using them to receive funds and could not actively use them because of the cost of travel. In this case, geographical exclusion results more in usage effects than access since the participants can still open accounts 80 km away. The use of bank accounts as mailboxes was confirmed, as seen in

Table 4. Supply-side factors: codes and illustrative narratives.

Codes	Theme	Illustrative Narrative
Cost of travel Ques at banks Sleep in town to access the bank	Geographical exclusion	FGD1 (80 km) We spend R64.00 on a taxi, and it takes 1 h 30 min to get to Kwa Nongoma out town. My child tells me that if R200 has been deposited in my account, how much will be left considering the R64.00 taxi fee and the charges? That is one of the reasons we prefer to keep our money.
		FGD2 (50 km) Distance makes the use of financial services unattractive. People are interested in using financial services; however, distances become an obstacle. There is an interest to use banks if they are close. People tend to sleep in town when they expect payment so that they can arrive first at the bank and avoid long queues. Most people in this community struggle to get transport when they go to town. Travelling costs make it difficult to use financial services. We spend R85.00 return to go to town.
		FGD3 (40 km)… It surely does, we spend most of our money on the transport hence we opt to withdraw the money at once.
High travelling costs Unaccounted deductions Loss of money Bank charges	Transactions costs	FGD5 Bank services increase daily, for example you are told there is R45.00 bank charges, and then you notice there are also amounts of R70.00 which are not accounted for.
		FGD20 There are unknown transaction which we call money mouse at the bank. These create a challenge for us. It’s the fluctuating bank charges which is not understandable.
		FGD8: There are unknown transactions which deduct money from clients, the banks do not assist, and they just give you a number to call rather than calling the number.
		It is not clear what the standard amount of bank charges are. Every time we lose our money, we are told about bank charges. It is better to keep my money in my wardrobe than to be a victim of those bank mice that eat our money.

.

Most participants relied on government grants and stipends as a source of income. An average grant is R1104.3 Although the absolute amounts cited in these quotations appear small (between USD 4.12 and USD 5.47),4 they are well over 5% of the average grant income, which is the primary source of income for many households in the villages. As a result, they justified withdrawing all their money and keeping it at home.

The cost of travelling was not only calculated in monetary terms. As pointed out in the illustrative quote from FGD2 in Table 4, the amount of time involved in travelling between the village and the service provider was quite significant, taking time away from economic activities that were generating a livelihood for them. Sleeping in town to prepare for queuing in the morning was not uncommon. Participants also indicated that using mobile banking was not a feasible option to lower costs because it is also associated with bank and usage charges.

5.2.3. Price Exclusion

All the groups complained that the bank charges were one of the main deterrents from using their bank accounts. The charges were perceived to be very high relative to the incomes. Some groups indicated that banks must remove bank charges for the poor or have sliding bank charges matched with income. Bank charges are not only perceived to be high but erratic as well. Many participants indicated that they did not understand how the charges were calculated and that money was often missing from their accounts for no reason. The frequent changes in charges made it difficult for them to understand how much they were expected to pay for maintaining their accounts.5 Many felt that bank charges were also extortionary. Consequently, much mistrust expressed about the use of banks (FGD8, Table 4).

5.3. Demand-Side Factors

Demand-side factors are related to consumer constraints. Four demand-side factors influencing the use of financial services were identified. These include lack of employment, access to income, financial literacy, and risk and trust perceptions. The codes and illustrative quotes used to derive the demand-side factors are shown in

Table 5. Demand-side factors: codes and illustrative narratives.

Theme	Illustrative Narrative
Employment	FGD4 I am not employed, so banks are not places I even think about when I do not have money. For example, you have already heard from the other people that they got loans because they worked for well-known companies. So it is all about status, and I am not at that level, Omashonisa.6 Works well for me.
	FGD17 We do not use financial services because we are unemployed. If you do not have money, a bank will close your account as they did to ours. Banks are for the people with an ongoing source of income. If I had a chance to advise them, I would say they should not close off accounts of people who do not have money.
	FGD10 I borrowed money when I was working, and they gave me. I wanted money again, but I did not get it. I think because I was relying on SASSA.7 I was not eligible as I am of a lower rank. I have borrowed money while working, and I never had challenges.
Income	FGD20 & FGD2 I opened a bank account to receive a salary. FGD9 I started to work, so I needed to transact. FGD12 SASSA and the allowance we get from our kids are our sources of income which drives us to use financial services.
	FGD18 I have tried in the second month of my job, the money they were willing to give me was way less than what I required. I think the people in well-paying jobs can get what they require. The limited salary I receive contributed to this. Due to my low status, I think I am not eligible to get loans from banks. Therefore, I would not even try.
	FGD13 I borrowed R100,000 they gave me because I was making deposits of R10,000 per week on the graduation attire I was selling. So, it is all about your record that makes you attractive to banks.

.

5.3.1. Employment

Formal employment was the most predominantly cited demand-side driver of use. Several participants expressed knowledge of the importance of employment as a criterion for accessing financial services. For that reason, as illustrated in Table 5, many indicated that they never even sought to open a bank account. The view that employment was a necessary condition was reinforced by the experiences of some of the participants who previously held bank accounts. They indicated that their accounts were closed once they retired or had lost their jobs. There was evidence that some understood that this is likely due to the loss of regular income. However, many of them separated the two factors of employment and income. Some participants indicated that it would be helpful to have appropriate accounts for those who did not have a regular income.

Beyond opening accounts, employment also emerged as a reason for not seeking formal credit from banks. Some comments suggested that the importance of employment went beyond access to regular income. For instance, some participants who held a regular income through government grants could not access credit despite having accounts at the same banks.

5.3.2. Income

The sources of income cited included income from microbusinesses, agricultural income, formal employment, and grant income. Only grant income and income from formal employment were considered to be regular. Access to regular income emerged as a motivation for opening bank accounts. Many participants indicated they had bank accounts mainly because they used them to receive salaries and government grants. Conversely, those who had not opened bank accounts indicated that the lack of regular income was one of the main contributing factors. Many participants argued that they did not feel the need to open bank accounts when they knew that there was no regular income (FGD4).

Participants also noted that the regularity and income size of income also affected credit access. Regularity of income was seen as a necessary but not sufficient condition for accessing credit. For example, those on government grants were on a regular income. However, they could not access credit facilities because the amounts were small. Two contrasting comments in Table 5 illustrate this point. In the first case, FGD18 was on a regular but low income and was excluded. In the second case, FGD13, a participant had a mid-level and regular income from a micro-business and a favorable credit history.

5.3.3. Financial Literacy

There was evidence of financial illiteracy related to the basics of bank charges, interest rates and criteria for lending. For example, although many of the participants understood that accounts were associated with charges, it was clear that almost all the participants were not clear on how bank charges work. They decried that deductions from their accounts were made without informing them. Participants showed a lack of understanding of how the charges are calculated and the responsibilities related to maintaining their accounts. Many did not understand that bank charges are applied monthly whether the account was used. They were unaware of what dormancy charges were. In some cases, bank charges resulted in complete depletion of savings and the accounts were closed due to dormancy.

In addition, while many understood that income was a necessary lending criterion, they did not seem to understand that income alone was not enough. FGD10 in

Table 6. Financial literacy and risk perception: codes and illustrative quotes.

Theme	Illustrative Narrative
Financial literacy	FGD2 Bank services increase on a daily basis, for example, you are told there are R45.00 bank charges, and then you notice there are also amounts of R70.00 which are not accounted for. FGD6 My challenge has been with bank charges; its only when I began to deposit money that charges kicked in and I was never aware of this. I feel that bank charges are unnecessary because people do not have money.
	FGD10, I have tried but I did not get it. Strangely the person who gets exactly the same salary as me did get a loan. I am therefore not interested.
	FGD4 I also have never borrowed from a bank. I believe that it is better to get money from a person you know than be told about interest rates that you have no idea of how they are calculated.
	FGD10 No, there is no reason for me to borrow from people I do not know who use strategies or systems known by only them to determine who should get what amount. At least I know the people I borrow from have a willing heart to give.
	FGD11 I do not intend to, the problem is that banks have a very complicated system of determining how much money they will give you. I am probably one of the people that would not get any loan due to fluctuating money that I get. This makes me not attractive to business of banks.
Risk and Trust perception	FGD15 I lost an amount of R1500 from a bank. This happened immediately after I withdrew money. When I went to the bank one of the people there told me that someone used my pin. Since I am the only one who has this pin, tell me, how did they know my pin? How can I trust them going forward? Banks are exposing us by giving our information to people whom we do not know. There should be a confidentiality clause. Our money ends up being deducted and when we track how they know our details, it’s through a bank.
	FDG17 It is not clear what the standard amount of bank charges are. Every time we lose our money, we are told about bank charges. It’s better to keep my money in my wardrobe than to be a victim of those bank mice that eat our money (group laughs).
	FGD6 I had an investment and I did not use the whole money when I took a small portion of it. To my surprise the banks could not trace the money I invested which was left. They could only see that I withdrew some. What angered me is that they never called me regardless of the promise to call.
	FGD17 The unknown deductions from my account were the most frustrating. If these are not known by the bank then who should know? An amount of R500 was deducted from my bank account and I was advised to open a new account as this could not be traced. Remember these are the same people I have given my trust and money to, yet there are things they don’t know.

illustrates this point where one participant felt he was entitled to the same loan as another person with the same income. Furthermore, there was an indication that many of the participants did not understand how interest rates work (FGD4, Table 6). Many of the participants indicated that the bank services were complicated to understand.

5.3.4. Risk and Trust Perceptions

There was a general misapprehension about the way banks administer bank charges. Almost all the groups indicated that the bank charges we applied without explanation or consultation. The general attitude was that there was something sinister about how this was done. FGD15 in Table 6 illustrates some of the frustrations expressed in various groups. There was an attitude of insecurity about keeping money in the bank. Many participants viewed banks with suspicion indicating that they did not treat personal information with care resulting in loss of savings. There was a general feeling that banks do not take responsibility when personal information is lost. A few quotations in Table 6 help to illustrate this point. Participants also expressed a distrust of digital finance. A participant from FGD15 commented, “We used to use books and papers during times of NBS; however, the arrival of cards brought along crime/corruption”, showing the extent of distrust.

6. Discussion

Distance and transaction costs have been highlighted as factors that affect access to financial services for many years. However, they have not been identified as significant contributors to usage. Our results show that distance is still a major factor in the usage of financial services. Many of the participants in the focus groups had opened bank accounts but could not use them because of the distance to the point of service. The literature proposes a concept of doorstep banking (Mushtaq and Bruneau 2019; Kochar 2018), which refers to banking at a location close to the consumer. The same concept of doorstep banking is behind the extensive use of automated teller machines (ATMs) placed away from bank branches but closer to the consumer. However, ATMs are rare in rural South Africa and have mainly remained an urban phenomenon.

Technology is now making it possible to provide financial services closer to consumers. Evidence shows that technology can overcome distance barriers and enable users to access financial services. Related results are found by (Bachas et al. 2018) who show that the use of card payments significantly reduced the distance effect for users in Mexico. Similarly, Aron (2018) shows that mobile money has the same effect. However, infrastructural voids are a barrier. Rural areas need better Internet connections and cheaper data rates for mobile banking to be a reality. Mobile money is a better option because it primarily uses USSD-based technology. Although there is evidence that network connections in rural areas are weak, it is better than Internet connections. However, the growth of mobile money in South Africa has been limited. FinMark Trust (2017) suggests that this is due primarily to the over-regulation of the sector, which mandates that only banks can issue e-money. As a result, the only viable technology is mobile banking or mobile money which relies on collaborations between mobile and network operators and banks. This dominance of banking in the provision of financial services has resulted in a concentrated financial services sector which is not favorable for expanding financial inclusion in rural areas.

Transaction costs mainly related to travelling and bank charges present a barrier to usage. The results suggest that bank charges as a barrier to the usage of financial services are likely related to the lack of knowledge about how financial services work. Many participants indicated that they did not clearly understand the administration and calculation of the bank charges. In effect, the existence of bank charges was not necessarily a barrier to usage but rather the fact that the administration of the charges was not clearly explained and that it seemed these charges were applied erratically. The Mzansi account allows small deposits with a small bank charge of about R3 per month.8 Nevertheless, there are many other small charges associated with the account, which add up to a significant proportion of their incomes which the users do not seem to be aware of. In addition, a dormancy fee is charged when the account is inactive. The dormancy fees could explain the unknown deductions indicated by the participants.

It is unclear whether the cost structure is clearly explained to the users or not understood. In addition, many participants indicated a lack of knowledge about financial products and services. The lack of understanding could emanate from complex bank processes and documentation. On the other hand, it could result from the lack of explaining the procedures to users. Both these factors are likely at play. Although the Mzansi account is meant to be a no-frills account, there are many related charges which could be simplified if combined into one account charge. In addition, clear communication of these charges and the benefits associated with interest rates could also be beneficial. The complexity of documentation and products have been noted in the literature, albeit as a supply-side factor (Abel et al. 2018; Karpowicz 2016). Our results suggest that care needs to be taken when designing post-adoption information to make it usable by customers who may have low literacy levels. In addition, these results also speak to the importance of financial literacy as a complement to service provision in attempts to increase financial inclusion beyond access.

The lack of knowledge about how financial services work could also explain the high levels of mistrust in banking that the participants expressed. The literature documents that a lack of trust sometimes drives low-income users to informal services due to the personal and community trust a (Grohmann et al. 2018) shows that financial literacy increases the uptake and usage of financial services. References to the existence of a ‘mouse’ in the bank indicated the participants’ belief that banks were not honest in dealing with them. This lack of trust resulted in reduced use of financial products and services.

Employment and income are cited in the literature as factors that influence the selection of clients by banks (Kombo 2021). Being employed can influence the use of financial services in two ways. First, in most cases, formal employment requires the ownership of a bank account to receive wages. Second, bank selection criteria favor individuals with formal employment and income stability. Rural communities are predominantly self-employed and in the agricultural sector.

For this reason, they are very likely to be excluded. However, the Mzansi account mentioned above was designed to circumvent such requirements. Nevertheless, there was significant evidence that several participants had self-excluded from opening an account for fear of rejection on account of being unemployed.

In addition, several participants self-excluded based on low and irregular incomes. Most rural communities have an irregular income from agriculture and microbusinesses. Those who have a regular income rely on very low government grants. The literature shows that the likelihood of being accepted by a bank increases with income (Wang and Guan 2016; Ramji 2009). Both income size and flows provide important information about the client’s ability to maintain their account and minimize risk exposure for the bank. Participants indicated they could not save in the bank because of their low incomes. However, almost all indicated that they saved with rotating savings and credit associations (ROSCAs) and other self-help groups. The prevalence of ROSCAs as a savings mechanism means that rural savers lose out on interest. However, the amounts they save are pretty low. Given the charges attached to account maintenance, the net effect could be harmful, deterring them from saving in the bank. The primary source of credit for the participants was money lenders. The money lenders charge very high interest rates. However, there was a sense that the money lenders understood the customers’ needs better than the banks. These informal services exploit social capital and provide very flexible lending terms. The literature shows that flexible lending terms and social capital are important in circumventing key market failures in financial markets (Simatele and Dlamini 2020).

7. Conclusions

In conclusion, our findings show that supply-side factors such as distance from the point of provision and high transaction costs are important in encouraging the use of financial services in rural areas. Distance from the point of service provision increases the transaction costs associated with bank accounts and discourages their use. In addition, relatively high and complex account fees deter customers from using their accounts even after they have opened them. The complexity of fees and associated documents has engendered distrust towards banks, further stifling the use of financial services.

However, we found that demand-side factors, including the lack of formal employment, low and irregular incomes, financial illiteracy, and risk and trust perceptions, play a more significant role. The introduction of the Mzansi account, a no-frills account targeting users such as those in rural areas, significantly increased the levels of financial inclusion in South Africa from an access perspective. Nevertheless, levels of usage are very low. Focusing on supply-side and access factors ignores the fact that financial inclusion benefits can only be fully realized if excluded populations have access to and use the services too. We found that various services are available but not being used for various reasons. For instance, regulating the production of accessible and user-friendly documentation which explains services and products can ease some of the discomforts that rural users have around finance, creating a more trusting environment. In addition, simplified structures which amalgamate account fees into a standard fee could be more appropriate for rural populations who typically have very low levels of education. Post-adoption financial education should accompany service and product provision if financial inclusion efforts are to bear much fruit.

Technology offers an important alternative for rural populations due to their geographical exclusion. However, the South African national payment system regulations present a barrier to the development of mobile banking, which has revolutionized access to financial services in many developing countries. Easing these regulations and investing in better telecommunications infrastructure in rural areas would go a long way to increasing the use of financial services. In addition, supermarket money markets are emerging as potentially important partners in the drive to increase financial inclusion. The transactions accounts offered by supermarkets offer accounts that have much lower and simpler fee structures. They could provide a more efficient doorstep banking model to increase financial inclusion for rural populations.

This study points to some important factors that could aid financial inclusion efforts in rural South Africa. However, the application of the results is limited as the study used focus group data in two provinces. While valuable insights have been gathered, further research is required to investigate how prevalent these factors are across the country and to what extent they impede or encourage the use of financial services in rural areas. A quantitative study with a larger dataset would complement the results of this study.