2.1. Methodology and Data Selection

The analysis primarily relies on UNCDF’s bank pricing data that were collected based on a market research approach by Africa Analysis,⁶ covering 197 commercial banks in 34 countries. The countries were selected based on UNCDF MAP’s country footprint in Africa and Asia (Table 1). Historically, the MAP program has operated in 20 countries, covering three regions—SADC, WAEMU, and ASEAN.⁷ For comparability, the bank pricing data collection was expanded to include all countries in these three regions. In each of the countries included in the database, data were collected for four to six retail commercial banks (see Annex II for details), depending on the size of the domestic financial sector,⁸ with the aim to cover more than 70 percent of the share in each country.

^{Table 1:}

List of countries in the sample

^{Table 1:}

List of countries in the sample

Region	SADC	WAEMU	ASEAN
Countries	Angola	Benin	Brunei Darussalam
	Botswana	Burkina Faso	Cambodia
	Comoros	Côte D’Ivoire	Indonesia
	Democratic Republic of the Congo (DRC)	Guinea-Bissau	Laos
	Eswatini	Mali	Malaysia
	Lesotho	Niger	Myanmar
	Madagascar	Senegal	The Philippines
	Malawi	Togo	Singapore
	Mauritius		Thailand
	Mozambique		Vietnam
	Namibia
	Seychelles
	South Africa
	Tanzania
	Zambia
	Zimbabwe

View Table

^{Table 1:}

List of countries in the sample

Region	SADC	WAEMU	ASEAN
Countries	Angola	Benin	Brunei Darussalam
	Botswana	Burkina Faso	Cambodia
	Comoros	Côte D’Ivoire	Indonesia
	Democratic Republic of the Congo (DRC)	Guinea-Bissau	Laos
	Eswatini	Mali	Malaysia
	Lesotho	Niger	Myanmar
	Madagascar	Senegal	The Philippines
	Malawi	Togo	Singapore
	Mauritius		Thailand
	Mozambique		Vietnam
	Namibia
	Seychelles
	South Africa
	Tanzania
	Zambia
	Zimbabwe

The data collection process involved a combination of secondary and primary research methods First, secondary research was conducted that relied on accessing publicly available information on pricing from various sources such as banks’ websites, publications by banks, regulators, and third-party sources. To the extent possible, validation of the bank product and service prices was undertaken by sourcing fees filed by the various banks with the financial sector regulatory authorities. In instances where bank product and service fees were obtained from the bank’s web site, the bank was contacted by e-mail to verify the fees published on the bank’s web site. This was done particularly in situations where it was unclear which specific year the bank fees were applicable to.

The secondary research served as the foundation for identifying any data gaps, which prompted the need for primary research. To ensure comprehensive data coverage, primary research involved direct contact with banks through methods like phone calls, emails, or online conferencing platforms. The collected data were systematically recorded into a structured template.

Sample selection of banks was done primarily based on a market share analysis, combined with local expertise in case data availability was an issue, with the objective of including the largest banks that represent the majority of the domestic market share in each country (i.e., at least 70 percent of the market share in terms of total assets).⁹ An additional criterion applied in selecting banks was the relevance of products and services that the banks provide; banks that offer products aimed at low- and medium-income groups were prioritized over those that focus on the high-income group, hence the selected banks may have lower market share than the targeted majority. The dataset, along with a detailed description of the methodology and the names of the selected banks, is publicly accessible through the UNCDF website.¹⁰

Product selection for the data collection included consideration of services that mainly target low-income individuals and micro to small businesses, based on the product features and pricing (i.e., typical products used by low-income consumers and micro and small business customers), or by being explicitly stated, for instance, in marketing materials and branding. No investment products, such as investment in stock exchange shares (equity) or unit trusts were included in the product suite. The product suite was informed by typical products offered to these customer groups by the largest banks in South Africa, with an additional plausibility check based on banks in Malaysia.¹¹ Definitions for the various bank products were developed to ensure that data are gathered for comparable products across all banks across all the countries researched (see Annex IV for a detailed description of definitions of products and services).

Once the data on bank fees and charges were collected for each individual bank, median, average, minimum, and maximum costs for each product and service were calculated at the country level. The data used in this paper are the median country value for each product fee line (also referred to as variables or indicators). Data were collected in terms of local currency denominations, and converted to USD,¹² and include the applicable sales tax / VAT for each country.

While pricing data are collected for a large number of indicators, only a subset is considered in this paper. This selection primarily stems from considerations of both their significance in the context of financial inclusion and the uneven data coverage within the dataset. For instance, among the total of 105 pricing variables available in the dataset for the low-income group, only 50 possess data for at least half of the countries, and only 20 variables encompass data for at least three-quarters of all countries. The coverage issue is even more pronounced for small businesses related indicators, with only 30 out of the 97 variables having data available for at least half of the countries.¹³ In light of these limitations, this paper primarily centers its analysis on essential financial metrics with better data coverage such as the monthly checking account fee, branch and ATM withdrawal fee, savings account fee, annual card fees, and mobile money account fees for low-income group individuals and SMEs. Moreover, this paper mainly focuses on the 2022 data as these are the most comprehensive, having the highest average number of indicators with available data per country. Summary statistics for the variables used in this study can be found in Table 2 below.

^{Table 2:}

Descriptive Statistics

Note: N denotes the number of countries with data for each variable. Mean, Minimum, and Maximum statistics denote dollar amounts.

^{Table 2:}

Descriptive Statistics

Statistic	N	Mean	St.Dev	Min.	Max.
Low Income Group
Checking Account Monthly Account Fee	31	1.3	1.8	0	8.3
In-branch Cash Withdrawal Charges from Checking Account per 100 USD	30	2	2.8	0	9.1
Mobile Money Monthly Account Fee	12	0.3	0.9	0	3
Send money to Mobile Money account (within own bank)	18	0.5	0.8	0	3
Savings Account Opening Fee	24	9.8	20.8	0	72.7
Savings Account Monthly Maintenance Fee	26	0.4	1.3	0	6.5
Withdrawal (own bank ATM)/ per 100 USD	32	0.3	0.6	0	3
Annual Card Fees	29	6.8	5.5	0	17.9
SME
General Business Account Monthly Maintenance Fee	30	8.7	10.9	0	55
Cash Withdrawal (in-Branch)/ per 100 USD	26	1.7	2.3	0	8.7
Inter-account transfer at a branch (within own bank)	27	1.3	2	0	7.4

Note: N denotes the number of countries with data for each variable. Mean, Minimum, and Maximum statistics denote dollar amounts.

View Table

^{Table 2:}

Descriptive Statistics

Statistic	N	Mean	St.Dev	Min.	Max.
Low Income Group
Checking Account Monthly Account Fee	31	1.3	1.8	0	8.3
In-branch Cash Withdrawal Charges from Checking Account per 100 USD	30	2	2.8	0	9.1
Mobile Money Monthly Account Fee	12	0.3	0.9	0	3
Send money to Mobile Money account (within own bank)	18	0.5	0.8	0	3
Savings Account Opening Fee	24	9.8	20.8	0	72.7
Savings Account Monthly Maintenance Fee	26	0.4	1.3	0	6.5
Withdrawal (own bank ATM)/ per 100 USD	32	0.3	0.6	0	3
Annual Card Fees	29	6.8	5.5	0	17.9
SME
General Business Account Monthly Maintenance Fee	30	8.7	10.9	0	55
Cash Withdrawal (in-Branch)/ per 100 USD	26	1.7	2.3	0	8.7
Inter-account transfer at a branch (within own bank)	27	1.3	2	0	7.4

Note: N denotes the number of countries with data for each variable. Mean, Minimum, and Maximum statistics denote dollar amounts.

2.2. Data Limitations

While the UNCDF’s bank pricing database is a unique and valuable resource that attempts to provide a better understanding of financial inclusion from an affordability perspective, it does have some limitations which are important to keep in mind.

First, although extensive efforts have been made to ensure the best coverage of commercial banks possible, the database may not capture majority of the market share in all cases.¹⁴ This issue arises primarily due to data availability gaps, particularly in regions with limited financial infrastructure or where data collection mechanisms are not well established. Despite attempts to mitigate this limitation by consulting field experts and utilizing alternative data sources, the coverage bias may still restrict the database’s ability to provide a comprehensive view of affordability in certain contexts.

Second, the database predominantly focuses on commercial banks, potentially overlooking other significant financial service providers such as state-owned banks, microfinance institutions, cooperatives, or non-bank mobile money operators. By primarily centering on commercial banks, the database may omit important pricing information and insights from these alternative financial institutions, which can be vital for a holistic understanding of financial inclusion with regards to low-income households in some countries.

Lastly, although the database provides detailed pricing information, it may not capture the complete range of product features and conditions. Factors like interest rates, loan terms, collateral requirements, or eligibility criteria, which profoundly impact the accessibility and affordability of financial services such as credit, are not adequately represented in the current database and consequently not discussed in this paper. Similarly, remittances and cross-border payments are crucial aspects from the perspective of financial inclusion. However, this paper does not delve into their discussion.

Recognizing these limitations can help contextualize the findings of this paper and encourage further exploration of other dimensions impacting financial inclusion beyond pricing information, prompting a nuanced understanding of the specialized focus required to collect more granular data.

3.1. Cost of Banking Services for Low-Income Households

The cost of banking services can be a significant burden for low-income households, given their limited financial resources. These households often struggle to access and afford basic banking services such as maintaining a checking or savings account, making electronic transfers, or using services such as ATMs and bank cards. This section takes stock of different types of fees and charges and provides a comparison of how they vary across countries in WAEMU, SADC, and ASEAN. We present the nominal costs (in USD) as well as the normalized costs relative to monthly GNI per capita (as a proxy for monthly income) to have a better understanding of affordability across countries.

Nominal costs of financial services

The data reveal that commercial banks in several countries impose a high monthly fee on maintaining a basic checking account. This fee can be in excess of USD 5 per month, as is the case in Niger and Cote D’Ivoire, and can also exceed 5 percent of monthly GNI per capita, as is the case in Guinea-Bissau, Niger, DRC, and Madagascar. On average, the monthly cost of maintaining a checking account is the highest in WAEMU countries (3.5 percent of monthly GNI per capita, on average), which corroborates with lower bank account ownership in the region as shown above, whereas it is the lowest in the case of countries in ASEAN, with countries such as Malaysia and Myanmar not charging any monthly fees on personal checking accounts (Figure 2). This high monthly maintenance fees can be a deterrent to opening bank accounts for low-income consumers, particularly in countries where this cost represents a relatively large proportion of their monthly income.

View Full Size

Checking Account Monthly Fee

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: UN Capital Development Fund (UNCDF); Authors’ calculations.Note: Fees represent median values, inclusive of VAT. Regional and overall averages are weighted using total population. Overall fee average is the average value of checking account monthly fees in USD for all countries in the sample.

• Download Figure
• Download figure as PowerPoint slide

Similarly, using bank debit cards, conducting in-branch or ATM transactions, and sending money via mobile money accounts can come with high fees in several countries, particularly in WAEMU and SADC regions. For example, in the case of Mozambique and Madagascar, annual median card fees can be more than $15, which can be a potential barrier for ATM usage. Although opening a mobile money account (operated by banks) can be free of charge in most countries, the transaction fees vary, albeit remaining low compared to traditional bank accounts, as highlighted in the right bottom chart in Figure 3. Lastly, data on the costs associated with in-branch cash withdrawals and ATM transactions also vary significantly across countries, with countries in SADC charging more on average than other countries (left bottom chart in Figure 3).

View Full Size

Costs of Banking Products and Services

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: UN Capital Development Fund (UNCDF); Authors’ calculations.Note: Fees represent median values, inclusive of VAT. Regional and overall averages are weighted using total population.

• Download Figure
• Download figure as PowerPoint slide

Affordability Analysis

To have a better understanding of bank pricing across countries, we construct two indices of affordability— basic affordability index and advanced affordability index—and use them to assess the affordability of financial services for low-income households as discussed in the rest of this subsection.

First, to capture the most basic forms of affordability for individuals, an affordability index is constructed that captures the monthly fees of maintaining a bank checking account along with the cost of doing one ATM withdrawal per month.¹⁵ The indicator is then again normalized by average monthly GNI per capita, to control for differences in income levels across countries. The composition of this basic affordability index corresponds with UNCDF’s MAP program findings (UNCDF, 2016) that most low-income customers use their bank accounts as a mailbox, whereby they receive their income in a bank account and then immediately withdraw all the money and spend it in cash.

The results of the basic affordability index show that even the most basic bank product package varies by country in terms of affordability, and the costs can reach in excess of 5 percent of monthly average income, as is the case for four countries in the sample (blue bars in Figure 4). These results suggest that low-income households in WAEMU pay on average higher costs (for the basic affordability product package) as compared to countries in SADC, while households in ASEAN countries pay on average the lowest costs (for the same product package).

While the basic affordability index is instrumental in assessing the initial accessibility of financial services, it may not adequately encompass the broader financial requirements and well-being of individuals. In light of this, we introduce an advanced affordability index that factors in a more intricate usage pattern of financial products and services. The selection of the additional usage criteria was primarily informed by the findings of the demand-side FinScope¹⁶ survey conducted in selected countries within our sample.¹⁷

The advanced affordability index assesses four distinct dimensions: (i) monthly cost of maintaining a checking account; (ii) monthly cost of having a bank debit card; (iii) two ATM transactions from own bank equivalent to $100 each; and (iv) two in-branch cash withdrawals equivalent to $100 each. By adding these dimensions to the analysis offers an opportunity to better explore the affordability of financial services, extending beyond mere account usage, and thereby providing a more nuanced perspective to allow for further context of what will be required to enhance participation of low-income households.

The advanced affordability index (yellow bars in Figure 4) shows that there is a lot of heterogeneity across countries, with countries in SADC having higher costs on average than countries in WAEMU, which is a switch from before highlighting countries in SADC pay more for services such as in-branch cash withdrawals and ATM transactions. In the case of a few countries in ASEAN, the advanced index corresponds to zero, suggesting that even the advanced usage package of financial services has minimal cost.¹⁸ Overall, for about one-third of the countries in our sample, costs of advanced usage can be in excess of 5 percent of monthly GNI per capita.

View Full Size

Affordability Index

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: UN Capital Development Fund (UNCDF); Authors’ calculations.Note: Fees represent median values, inclusive of VAT. Regional and overall averages are weighted using total population.

• Download Figure
• Download figure as PowerPoint slide

Barriers to financial access: Possible impacts and drivers

These financial barriers associated with affordability can further exacerbate financial exclusion, calling for policymakers and financial institutions to find innovative and affordable solutions to address cost issues. Indeed, the affordability index is found to be negatively associated with financial inclusion (Figure 5), highlighting that higher costs associated with banking services can hinder financial inclusion.¹⁹ Our econometric estimations also show that higher costs of financial services are negatively associated with financial inclusion, even after controlling for a broad range of variables such as income level, institutional quality, labor market dynamics, human development, legal rights, and digital infrastructure (see Annex III for details). However, given the limited number of country-level observations and cross-sectional nature of the data, we are not able to draw causal inferences at this juncture, and our analysis is primarily focused on identifying correlations.

Another crucial aspect to explore is the factors influencing the variability in the affordability of financial services across different countries. We examine correlations between the cost of financial services and pertinent macroeconomic indicators to gain insights into these relationships. The results indicate that higher costs of basic financial services are associated with factors such as regulatory quality, competition in the banking sector, and the level of digitalization within a country, among other factors. These findings suggest that these elements potentially play significant roles in shaping the financial access and affordability landscape.

View Full Size

Higher costs of financial services are negatively associated with financial inclusion.

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

• Download Figure
• Download figure as PowerPoint slide

Results highlight that a higher level of regulatory quality²⁰ in a country is associated with lower costs of accessing basic financial services (Figure 6). Similarly, we find that an increase in digital connectivity in a country, measured by the enhanced digital access indicator (EDAI),²¹ is correlated with lower costs of financial products and services (Figure 6). This highlights the importance of promoting affordability and focusing on enhancing access to new technology-based solutions such as digitalized products and channels to lower operational costs and promote increased access to financial services. Lastly, we also demonstrate that a higher number of banks and bank branches within a country, which serves as a rough proxy for competition in the banking sector, is associated with lower banking fees (Figure 6). This suggests that increased competition among banks may lead to more favorable pricing for consumers, enhancing the overall affordability of financial services.

Due to the limited availability of bank pricing data and relevant control variables, this paper does not perform any econometric analysis to further understand these drivers. This gap underscores a vital area for future research. Delving deeper into these correlations with more detailed and robust datasets could provide valuable insights into developing strategies to enhance the affordability and accessibility of financial services globally.

At more micro levels, detailed analyses of the affordability and costs of financial services across 19 MAP program countries and cross-country consumer behavior insights are available through detailed country diagnostics and knowledge series publication.²² The findings from the first six MAP pilot countries indicate that the pricing models for bank accounts serve as a disincentive for the majority of adults. While these models vary significantly, qualitative research in these countries shows strong opposition to monthly service fees because they erode the value stored in accounts. This resistance points to a broader issue within the financial services ecosystem that fails to cater effectively to low-income individuals. Moreover, the necessity to create and strengthen a market-based institutional environment becomes evident, as exemplified by findings from Nepal, where bank savings products lack diversity and do not offer tailored solutions for low-value savings, highlighting a key area for improvement in market affordability strategies (UNCDF 2021; UNCDF 2020).

View Full Size

Higher costs of financial services are negatively associated with regulatory quality, digital access, and competition in the banking sector.

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: IMF’s Financial Access Survey (FAS); UN Capital Development Fund (UNCDF); World Bank; ITU; Authors’ calculations.Note: A higher number for basic affordability index (percentage of monthly income) suggests that the costs associated with availing basic financial services are high.

• Download Figure
• Download figure as PowerPoint slide

3.2. Cost of Banking Services for SMEs

The data reveal that the cost of banking services for SMEs is also quite significant, with substantial variations across countries and regions. The overall average monthly maintenance fee for a general business account stands at USD 6.1 (equivalent to an average of 2.1 percent of monthly GNI per capita), with the highest fee observed in the WAEMU region at an average of USD 14.3 (15.3 percent of GNI per capita), while the lowest average fee is found in the ASEAN region (USD 1.4, or 0.4 percent of GNI per capita). At the country level, differences become even more pronounced. For instance, general business accounts for SMEs in Brunei Darussalam incur a higher monthly maintenance fee,²³ whereas the service is provided free of charge in the case of Philippines and Mozambique (Figure 7).

View Full Size

SME Monthly Account Fee

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: UN Capital Development Fund (UNCDF); Authors’ calculations.Note: Fees represent median values, inclusive of VAT. Regional and overall averages are weighted using.

• Download Figure
• Download figure as PowerPoint slide

Cash withdrawal and transfer fees are also not negligible for general business accounts, although they do not necessarily follow the same pattern as monthly maintenance fees across countries and regions (Figure 8). On average, in-branch cash withdrawal from general accounts costs about USD 4.5 per USD 100 transaction. Interestingly, all WAEMU countries with available data offer free withdrawal services, while SMEs in SADC and ASEAN pay an average of USD 2.5 and USD 8, respectively, for a withdrawal of USD 100 from their general account. Similarly, inter-account transfers (within the same bank) at bank branches are the highest for SADC with a regional average of USD 2.4, followed by ASEAN (USD 0.5). Notably, this service is also free of charge for all WAEMU countries.

View Full Size

Costs of Banking Services for SMEs

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: UN Capital Development Fund (UNCDF); Authors’ calculations.Note: Fees represent median values, inclusive of VAT. Regional and overall averages are weighted using total population.

• Download Figure
• Download figure as PowerPoint slide

4.1. Overview of Access to Banking Services in Botswana

While the level of financial access in Botswana is higher than for many of its peers in the SADC region—in part reflecting the efforts of the government to improve financial inclusion²⁵ and the relatively lower cost of banking services than its regional peers—a significant proportion of the population remains unbanked in the country (Figure 9). According to the latest FinScope data, as of 2020, about 44 percent of adult population in the country was without a formal bank account²⁶ compared to 18 percent in South Africa and 32 percent in Namibia. Similarly, supply-side data from the Financial Access Survey (FAS) show that access to banking (measured as the number of depositors with commercial banks per 1,000 adults) is yet to reach the entire adult population in 2022. This underscores the ongoing challenges and the need for further interventions aimed at improving access to banking and financial inclusion in Botswana.

View Full Size

Access to Bank Accounts in Botswana vis-a-vis Regional Comparators

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Sources: Financial Access Survey (FAS); FinScope Data; Authors’ calculations.Note: FinScope data for each country represents the latest available information on percentage of banked population between 2017 to 2022.

• Download Figure
• Download figure as PowerPoint slide

To some degree, this is driven by prevailing incidence of poverty and low-income levels, making it crucial to closely look at the affordability landscape to better understand costs associated with banking services as a barrier to financial inclusion. As of 2021, about 21 percent of the population lives below the national multidimensional poverty line. ²⁷ According to the FinScope data, a large proportion of these low-income households seem to have lower access to bank accounts as compared to households with relatively higher incomes. For instance, the proportion of people with bank accounts is only 25 percent for the lowest income group while it reaches 94 percent for the highest income group in Botswana (top left chart in Figure 10). This implies that affordability can play an important role in the uptake of bank accounts; it is therefore crucial to analyze affordability for different income levels to better understand barriers for financial inclusion.²⁸

The usage of bank accounts is also strongly associated with other socioeconomic characteristics including age, sex, and location as shown in Figure 10. The likelihood of having a bank account is the lowest for those above 60 years (34 percent) followed by the youngest age group while access to a bank account is the highest for those in the middle age groups. In addition, the proportion of people with a bank account increases with urbanization levels while males are also more likely to have a bank account.

View Full Size

Access to bank accounts across different socioeconomic characteristics (2020)

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Source: FinScope Botswana 2020; Authors’ calculations.²⁹

• Download Figure
• Download figure as PowerPoint slide

4.2. Bank Pricing in Botswana

The cost analysis presented in Section 3 shows that fees charged by banks in Botswana for both the low-income group and small businesses are in general lower compared to fees in most SADC countries.³⁰ However, in Botswana, like most SADC countries, commercial banks charge higher fees for maintaining general business accounts compared to checking accounts for low-income individuals. A few additional takeaways from an analysis of bank fees in Botswana and its regional comparators include the following:

• The fees for all the services covered in the analysis are lower than the average fees for SADC, the differences (from the regional average) are particularly significant for checking account monthly maintenance fees, annual card fees, and in-branch cash withdrawals. On the other hand, the gap for mobile money account transfers (within the same bank) is relatively modest.

• In contrast to countries like Comoros, Madagascar, Mozambique, and Seychelles, where in-branch cash withdrawals are free, customers in Botswana typically incur a fee for this service.

• Finally, similar to most SADC countries, in-branch withdrawals cost substantially more than ATM withdrawals. Sending money to a mobile account (within the same bank) is only marginally higher than an ATM cash withdrawal. Payments to an account and inter-account transfers in general are much cheaper at an ATM or through online banking, compared to in-branch.

4.3. A Closer Look at Affordability in Botswana

With the above background in mind, this section presents affordability analysis for different income groups of the banked population. We use the bank pricing data discussed in Section 3 and the FinScope Botswana 2020 Consumer Survey data—a complementary demand-side survey that includes detailed information on usage of banking services by income group. This analysis expands on the affordability indices developed in Section 3 by exploring the actual reported product usage across different income groups.

By dividing the adult population into roughly four equal income quartiles using data from FinScope,³¹ we highlight the differences in the usage and affordability of financial products and services across different income groups in Botswana (left-hand side chart in Figure 11).³² Given their banked status, all income groups own a checking account. Most of the population also seems to withdraw money from their checking account at least once a month, but the frequency of ATM withdrawals declines with income levels with the lowest income group not making any withdrawal at an ATM. Services such as ATM transactions, savings accounts, and internet banking³³ are used mainly by the two higher income groups. Cellphone banking, although significant—is used by less than 40 percent of adults even in the highest income group.

Based on the usage pattern and frequency of using bank services (as reported in FinScope), we identified a more suitable bundle of (basic) services (as compared to the affordability indices in Section 3) for each income group in Botswana as follows:

• ▪ The low-income group (1^st income quartile): Maintaining a checking account and one in-branch withdrawal per month.

• ▪ The 2^nd income quartile: Maintaining a checking account, one ATM withdrawal, and one cellphone banking transaction per month.

• ▪ The 3^rd income quartile: Maintaining a checking account, one ATM withdrawal, one cellphone banking transaction, and a savings account maintenance fee per month.

• ▪ The highest income group (4^th income quartile): Maintaining a checking account, three ATM withdrawals, three cellphone banking transactions, and a savings account maintenance fee per month.

We then estimated the cost of the common bundle of services for each income group (normalized by median/mean income of each group) according to the product suite and prices discussed in the previous section. Despite there being an increase in both product usage and the number of products used in higher income quartiles, costs as a percentage of their income tend to decline significantly as we move to higher income quartiles (right-hand side chart in Figure 11). We find that the lowest income group (over a quarter of income earning adults) has to pay about 5 percent of their monthly income, on average, to just have a checking bank account and make one withdrawal (in-branch) per month. The higher income groups spend a significantly lower percentage of their income, though they use a higher number of services, and use these services more frequently. Our results show that, overall, about 36 percent of income-earning adults in Botswana would need to pay more than two percent³⁴ of their monthly income to access a basic bank account and conduct one in-branch withdrawal.

View Full Size

Bank product monthly usage packages and affordability per income group

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Source: FinScope Botswana 2020³⁵; Authors’ calculations.

• Download Figure
• Download figure as PowerPoint slide

4.4. Affordability for Micro and Small Businesses

A similar analysis is conducted for micro and small businesses in Botswana. Using FinScope 2020, we derive two groups—micro businesses with no hired employees (69 percent of the sample), and micro and small businesses who have employees (31 percent of the sample). MSMEs tend to use formal financial services more than households, and usage is therefore similar to the usage of personal accounts by higher income groups, where the highest usage reported is for ATMs, savings accounts, and cellphone banking, with significant usage also of current accounts, internet banking, and mobile banking products.

Figure 12 below portrays differences in the usage of financial products and services across micro and small businesses with and without employees, based on the reported usage and frequency of products in FinScope for each group (similar to the above methodology for personal accounts). Both groups that have accounts would pay a monthly maintenance fee for a business account. In addition, there is likely to be between 3 to 4 withdrawals at an ATM per month; this is based on the detailed data on frequency of ATM usage where more than 50 percent of each group who do use ATM’s report using it only on a monthly basis, but between 33 and 41 percent of the two groups report using ATMs more frequently (weekly or to a lesser degree daily). A similar approach yields on average 4 instances of cellphone banking per month for the first group, and 8 for the second group. For savings accounts, we identify one interaction per month for the first group, but two for the second group.

Using the same average personal income from the same FinScope survey as a proxy for business earnings,³⁶ the affordability of the resultant product packages for micro and small businesses are depicted in the right-hand side chart in Figure 12. The results show that the proportion of monthly income spent by micro businesses with no employees on banking costs is higher (3.3 percent of their monthly income on average) than the one spent by micro or small businesses with employees (2.3 percent of their monthly income).

View Full Size

Micro and Small business bank products usage and affordability, Botswana (2020)

Citation: IMF Working Papers 2024, 150; 10.5089/9798400280627.001.A001

Source: FinScope Botswana 2020; Authors’ calculations.

• Download Figure
• Download figure as PowerPoint slide

In sum, the above analysis shows that while the cost of banking services in Botswana is comparable to or even lower than those in SADC countries, it still poses a significant financial burden, particularly for low-income households. The cost of basic banking services (measured as a percentage of monthly income) is high for the low-income group—on average, more than a quarter of the bottom income earning adults spend about 5 percent of their monthly income to access basic services (to open a checking account and make one in-branch withdrawal per month). Similarly, micro-businesses (with no employees) incur relatively higher fees for the usage of banking products and services compared to micro and small businesses with employees, underscoring a greater financial burden faced by these micro-businesses. These findings seem to explain, at least partly, why a significant fraction of the population still does not have access to banking services.

4.5. Considerations for policymakers and regulators

Botswana has a National Financial Inclusion Strategy (NFIS) and roadmap spanning from 2015 to 2021 and is currently in the process of updating the NFIS for a new period.³⁷ The existing roadmap has six priority areas, two of which are relevant when looking at bank pricing: Priority 1 is about the payment ecosystem development, and Priority 2 is about development of low-cost accessible saving products. For the latter, the analysis above showcases how banks can contribute, including by having zero monthly fees for savings accounts as well as a zero rate for the first withdrawal. For Priority 1, the bank pricing analysis can provide further insight and potential policy directions as follows:

• ▪ ATMs: The roadmap calls for partnerships to develop infrastructure for distribution (which includes ATMs), but the bank pricing analysis suggests that the pricing aspect also plays an important role. The total cost of accessing an ATM—including annual card fees—need to be considered in order to catalyze the usage of ATM infrastructure. In this regard, the roadmap does call for revisiting bank charges for sustainable rural provision to reduce overall bank access costs, and the ATM related costs would be key to achieving this (in addition to utilizing digital channels and partnering to leverage agent networks).

• ▪ Interoperability: Charges for payments or transfers to other banks, as well as transactions that are conducted through ATMs of other banks are still much more expensive than those conducted within the same bank (or to accounts within the same bank). Relating to this, the roadmap does call for improved interoperability, not only between banks but also between banks and MNOs—greater interoperability could contribute to further lowering the cost of financial services.

• ▪ Bank card fees: While the bank pricing database does not encompass the fees merchants incur for transactions through points of sale (POS), the NFIS suggests a strategy for banks to collaborate with retailers. This partnership aims to enhance the card’s value proposition for merchants, thereby boosting financial access for their customers. Concurrently, our analysis reveals that the annual fees charged to consumers for card use—excluding the merchant fees for POS transactions—could discourage the use of ATMs. This, in turn, poses a challenge to broadening financial inclusion, particularly in rural areas where such barriers can significantly impede access to banking services.