Digital Finance in Africa: What Accounts Reveal About Borrowing to Start a Business

Research across 34 African economies explores how identity ownership and digitally enabled accounts relate to borrowing for business purposes.

Getting money to start a business can involve much more than having an identity document or opening an account. A person may be recognised by financial institutions and still struggle to find affordable borrowing that suits their plans. Research across 34 African economies explores how identity ownership and digitally enabled accounts relate to borrowing for business purposes.

The study, "Entrepreneurship and Digital Financial Inclusion in Africa: Foundational Identity, Digitally Enabled Accounts and Business-Purpose Borrowing," appears in the International Journal of Financial Studies. Researchers Vivence Kalitanyi and Misheck Musaigwa analysed survey responses from 29,883 adults aged 18–64 using data collected in 2024 for the World Bank's Global Findex. The finding was that digitally enabled accounts had a clearer connection with reported business borrowing than identity ownership alone.

That finding speaks to a practical question about financial inclusion: what does being connected to the financial system actually tell us about someone's financing activity? Having identification, holding an account and borrowing to run a business describe different parts of a person's financial life. The research examines those connections without claiming that one automatically leads to the next.

What the Researchers Measured

Participants were asked whether they had borrowed money during the previous 12 months to start or operate a business. Their answers did not reveal whether the money came from a bank, relatives, informal lenders or another source. The survey measure could not establish loan approval, borrowing costs, current business ownership or whether the money helped a business succeed.

Foundational identity meant reported ownership of a recognised identity document, with documentation varying across countries. A digitally enabled account covered mobile money accounts or financial institution accounts used for specified digital payments by card or phone. This combined measure captured different forms of financial connection without showing transaction frequency, account-opening dates or whether borrowed money passed through an account.

Each economy received equal overall weight, giving smaller economies the same influence as larger ones in the main analysis. The researchers accounted for measured personal characteristics and country differences, with additional checks using other samples and weighting methods. Income was excluded from the main model because Lesotho lacked that information, and income-adjusted checks covered 29,037 respondents across the remaining 33 economies.

Digital Accounts Had the Clearest Link to Business Borrowing

In the equally weighted sample, 83.17% reported owning foundational identification, 48.99% held a digitally enabled account, and 16.33% reported business-purpose borrowing. Borrowing rates varied widely, ranging from 3.29% in Morocco to 32.26% in Uganda. These figures describe the included economies under the study's weighting approach, rather than a single borrowing rate for Africa's entire adult population.

Holding a digitally enabled account was associated with a 6.10-percentage-point higher predicted probability of reporting business borrowing after adjustment for other measured factors. The reported uncertainty range ran from 4.62 to 7.59 percentage points, supporting a positive association within the model. This difference does not mean opening an account would increase an individual's borrowing probability by that amount.

Identity ownership had a smaller and less certain connection with borrowing, estimated at 1.33 percentage points before account status entered the model. That estimate fell to 0.64 percentage points after account status was considered, with an uncertainty range compatible with either a small negative or positive relationship. Identity ownership was associated with a 10.43-percentage-point higher predicted probability of holding a digitally enabled account, showing that identification and account access were related but distinct.

The positive account-borrowing connection remained across income-adjusted analyses, alternative weighting methods and different age or workforce samples. Removing each economy in turn produced estimates between 5.81 and 6.32 percentage points, suggesting that no single economy drove the pooled finding. Separate checks of financial institution accounts, mobile money accounts and digital payment use found positive connections, without establishing that these services offered equivalent experiences.

What the Findings Cannot Tell Us

The research used a snapshot of people's circumstances, leaving the order of identity ownership, account ownership and borrowing unresolved. Someone seeking business finance might obtain an account during that process, and factors such as financial knowledge, entrepreneurial experience or local lending conditions could influence both. The results cannot prove that accounts caused borrowing or that accounts explain a causal pathway from identification to business finance.

Exploratory comparisons found no reliable evidence that the associations differed by sex, rural or urban residence, or income group. Limited precision means those findings cannot establish that the relationships are identical across groups. Similar borrowing rates could conceal very different interest charges, loan sizes, repayment pressures or treatment by lenders.

The 34 economies do not represent every African country, and differences in identity documentation and account measures limit direct comparisons. Online digital identity information was available only for Benin, Morocco and Uganda, so it was excluded from the main analysis. Self-reported answers and unmeasured personal circumstances leave gaps that the researchers' checks could not fully resolve.

Financial Inclusion Needs More Than Ownership Numbers

An identity document can help someone meet verification requirements without demonstrating business cash flow, reducing lending risk or creating a suitable credit product. An account may place that person closer to everyday payments and financial services, making it a more informative marker of participation in this study. The uncertain identity result does not show that identification is unimportant; it shows that identity coverage alone tells us relatively little about reported business borrowing.

Researchers discuss affordable digital services, accessible payment systems and proportionate customer checks as considerations for broader financial-inclusion efforts. Data protection, cybersecurity and accessible complaint processes belong in that discussion, though the study did not directly test their effects. Evaluating these proposals requires evidence about actual account use, available credit, borrowing costs and repayment outcomes.