Entrepreneurship and Digital Financial Inclusion in Africa: Foundational Identity, Digitally Enabled Accounts and Business-Purpose Borrowing

This study examines whether reported foundational identity ownership and digitally enabled account ownership are associated with business-purpose borrowing in Africa. Drawing on 2024 Global Findex microdata for 29,883 adults aged 18–64 in 34 economies, it applies country-normalised survey weights so that each included economy contributes equal aggregate weight, together with logistic regressions with economy fixed effects and economy-clustered standard errors. The conceptual framework distinguishes institutional legibility, represented by reported foundational identity ownership, from transactional capability, represented by digitally enabled account ownership; it does not assume that these components form a causal or temporal pathway to borrowing. The primary specification omits income, which is structurally unavailable in Lesotho; income-adjusted sensitivity analyses use 29,037 respondents in the other 33 economies. Before account adjustment, the identity estimate is positive but statistically uncertain (average marginal effect [AME] 1.33 percentage points; 95% CI −0.09 to 2.76; p = 0.066); after account adjustment, it is 0.64 points (95% CI −0.92 to 2.20). Holding a digitally enabled account is associated with a 6.10 percentage-point higher predicted probability of reporting business-purpose borrowing (95% CI 4.62 to 7.59). A separate account/outcome model associates reported identity ownership with a 10.43 percentage-point higher account probability (95% CI 6.57 to 14.30). Null-imposed wild-cluster-bootstrap tests support the account-borrowing and identity-account coefficient associations, while the identity-borrowing estimate remains uncertain. The account association remains positive across the reported alternative samples and weights, while exploratory interaction analyses provide no statistically reliable evidence of variation by sex, residence, or income group. Because the outcome captures self-reported borrowing for business purposes rather than formal credit approval, source, price, or productive use, the findings are interpreted as conditional associations rather than evidence of mediation or causal sequencing. The strongest evidence-based implication is therefore narrow: identity coverage alone is not a reliable marker of business-purpose borrowing, whereas digitally enabled account ownership is a more robust correlate under the reported specifications.