Digital financial inclusion beyond repayment: trust, task diversity and borrower autonomy in Indian microfinance

Abstract

Abstract

Purpose - This paper examines whether participation in digital microfinance transactions translates into broader borrower capability, and whether technology-acceptance, trust, risk and enabling conditions are associated with behavioural intention.

Design/methodology/approach - The supplied cross-sectional survey file for Ranchi district, Jharkhand, India was audited before analysis. It contains 662 complete respondent records. Composite scores were formed from 29 five-point Likert items. The analysis uses task-level descriptive statistics, Cronbach's alpha, a paired-samples test of institutional versus technological trust, HC3-robust OLS, variance inflation factors and an unweighted sensitivity analysis based on unique multivariate response profiles.

Findings - Balance inquiry (96.1%) and loan repayment (87.9%) were common, while fund transfer (55.7%), bill payment (36.0%) and new digital loan application (16.0%) were less prevalent. The repayment-transfer difference was 32.2 percentage points. Trust in the MFI (M = 4.37) exceeded trust in technology (M = 3.38), with a paired mean difference of 1.00 point (p < .001). In the full-file HC3 model, several constructs were statistically associated with behavioural intention, but severe multicollinearity and exact repetition of response profiles materially limit coefficient-level inference. In the unique-profile sensitivity analysis, perceived financial benefit, social influence and facilitating conditions retained positive associations at the 5% level.

Practical implications - MFIs and regulators should evaluate digital inclusion with measures of task diversity, independent initiation, authentication, security and error recovery, rather than relying on electronic repayment volumes alone. Data-quality controls are equally important when borrower-level evidence is used for policy inference.

Originality/value - The paper separates institutional digitisation from borrower-level financial agency through the Compliance-Autonomy Gap and the Trust-Enabled Phygital Adoption Framework (TEPAF). It also demonstrates how file-level integrity checks and sensitivity analysis can change the interpretation of highly fitted technology-adoption models.