From Digital Compliance to Financial Autonomy: Trust, Task Diversity, and Phygital Adoption in Microfinance

Abstract

Digitally recorded transactions do not by themselves establish borrowers’ independent control of digital financial services. This study re-estimates an interviewer-assisted survey of 662 active microfinance borrowers in Ranchi district, Jharkhand, using constructs drawn from the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT). Descriptively, digital loan repayment was reported by 582 borrowers (87.9%) and fund-transfer use by 369 (55.7%), a paired difference of 32.2 percentage points (95% CI 28.6–35.7; exact McNemar p < .001). Across six digital task indicators, respondents reported a mean of 3.71 tasks (SD = 1.61; median = 4; IQR = 3–5). Trust in the MFI was high (M = 4.37, SD = 0.49) and exceeded trust in technology (M = 3.38, SD = 1.16) by 1.00 scale point (95% CI 0.88–1.11; t(661) = 17.38, p < .001, dz = 0.68). An eight-predictor behavioural-intention OLS model yielded R² = .9851 and adjusted R² = .9849; HC3 robust standard errors were used because heteroscedasticity was present. However, VIFs ranged from 1.86 to 86.89, and only 99 distinct substantive response profiles occurred among 662 records, so the partial coefficients are not interpreted as independent causal effects. Profile-grouped ridge and LASSO models were used as stability and prediction checks, and a ridge-logistic model examined fund-transfer participation directly. These models show strong predictive structure but do not identify transaction need, assistance, or independent execution. Accordingly, the Compliance–Autonomy Gap is interpreted as a task-use contrast rather than a validated autonomy measure, while the Trust-Enabled Phygital Adoption Framework (TEPAF) remains a conceptual framework for future testing of the transition from assisted participation to independent financial agency.

Keywords: digital microfinance; financial inclusion; borrower autonomy; Compliance–Autonomy Gap; Technology Acceptance Model; UTAUT; trust asymmetry; regularized regression; phygital finance; TEPAF; Jharkhand