When Does Digital Finance Empower Women? Causal Evidence and Mechanisms From Rural Bangladesh

ABSTRACT Bangladesh's rapid expansion of digital financial services (DFS) raises a central question: do these tools empower rural women, and under what conditions? We analyze survey data from 1202 rural households, complemented by six focus group discussions, and estimate the causal effect of DFS adoption on women's empowerment using two‐stage least squares with agent‐outlet density and mobile network coverage as instruments. We test mechanisms via structural‐equation mediation (DFS → financial inclusion → empowerment) and examine heterogeneity by digital literacy and proximity to agents. DFS adoption increases women's empowerment by 0.37 SD (IV, p < 0.01). Mediation analysis indicates that approximately 54% of the total effect operates through financial inclusion, consistent with gains in savings autonomy, control over remittances, and liquidity during shocks. Effects are largest among women with above‐median digital literacy and among households residing within ~2 km of an agent outlet. Robustness checks, including placebo outcomes and alternative index constructions, yield similar conclusions. This study extends empowerment theory by conceptualizing the impacts of DFS as relational and conditional, rather than automatic. Empowerment primarily emerges through financial inclusion pathways and is moderated by literacy and access. Furthermore, Bangladesh's multi‐channel ecosystem—mobile money, agent banking, and G2P digitization—demonstrates that empowerment outcomes depend on how platforms interact, not merely on access to a single service. In Bangladesh, DFS can shift intra‐household power, but access alone is insufficient. Policy should pair network expansion with targeted digital‐literacy programs and ensure reliable agent liquidity and proximity to deliver inclusive empowerment gains.