Introduction

Access to finance is critical for economic development, yet formal financial markets are often absent in rural areas of low-income countries. While informal institutions such as Village Savings and Loan Associations (VSLAs) mitigate some frictions, they typically operate at a small scale and provide limited services. This research investigates whether and how these markets can be upgraded, not only by expanding access but by increasing usage and service quality through consumer accountability. The study, conducted in rural Sierra Leone, tests a scalable approach that upgrades existing VSLAs to operate as formal digital financial service (DFS) agents.

Background

Sierra Leone has historically struggled to expand formal financial services to its rural interior, where over 60% of the population resides. Although the mobile money sector is growing, retail agent networks remain concentrated in urban hubs due to high liquidity costs and fragmented demand. The project partnered with a consortium led by the Catholic Agency for Overseas Development (CAFOD), Orange Money, and GT Bank. The study focused on 195 ‘virgin’ communities that lacked pre-existing DFS agents and had populations ranging from 115 to just over 5,700 people (with a mean under 750).

Experiment

The experiment utilised a community-level randomised controlled trial with two distinct treatment arms. Treatment I (Upgrade) encouraged VSLAs to enrol as formal retail agents for Orange Money and GT Bank, providing them with necessary infrastructure such as solar panels, handsets, and initial liquidity. Treatment II (Upgrade + Feedback) augmented this infrastructure with a monthly, anonymised feedback mechanism. Innovations for Poverty Action (IPA) enumerators conducted phone audits with DFS users to solicit evaluations across five service dimensions: fees, privacy, transparency, reliability, and satisfaction. These reviews were then aggregated and presented publicly during community ‘barray’ gatherings to create reputational incentives.

Conceptual framework

The authors developed a framework of delegated intermediation to clarify how infrastructure and active accountability interact. In this model, the provider acts as a delegated agent whose effort in maintaining service quality and consumer protection is costly to monitor. The framework suggests that while providing infrastructure is a necessary condition for usage, it does not inherently align provider incentives with consumer welfare. Accountability is required to discipline agent actions, bolster consumer trust, and drive intensive-margin usage.

Data

The researchers combined three data sources: high-frequency behavioural audits, detailed household and firm-level surveys, and administrative records from commercial and NGO partners. Administrative transaction data from Orange Money and GT Bank provided objective measures of agent activity, including transaction volumes and commissions earned. Baseline data confirmed the target communities were cash-constrained and poor, with 62.9% of households estimated to fall below the national poverty line.

Results

The study documented five main results. First, connecting communities to DFS significantly expanded access, with formal account ownership rising by 5.2 percentage points (+23%). Second, introducing structured feedback substantially amplified these gains, increasing account ownership by 9.1 percentage points (+39%) and mobile-money usage specifically by 10.4 percentage points (+29%). Feedback induced providers to improve service quality; Upgrade + Feedback agents were 13.2 percentage points less likely to allow a third person to be present during a transaction and reported a 20.5 percentage point (85%) increase in making changes in response to feedback.

Thirdly, while provider entry alone raised trust in mobile-money agents by 6.7–6.9 percentage points, feedback sharpened provider conduct without necessarily changing general stated trust. Fourthly, upgraded VSLAs showed significant spillover effects, including greater digitisation of member contributions, lower default rates, and higher group-level income. Finally, structured feedback improved household-level shock mitigation; households in these communities were 3.8–4.6% less likely to report a financial shock they could not remedy.

Interpreting the results and scalability

The results indicate that accountability drives market development by aligning provider incentives with consumer welfare. Upgrading alone generates limited gains, but adding structured feedback produces larger, more robust improvements in behaviour, trust, and usage. The intervention is highly cost-effective, with an implementation cost of approximately $5.58 per household for the Upgrade arm and $10.43 for the Upgrade + Feedback arm. The aggregate annualized return was estimated at 276.1% under Upgrade and 323.4% under Upgrade + Feedback. Based on these gains, CAFOD has begun scaling the model to additional villages across Sierra Leone.

Conclusion

Successful financial market development depends not only on increasing the supply of services but on ensuring providers respond to consumer experiences. The study demonstrates that pairing access with structured feedback creates a functional market that benefits consumers, VSLAs, and commercial providers alike. This research highlights the underappreciated role of quality and consumer voice in converting financial access into active adoption and sustained economic benefits for the poor.