Why an evidence approach matters more than ever

Inclusive credit, defined as formal and semi-formal products for low-income borrowers, operates at a global scale with an estimated total portfolio of approximately $1.5 trillion. At this magnitude, even modest design or governance failures affect hundreds of millions of households. Historical debt crises in places like Andhra Pradesh (2010) and Cambodia demonstrate that aggressive expansion without attention to borrower context leads to over-indebtedness and trust erosion. This report argues for a shift from asking whether credit “works” to determining under what specific conditions, for which borrowers, and through what delivery modes it generates benefit rather than harm. Modern analytics, including the Impact Pathfinder and Precision Causal Modelling (PCM), now allow for more outcome-oriented decision-making based on longitudinal data.

How the evidence has evolved: From transformative claims to conditional impact

The evidence base has transitioned through several phases. Early optimism was challenged by a generation of randomised controlled trials (RCTs) between 2003 and 2012, which found that average effects were often negligible and did not consistently reduce poverty or improve social indicators. However, subsequent meta-analyses revealed a “heterogeneity breakthrough”, showing that while average effects might be modest, significant gains are concentrated among specific groups, such as borrowers with prior business experience. Conversely, market failures occur when rapid commercialisation outpaces consumer protection, leading to coercive collection and debt spirals. The evidence establishes that serious harm follows predictable patterns of poor governance and aggressive growth targets.

Five factors that shape outcomes

The report identifies five intersecting factors that determine credit outcomes: who borrows, how it is structured, what it finances, where it operates, and the time horizon. Regarding borrower characteristics (WHO), gender is a critical modifier; credit to women often yields higher household welfare gains, but only if they retain genuine control over the funds. If men appropriate the proceeds while women bear the repayment liability, it creates a harmful “risk-without-benefit” pattern. Economic status and productive capacity also dictate impact, as those with existing assets or business experience are better equipped to absorb credit. Households in phases of asset accumulation are more likely to generate sustained gains than those facing high dependency burdens.

How credit is structured and delivered

The design and delivery (HOW) of credit are potent levers for financial services providers (FSPs). Repayment flexibility, such as grace periods and seasonal schedules, is associated with higher profits and reduced default rates. In Pakistan and Bangladesh, flexible repayment increased business capital by up to 80 per cent and profits by 25 to 55 per cent. Digital disbursement into borrower-controlled accounts improves privacy and autonomy; a Ugandan RCT found it increased business profits by 15 per cent. Furthermore, transparency interventions, such as those in Ghana that reduced misconduct by 72 per cent, are essential. Bundled services like training or insurance can multiply effectiveness by addressing capability gaps and risk exposure.

What credit finances: Use of funds

Productive investment (WHAT) in equipment or technology produces the most consistent returns, provided borrower capability and market conditions align. Human capital and mobility credit can transform long-term trajectories, while consumption smoothing serves a protective function during isolated shocks but becomes harmful if used repeatedly to bridge a structural income gap. Recovery credit presents a distinct challenge; well-designed structured programmes, like those in Haiti and Nepal, achieved repayment rates of 95 to 99 per cent and significant income gains by using criteria calibrated to post-shock opportunities rather than past income.

Where credit operates and when it helps

Contextual factors (WHERE) like regulation, market infrastructure, and climate risk are primary determinants of impact. In Uganda, credit for women farmers failed because of a lack of output markets, not a lack of credit access. Social norms also act as filters; credit produces strongest gains where norms support women’s economic roles. Time horizon (WHEN) is frequently neglected; gains from productive investment often take years to manifest. For instance, the Hyderabad study found decisive firm-level gains at six years that were invisible at 18 months. Long-run welfare is shaped by borrowing trajectories—how clients progress across successive cycles with graduated loan sizes.

Making it work: From evidence to precision-led impact

The report provides recommendations for key stakeholders. FSPs should screen for absorptive capacity rather than just repayment eligibility, design products around borrower cash flows, and monitor financial health trajectories. Funders and impact investors should prioritise “funding precision” over volume, align evaluation horizons with the logic of the credit being financed, and require distributional reporting. Policymakers and regulators should build enabling infrastructure, such as interoperable instant payment systems and open finance frameworks, and enforce transparency. They must track borrower trajectories and intervene before harm accumulates, treating vulnerability as a product design challenge rather than a reason for exclusion.