End of extreme poverty: A case for action and a plan to start

A decades-old ambition

For 50 years, from 1960 to 2010, poverty reduction was driven by rapid economic growth in poor countries, with most of the benefits accruing to the poor through more employment, better wages, and higher agricultural productivity. By 2015, there were 1.1 billion fewer people in extreme poverty than in 1990. The hope that the remaining 840 million people could be lifted out of extreme poverty by 2030 became the foundation of the U.N.’s Sustainable Development Goals (SDGs).

Since 2015, however, poverty has become increasingly concentrated in a few fragile states with limited prospects for private investment or growth. Over the past decade, poverty reduction has stalled: About 800 million people still lived in extreme poverty in 2025, with sub-Saharan Africa accounting for roughly 67% of all people living in extreme poverty.

A new opportunity

A new model for poverty reduction, based on targeted cash transfers to poor households, has been taking shape for the past two decades. Thanks to AI, improvements in digital payment technologies, and the ubiquitous access to phones that can receive SMS texts, money can be transferred efficiently and effectively to intended beneficiaries.

Evaluations of this model have documented long-term, consistent positive effects on consumption, food security, savings, asset accumulation, children’s school attendance, health service utilization, and resilience to economic shocks. However, most evaluations are based on pilot programs or relatively small-scale interventions implemented under controlled conditions. Only a limited number of cash transfer programs have been scaled geographically and sustained over long periods, particularly in low-income countries. As programs expand, new challenges emerge relating to fiscal sustainability, targeting accuracy, administrative capacity, political economy, inflationary effects in local markets, and the extent to which transfers generate lasting productive and labor-market outcomes.

Consequently, important questions remain around the optimal design of such programs, including the appropriate transfer size and frequency, duration of support, targeting mechanisms, conditional versus unconditional transfers, digital delivery systems, and integration with complementary interventions such as livelihood support, financial inclusion, and human capital investments. Answers to each of these questions shape the long-term effects on poverty reduction, employment, productivity, and economic transformation. A better understanding is essential for designing cash transfer programs that not only alleviate poverty but can also be scaled to promote broader goals of societal inclusion, resilience, and strong economic growth.

New research reframes the question of scale from an academic assessment of the “poverty gap” (the theoretical amount to bring all households above the extreme poverty threshold to meet the goal of SDG 1) to costing out, on a country-by-country basis, a program that would bring extreme poverty rates below 1% of the population, using data from the practical experience of targeting in existing programs. The headline finding is that in the 23 countries where about half the world’s poor live, it would cost approximately $170 billion per year to reduce poverty from 12% to 1% of the population. This amounts to less than $500 per person.

Extrapolated to the whole world, the cost would represent about 0.3% of global GDP. Put differently, that’s $30 for every $10,000 of global GDP, or just 14% of the wealth gained by global billionaires in 2025 alone. The financing needed to end extreme poverty is not a moonshot; it’s a rounding error that the world has simply chosen not to correct.

Calculations of this type reframe poverty eradication as a feasible, targetable financing problem. It is ambitious but not hugely expensive by global standards, small enough that a handful of high-net-worth individuals could have a material impact, and simple enough to be scalable and easily implementable without having to delve into the complexities of institutional reform, structural transformation, and long-run growth that characterized the old model.

A proposal: Prove it’s possible, then go global

Already a decade ago, there were more than 200 evaluations of cash transfer programs in Latin America, sub-Saharan Africa, and South Asia. A 2025 meta-study by Crosta et al. covered 72 programs in 34 low- and middle-income countries (LMICs). The authors find that unconditional cash transfers (UCTs) have positive average effects on most major outcomes, including consumption, income, labor-force participation, school enrollment, food security, psychological well-being, assets, and child height-for-age, with no evidence on average that cash discourages work. The paper also shows that design matters: Streamed transfers are consumed more, lump sums are invested more, targeting women appears to strengthen effects on consumption and income, and ongoing transfers tend to sustain gains better over time.

These evaluations, although largely based on fragmented and small-scale programs, are helping to narrow the band of uncertainty around impact. Large cash infusions equivalent to roughly 15% of local GDP in Kenya produced minimal price inflation. That result suggests macroeconomic risk may not be a major problem. Exchange rate effects from sustained foreign-currency inflows, political economy dynamics around targeting and phasing, and the design choices between lump-sum and recurring payments all remain serious concerns—but have not been tested at national scale. Whether transfer effects are heterogeneous across household types, and whether the macroeconomic results observed at the village level would replicate where fiscal and monetary feedback loops operate differently, are genuinely open empirical questions.

Acknowledging these limits is a reason to design for learning rather than to delay. The case for a national-scale program is precisely that these questions require national-scale evidence to answer. A program structured from the outset around rigorous evaluation—and run in partnership with a government that can take it to scale if it succeeds—is more likely to generate usable knowledge and durable outcomes than a further accumulation of small pilots.

Government partnership converts a philanthropic experiment into a test of state capacity, building the administrative infrastructure for digital transfers, the targeting mechanisms, and the monitoring systems within a national government in ways that outlast any individual funder. A program that runs in parallel to the state, without embedding capability within national systems, cannot be replicated at scale even if it succeeds.

The case for action is a no-lose proposal. Even under pessimistic assumptions, millions of people would receive direct cash income. If only 10% of recipients escape extreme poverty permanently, that’s still 500,000 to 1 million people lifted out of deprivation by a modestly sized program. The case for trying does not require agreement on the probability of full success.

And if it does succeed, the result would carry the evidentiary weight needed to reshape how international financing approaches extreme poverty—something no subnational pilot, however carefully designed, can achieve. Put differently, a national-scale program can serve as a policy prototype: a real government-embedded intervention that improves lives immediately while generating evidence on the questions that only implementation at scale can answer.

None of this is a claim that cash transfers are sufficient on their own to build durable development. They do not build roads, staff clinics, or train teachers. The argument here is narrower: For the several hundred million people currently below the international extreme poverty line, a well-designed digital cash transfer program at national scale is likely to produce lasting and measurable improvements in well-being.

The approximate cost of a national-scale proof-of-concept—structured as a government partnership with evaluation embedded from the start—is in the range of $1 billion per year in philanthropic seed capital, preferably running for a five-year duration. The math of this moment is worth stating plainly: The financing needed to run a decisive test at national scale is comfortably within the reach of a very small number of wealthy individuals. It does not require multilateral consensus or government action to initiate. It requires a handful of actors with the means and the will.

The case for action and next steps

The case for action is clear—not because the path is certain, but because the cost of trying is affordable and evidence-backed, while the potential benefits are large, scalable, and durable.

A proof-of-concept that can potentially impact the lives of hundreds of millions of people could be enacted through the following
sequence of steps:

1. Solicit interest from one or more high-net-worth individuals to provide initial seed capital for a national program. Target $1 billion annually for five years.
2. Identify a willing government prepared to trial a national-level program, potentially with some matching funds.
3. Design a program with learning at the center. Rigorous evaluation needs to be built in from day one. Technical design issues described above (macro and political economy risks, spillovers, lump-sum or streaming, etc.) should provide new information on what works best at scale.
4. Identify success criteria and a monitoring framework. The gap in the number of people remaining in poverty should be meaningfully closed. This does not require perfection. It does need an upfront definition of what “gold, silver, bronze” looks like.
5. Create an independent expert advisory group. Outside eyes build honesty and credibility in the findings.

That is Act 1: a national-scale proof-of-concept, government-embedded and built for learning. If it works well—and the evidence suggests it will—the focus will shift to how fast such a concept can be globalized to accelerate the end of extreme poverty worldwide.

Understanding the impact of cash transfers at scale is the missing evidence link between existing programs and the rollout of national programs with multiple funders. The proposal above would provide that evidence.