Summary
- 💸 We’ve sent nearly $1b to people in poverty — next up: $5b by 2035 through smarter, faster, recipient-first programs in our mission to accelerate the end of extreme poverty.
- 🛠️ From cash for new moms to planning nudges and AI tools, we’re piloting designs that could have much more impact per dollar and be scaled by governments.
- 🎯 We continue to hold new innovations to a high bar: they must be respectful, scalable, and more impactful than our standard program.
- 🚀 We’re seeking another $25m to accelerate our innovative pilots and have even more room for funding for our ever-improving flagship Poverty Relief programs.
This year, GiveDirectly will hit a milestone few thought possible: $1 billion committed to people in poverty, alongside 25+ independent studies rigorously testing how that cash was spent, what changed, and why.
Now we’re raising the bar. GiveDirectly has set an audacious new goal: deliver $5 billion by 2035 to accelerate the end of extreme poverty and meet humanitarian needs.
Getting there won’t just require more of the same. It demands we grow faster, test smarter, and build for the future. Three shifts are pushing us forward:
- Rising need, shrinking resources: Crises are compounding, aid budgets are not. We must deliver more impact per dollar to reach more people in need.
- Faster tech, new tools: AI and digital finance are changing how people work, save, borrow, and learn. Our cash delivery should take advantage of these without exacerbating inequity.
- Deeper evidence base: With 1,000 studies and counting, we now know far more about what cash can do and are designing programs that push beyond “does it work?” to “what works best in which contexts?”
Call it “GiveDirectly 2.0.” We’re continuing to scale what’s working — village-wide, unconditional lump sums — with greater speed and reach. We’re also adapting for the places and people where traditional aid falls short: displaced families, pregnant women, climate-vulnerable regions.
That means more targeted designs, like transfers timed to pregnancy or structured to test the most cost-effective paths to lift incomes. To drive this, we’ve built a dedicated product function focused on optimizing our model: testing more, iterating faster, and shaping programs around the outcomes that matter most to recipients and donors.
Even as we evolve, our guardrails remain clear: every GiveDirectly program must be respectful to recipients, effective at improving lives, and scalable to many places.
✅ What we’re still doing: Scaling what works
💸 Large lump sums to people in extreme poverty
Our core program remains one of the most effective and efficient ways to support people in extreme poverty. The evidence base is robust:
- Recipients see lasting increases in income, assets, and consumption, even seven years after receiving cash
- Spillover effects are real: every $1 in cash can generate up to $2.50 in economic activity for communities
It works at scale: we’re set to deliver $90m in lump sum transfers in 2025 (about 70% of our work), and could responsibly absorb and deliver an additional $250m+ in the next year
In GiveDirectly 2.0, we’re not just scaling our core program — we’re improving it. We’re investing more in our core technology so that we can launch programs faster and better adapt to local needs. That unlocks the ability to test questions big and small that matter for outcomes:
- 💸 What’s the ideal transfer size for different communities or contexts?
- 🤖 How can we leverage AI to drive more efficient and inclusive programs?
- 🗣️ Which eligibility explanation best reduces confusion or stigma?
- 📝 Which survey question yields the most honest answers?
🆘 Rapidly responding to disasters around the world
Since 2020, we’ve delivered $25m to over 70,000 people in response to rapid onset crises in increasingly sophisticated ways, leveraging partnerships and technology to get cash to victims of hurricanes & cyclones, fires, floods, earthquakes, and conflict.
And, we’ve gotten faster — faster at making decisions, faster at raising money, and faster at deploying cash.
This year, we’re doubling down on the technical infrastructure and partnerships needed to send life-saving cash anywhere in the world within 5 days of a crisis. Not just to show it’s possible, but because this kind of speed will increasingly define whether or not people survive and help shrinking aid budgets stretch further.
We’re also investing in research to help us understand the relative impact of cash, information, and quick action.
🌍 Delivering cash in conflict-affected areas
More than 60% of the world’s poorest people are now living in fragile or conflict-affected states. That means ending extreme poverty requires reaching those places quickly, safely, and reliably for both our large lump sums and our crisis payments.
In 2025, we’re testing:
- Remote targeting to identify and pay people displaced by conflict in DRC
- Giving cash for livelihoods in the insurgency-affected Memba district of northern Mozambique
- Flexible delivery models so recipients can receive funds when and where they’re safe
🆕 What we’re starting: Designing for higher impact per dollar
Our large transfer program is already one of the best bets in global development. We think cash can do even better on specific outcomes with the right adaptations.
👩🏾🍼 Cash for new moms and babies
1 in 6 babies is born into extreme poverty each year. More than 4.5 million of them will die before their fifth birthday, most from preventable causes.
Early evidence suggests that cash delivered during pregnancy or infancy could cut child mortality in half. That’s what we saw in Kenya, where large transfers coincided with better access to care and nutrition. In Rwanda, similar designs led to a 70% drop in child deaths.
We believe this could be one of the most cost-effective ways to save lives and reduce poverty, which is why pursuing new pilots that target cash specifically to pregnant women.
In September, we’re partnering with a Kenya community health organization, Lwala Community Alliance, to deliver cash to pregnant women alongside care at trusted, local health facilities. Each woman will be randomly assigned to one of two transfer structures, letting us test which model delivers better outcomes for moms and babies. A similar pilot is in early design in eastern DRC.
We’re seeking funders and partners to scale what we learn and deliver one of the most cost-effective ways to save lives.
🤝🏾 Cash with planning nudges
Over the past 15 years, studies have found that adding personalized coaching or nudges to cash can boost impact, especially on long-term income gains. Lighter-touch models that focus on mindset and planning, not technical advice, have outperformed cash alone in places like Kenya, Liberia, Niger, and Uganda.
We’re now testing a “cash with planning nudges” model that layers simple, behavioral support on top of a large lump sum transfer. Unlike complex “graduation” programs — 1 to 2 years of in-person coaching plus job training while giving cash and assets (🐄🛠️) — ours is a lighter, more scalable model.
For now, we’re only coaching recipients on planning and executing their own ideas of how to spend their lump sum, using existing channels like community meetings, call centers, and follow-ups to keep execution simple and costs low. The goal is to find out whether a lighter-touch approach can match the impact of more intensive graduation programs but at a fraction of the cost.
🔍 Innovating for 10x the impact
We’re piloting new program designs that could deliver significantly more impact per dollar, up to 2, 3, or even 10 times as much. These pilots aim to outperform our core cash program on cost-effectiveness:
- 🗓️Reaching people at key moments like becoming a parent or finishing school
- 🌱Giving cash alongside existing education, agriculture, or water access interventions
- 📲Pairing cash with smartphones and AI-enabled chatbots giving farming advice
- 🧱Giving village committees $14k to spend on local infrastructure projects
We’ll incorporate the ones that work into our existing programs or spin them off as new ones. This effort is funded in part by GiveWell, and will shape our pipeline of innovation going forward.
Our next billion dollars will be even more impactful
Our goal is maximizing impact for recipients in increasingly scalable and cost-effective ways.
We’re doing this by systematically building and testing program variations that meet our three-part bar: they must be respectful, scalable, and more impactful than simply giving the additional costs to recipients as cash. That benchmark challenged the sector a decade ago and we’re still holding ourselves to it.
If you want to back proven, repeatable impact at scale, our Poverty Relief program is shovel-ready and underfunded: we have $90m to deliver in 2025, but capacity to give out $250m+.
If you want to back innovation — supporting fragile states, new mothers, or next-generation program technologies — we’re raising an additional $25m to move from early-stage pilots to scalable, cost-effective models. Some of these haven’t been tested yet; this fund is how we do that.
We’ve delivered nearly a billion dollars in highly impactful ways. Now we’re ensuring the next billion will do even more.
Appendix: 🛑 What we’re stopping: Broad-based basic income in Africa
Our historic 12-year universal basic income (UBI) study in Kenya remains a landmark in cash research, which we plan to run to its full conclusion, but evidence and context have evolved. Researchers’ early findings show:
-
- Lump sums outperformed basic income on most economic indicators at the same cost
-
- Short-term UBI designs, while still helpful, produced less impact per dollar on most measures
-
- Other research found 95% of recipients prefer lumpier payments to smaller monthly ones
In response, we’ve wound down broad-based basic income designs, including our program in Maryland, Liberia, and won’t be launching new ones in Africa. Instead, we’re prioritizing:
-
- Larger transfers that enable investment, not just survival
-
- Testing basic income to specific moments of high leverage, like pregnancy, early childhood, or post-displacement