π
Extreme Poverty
Roughly 800 million people live on less than $3.00 a day
β Improving
πΎ Poverty
π§ Understand
The trend
The share of humanity in extreme poverty fell from ~38% in 1990 to under 10% today β one of the great achievements of our era, though progress has slowed since COVID.
The scale
On the order of 800 million people β about 1 in 10 humans β live below the World Bank extreme poverty line of $3.00/day (2021 prices, adjusted for local costs). Most live in Sub-Saharan Africa and fragile, conflict-affected states.
Root causes
Conflict and displacement, weak institutions and infrastructure, limited access to markets, credit, education and healthcare, and climate shocks that wipe out savings held as crops or livestock.
Who suffers most
Rural families, children (who are over-represented among the extreme poor), and people in conflict zones. Poverty at this level means chronic hunger, untreated illness, and children working instead of learning.
Common misconception
"Global poverty keeps getting worse." The opposite: extreme poverty has seen the steepest sustained decline in human history over the last generation. The work is unfinished, not failing.
β‘ What actually works
Direct cash transfers
Strong evidence
Giving money directly to the poorest households, unconditionally. Recipients invest in food, roofs, livestock and small businesses.
Cost & effect: Roughly $1 delivers ~$0.85β0.90 into a family's hands; well-studied across dozens of trials.
Graduation programs
Strong evidence
A package: a productive asset (like livestock), training, coaching and a small stipend, to help families build a livelihood.
Cost & effect: Higher cost per household (~$300β2,000+, varying widely by country and package), with benefits persisting years after the program ends.
Microfinance
Debated
Small loans to poor entrepreneurs. Popular for decades.
Cost & effect: RCTs find modest average effects β helpful for some, not transformative at scale; less effective than the hype suggested.
π§ Act
πΆ With your money
Set up a small monthly gift to org types doing direct cash transfers or graduation programs β consistency beats one-offs.
Use the "What Would $X Do?" tool to see what a specific amount achieves against poverty.
β° With your time
Volunteer with local organizations serving low-income families β poverty is global and local.
Join or start a giving circle focused on global poverty.
π οΈ With your skills
Offer professional skills (accounting, software, design) pro-bono to poverty-focused nonprofits via skills-based volunteering platforms.
π£ With your voice
Share the real trend β that extreme poverty is beatable β to counter fatalism.
Advocate for development aid and open trade policies with your representatives.
π£ Share this
β Questions people ask
How can I help with extreme poverty?
There's a concrete step for whatever you can offer. With your money, set up a small monthly gift to org types doing direct cash transfers or graduation programs β consistency beats one-offs. With your time, volunteer with local organizations serving low-income families β poverty is global and local. With your skills, offer professional skills (accounting, software, design) pro-bono to poverty-focused nonprofits via skills-based volunteering platforms. With your voice, share the real trend β that extreme poverty is beatable β to counter fatalism.
What is the most effective way to reduce extreme poverty?
The approaches with the strongest evidence: Direct cash transfers: Giving money directly to the poorest households, unconditionally. Recipients invest in food, roofs, livestock and small businesses. Roughly $1 delivers ~$0.85β0.90 into a family's hands; well-studied across dozens of trials. Graduation programs: A package: a productive asset (like livestock), training, coaching and a small stipend, to help families build a livelihood. Higher cost per household (~$300β2,000+, varying widely by country and package), with benefits persisting years after the program ends.
Where should I donate to help with extreme poverty?
Impact Compass doesn't name individual charities. The higher-leverage path is to back the interventions that work best here (Direct cash transfers, Graduation programs) and to choose organizations by how transparently they deliver them. Compare organization types for this cause with the free tools linked above, or give useful items directly through Givelink.
βοΈ One world problem a week
Get one problem, and one thing that actually works against it, in your inbox. Free, once a week, unsubscribe anytime.
Rough figures for context, drawing on: World Bank Β· Our World in Data Β· GiveWell. Approximations, not citations. Last reviewed 2026-07-16.
π Embed this problem card on your site β free
Paste this snippet anywhere. It renders a live card for this problem linking back here. Drop the ?problem=β¦ part to rotate a different problem every day.