Keep a family's drinking water clean all year.
A dispenser of chlorine solution stands at the water source. People turn the valve as they fill their container, and the dose is already measured for them.
Choose your ripple
Direct only: Every.org lists Evidence Action as a whole, and only its own form can earmark this program. $25 is filled in.
Diarrhoeal disease is one of the leading killers of children under five, and much of it comes from drinking water carrying faecal bacteria. Chlorine kills those bacteria cheaply and keeps working in the container on the walk home. The chemistry has never been the hard part. Getting people to use it every single day is.
- Drinking waterWorld Health OrganizationHow many people drink contaminated water and what it does to children.
- Clean waterOur World in DataAccess to safe water by country, and the diarrhoeal disease burden it drives.
- Water quality interventionsGiveWellThe evidence on chlorination and child mortality, and why GiveWell revised it upward.
GiveWell's 2022 cost was $1.22 to $1.87 per person per year. Its 2025 review cut estimated reach by about 20%, and our figures apply that cut, so they understate.
The figures
What Evidence Action's Dispensers for Safe Water does
Evidence Action installs chlorine dispensers beside rural water sources in Kenya, Uganda and Malawi, keeps them stocked, and employs local promoters to encourage use. Placing the dispenser at the source is the whole idea: nobody has to remember to buy chlorine or carry it anywhere.
How GiveWell gets there
Water chlorination has strong trial evidence behind it, and a 2022 meta-analysis raised GiveWell's estimate of how much it reduces child mortality. GiveWell models the number of people drinking treated water, how much diarrhoeal disease that prevents, and how many deaths follow from that, then discounts heavily for uncertainty about the mortality effect.
What could be wrong
- In 2025 GiveWell cut its estimate of the program's reach by about 20%, after an independent study and Evidence Action's own survey found far fewer people were drinking chlorinated water than routine monitoring reported. In Kenya the gap was around 40%. This is a measurement correction rather than a decline in the program, but it means earlier per-person figures were too optimistic.
- At roughly 5 times cash, this sits below GiveWell's bar for a Top Charity. GiveWell keeps funding it, but it is not one of their four strongest picks.
- The mortality benefit rests on a meta-analysis whose central estimate GiveWell itself treats as uncertain, and which it discounts substantially in its model.
- A dispenser only helps if people use it. Adoption is the number that moved, and it is the number most worth watching.
Every figure is an average program cost, not a receipt. How we phrase things.