Working with the [Innovations for Poverty Action] team, [Harvard economics professor Michael Kremer] first tried making the water cleaner through making improvements to water-supply sites (largely through engineering features that filtered out contaminants while the water was still underground). The water was clean when it came out of the source, but by the time it reached people’s homes, it was contaminated again. Guessing that the weak link might be the containers people were using to carry the water home, the team distributed a different kind of container, only to realize that other sources of contamination meant people were still drinking water that could make them sick: for instance, pouring it into a contaminated drinking glass, or storing it in an open container into which a child might plunge a grimy hand. They decided they had to distribute chlorine, which would not only sanitize the water, but make the water itself into an agent that would sterilize vessels with which it came into contact.
The next trial involved distributing chlorine in small plastic packages for people to add to their water at home. Not even 10 percent of the people in the trial actually used the chlorine; on follow-up surveys, they said they neglected to use it because they forgot the directions for how much to add and when, or simply forgot altogether. In tests of different advertising campaigns—some based on fear (the idea that unclean water is dangerous), some appealing to people’s consciences (good parents provide safe water for their children), some invoking aspirational feelings (“City people use WaterGard”)—none proved particularly effective in getting people to use the chlorine.
The team went back to the drawing board, deciding to try placing the chlorine at central collection sites like the one near Busia. In the pilot phase of testing, the scheme proved far more effective—and far less expensive—than giving the chemical to people in containers sized for use in an individual household. Because such a small amount is needed—“a bottle of bleach you buy at CVS could treat water for one person for 40 years,” says Kremer—the packaging for the earlier trial cost more than the chlorine itself. Installing the dispensers in a public place has the added benefit of potentially creating a new social norm: as people see others doing it, they become less fearful of putting this unfamiliar substance in their water.
Tests of the drinking water at people’s homes indicate that this operation is successful: pre-intervention testing (after distribution of chlorine for household use; before installation of the dispensers) found chlorine in the drinking water at just 8 percent of homes during unannounced visits; post-intervention, it was 62 percent.
--Elizabeth Gudrais, Harvard Magazine, on why it's not enough to just clean up developing country water supplies