Small bribes may help people build healthy handwashing habits

By | May 12, 2021

Good habits are hard to adopt. But a little bribery can go a long way.

That’s the finding from an experiment in India that used rewards to get villagers hooked on routine handwashing. While tying rewards to desired behaviors has long been a staple of habit formation, handwashing has proven difficult to stick.

The rewards worked. “If you bribe kids, handwashing rates shoot up,” says developmental economist Reshmaan Hussam of Harvard Business School. And even just making handwashing a pleasant, easy activity improved health: Children in households with thoughtfully designed soap dispensers experienced fewer illnesses than children in households without those tools, Hussam and colleagues report in a paper to appear in American Economic Journal: Applied Economics.

Significantly, good habits lingered even after researchers stopped giving out rewards. “The fact that they found persistence suggests to me that participants did form habits,” says Jen Labrecque, a social psychologist at Oklahoma State University in Stillwater who was not involved with the research.

The study involved 2,943 households in 105 villages in the state of West Bengal between August 2015 and March 2017. All participants had access to soap and water. Nearly 80 percent said they knew soap killed germs, but initially only 14 percent reported using soap before eating.

To objectively assess habits, Hussam’s team devised a way to monitor handwashing in the absence of observers — whose presence typically makes people behave better. In collaboration with the MIT Media Lab, the team built a soap dispenser with a hidden sensor that recorded whenever somebody used it.

A young boy in West Bengal, India, uses a soap dispenser with a built-in sensor. The dispenser, designed by the MIT Media Lab, can track when and how often members of a household use soap, providing useful feedback for users and researchers alike.Reshmaan Hussam

They then educated families on how to build good handwashing habits, such as establishing a trigger (dinner time) and a routine (handwashing right before meals). They also made the handwashing experience as simple and enjoyable as possible, such as by using scented soap and mounting the sensors where children could easily reach them. Researchers visited households every two weeks to collect data on children’s health and refill the dispensers.

Hussam’s team divided households into multiple groups. Some households received only a dispenser. Others received automated reports on their daily handwashing performance, a social incentive to gently prod routine activity. Still others got tickets each time somebody pressed the dispenser around dinner time — these tickets could be traded for toothbrushes, backpacks and other useful items. A control group received no dispensers.

In households that got no incentives, the team found that people used soap at dinnertime 36 percent of the time, one to four months after receiving a dispenser. Those who got automated reports used soap 45 percent of the time. And those earning tickets used soap 62 percent of the time.

Once rewards and feedback ceased, soap use abruptly plummeted. With little to lose, the researchers kept the sensors on. As months progressed, handwashing rates among households that had received incentives ticked slightly upward. Nine months after incentives ceased, households that had received tickets washed their hands 16 percentage points more than households that received dispensers only.

The team suspects that the return of cold and flu season reminded parents to use soap. Perhaps “when parents see that kids are sniffly or sneezing, that’s when they’re triggered to use the device,” Hussam speculates. Often, “habits are tied to specific cues.”  

This study shows the value of spending a limited pool of money up front versus spreading it more evenly across time, as is common in public health campaigns, says medical epidemiologist Stephen Luby of Stanford University. “I do see the value of front-loading habit adoption.”

Even children living in households with just a dispenser and no rewards had better health than children in households without a dispenser. Eight months after incentives ceased, children with soap dispensers in their households experienced 38 percent fewer days with diarrhea and 16 percent fewer days with respiratory infections than children without dispensers. Access to a well-designed dispenser also tracked to healthier height and weight for children.

For product designers hoping to steer people toward good habits, a valuable lesson emerges: “Think carefully about human-centered design,” Hussam says.