Présentation de Anett John, CREST, ENSAE
Abstract : We test the effect of light-touch psychological interventions on water chlorination and related health and psychological outcomes using a randomized controlled trial among 3750 young women in rural Kenya. We randomly allocate participants to two light-touch psychological interventions: one targeting planning skills (executive functions); and one targeting present bias and impatience (time preferences). A third group received only information about the benefits of chlorination, and a pure control group received no intervention. Ten weeks after the interventions, the time preferences and executive function interventions led to significant 27 and 18 percent increases, respectively, in the share of households who chlorinated their drinking water, compared to the pure control group. This increase was accompanied by significant 25–26 percent (executive function) and 35–38 percent (time preferences) reductions in the number of child diarrhea episodes relative to both placebo and control. The time preferences intervention also significantly increased participants' savings per week by 26 percent. We further study the psychological channels through which effects occur. The executive function intervention did not affect lab measures of planning, and the time preferences intervention did not affect choices in lab discounting tasks. However, both interventions increased self-efficacy, i.e. beliefs about one's ability to achieve desired outcomes. We show effects are not driven by changes in information about the benefits of chlorination. Together, these results suggest that low self-efficacy may be a psychological barrier to health behavior.