Bright Lines, Risk Beliefs, and Risk Avoidance: Evidence from a Randomized Intervention in Bangladesh

Abstract

We conducted a randomized controlled trial in rural Bangladesh to examine how household drinking water choices were aected by dierent methods of communication about risk from naturally occurring groundwater arsenic. Households in a first group of villages received information about arsenic in tubewell water in a bright-lines format that emphasizes that arsenic levels above the Bangladesh national standard is unsafe. In the remaining villages, we added a gradient message that emphasized how risks from exposure increase with arsenic levels. Unexpectedly, households receiving the gradient message were 9 percentage points less likely to take mitigating action by switching sources. This average masks significant heterogeneity in treatment eects across the arsenic distribution that does not appear to be explained by heterogeneity in households’ observed characteristics. The gradient message led to 52% more switching of water sources when the arsenic level was moderately unsafe, but 38% less switching at high arsenic levels, although only the latter dierence is significant at the five percent level. The dierences in behavior are at least partially explained by discrepancies in risk perception that