Spreading the Word! Spillover Effects of a Randomized Normative Informational Campaign on Residential Water Conservation

This paper reports the direct and spillover effects of a campaign with a normative based report on residential water use in Colombia. The experimental design introduces an exogenous variation in the percentage of households receiving reports within a water utility, leaving households without reports but affected by the diffusion of information. The randomization is at two levels. First, assigning saturation percentages across water utilities and second, the treatment status across households conditional to that saturation. Findings show that accounting for spillovers leads to a higher impact of social norm campaigns than previous interventions. In addition, shorter periods between reports remind households to act pro-environmentally. Households receiving feedback every 30 days show a significant impact than every 60 in both treatment units. This paper emphasizes using social norm-based as an effective tool in promoting socially desirable pro-environmental behaviors and addressing spillovers to diffuse behavioral changes. In particular, in small urban settings with high social capital levels or during weather shocks that could affect water supply