Spreading the word! Direct and spillover effects of a normative message on water conservation

Individuals care about what others do or think of what should be done. Provide information in this regard has proven effective in promoting pro-environmental behaviors in water, energy, and green product consumption. However, the presence of spillover effects in previous field experiments has lead to biased or lower-bound estimators. This paper reports the results of a randomized field experiment designed to produce and measure the spillover effects of a normative informational campaign on residential water use in Colombia. Randomization was at two levels. First, I randomly assign a level of saturation among villages. Saturation refers to the percentage of the households that will be exposed to treatment directly; the rest are the spillovers. Second, I randomly assign the treatment status according to this saturation. Results show that accounting for spillovers leads to higher effects than in previous interventions. On average, directly exposed households reduce water consumption by 7.5% and spillovers by 4.5%. Spatial proximity to others directly treated in the same street reinforces the effects on the directly exposed. Overall the reactions of the spillover households could be explained by communication with other family members and interaction in public places. This paper remarks on the importance of addressing spillovers in normative informational interventions. Using them accurately could help to diffuse desirable or to avert bad behaviors in environmental concerns at a lower cost.

Work in progress

Early Research

Follow me


T: +57 (1)  3394949 Ext. 3407

  • LinkedIn - White Circle
  • Twitter Clean