This was created in partnership with Environment for Development .
We investigate whether a social information campaign aimed at reducing water use causes a spillover effect on the use of electricity. On average, water use decreased by 6 percent for a treatment group for whom we conducted a social information campaign on their use of water, compared with that of a control group. We identify a spillover effect on electricity use among households that had efficient use of water before the campaign. The effect is sizeable; this group has almost 9 percent lower use of electricity after the campaign compared with the control group. Other types of households also decrease their water use, without a spillover effect. We argue that these results are consistent with a model of cognitive dissonance where the efficient households infer information about electricity use from the water use information.