¹ÏÌïºÚÁÏ

This study evaluates two interventions for residential water conservation. Comparing households across an enforcement algorithm's cutoff using a regression discontinuity design, we find that automated irrigation violation warnings cause substantial water conservation but also shift some consumption from regulated to unregulated hours within the week. In contrast, we show using data from a randomized experiment with the same customers that normative Home Water Reports reduce water use by a much smaller amount, but that this social pressure is effective during all hours both before and after automating irrigation policy enforcement. Our findings highlight the merits of implementing multidimensional conservation programs.