From Water Cops to Smart Meters: an Experiment on Water Conservation Policy in the New Era of Automated Enforcement

Ludovica Gazzè, University of Chicago

Globally environmental regulations are generally enforced through costly manual inspections that result in detecting only a small fraction of violations. However, recent advances in monitoring technology and analysis of big data are poised to transform environmental enforcement. We run a randomized field experiment in a major US city among 80,000 households designed around the recent adoption of real-time smart water meters that enabled perfect detection of outdoor watering violations - replacing a system of visual inspections that identified under 1% of all violations. The households are randomized into 12 treatment groups defined by a) automatic detection through the smart meters (relative to visual detection), b) fine size, and c) consumption level that defines a violation. Consistent with the Becker rational model of crime, we find that the automatic violation detection treatment decreases water use by 3.1% and doubles city service requests for help with conservation. Additionally, the automatic detection and consumption violation level are more effective at increasing conservation than fine levels.