Using Targeting to Optimize Program Design: Evidence from an Energy Conservation Experiment
Dyson School of Applied Economics and Management
Friday, March 12, 2021
On-line via Zoom
12:00 pm-1:15 pm
We investigate the potential for targeted treatment rules to improve the performance of a large-scale behavioral intervention to encourage energy conservation. We first verify that receiving personalized feedback on energy use causes households to reduce their electricity consumption on average, and we present evidence of treatment effect heterogeneity. We then use a policy learning approach to derive treatment rules based on observable household characteristics that exploit this heterogeneity to maximize the expected benefits of the intervention. Targeting using transparent and easily implemented treatment rules yields significant gains relative to the treatment assignment that was actually implemented.