The Lightbulb Paradox: Using Behavioral Economics for Policy Evaluation

In an e¤ort to reduce energy use, governments in the United States and many other countries subsidize energy e¢ cient lightbulbs or ban traditional incandescent bulbs. We evaluate one important rationale for these policies: that imperfect information or attentional biases cause consumers to purchase more incandescents than they would in their private optima. We formalize a model of consumer choice that allows for such biases and develop formulas that allow optimal policy to be approximated using su¢ cient statistics from simple randomized control trials (RCTs). We then present results from two RCTs that provide information about and draw attention to the energy costs of di¤erent lightbulbs. The treatment e¤ects could provide support for subsidies on the order of those currently in place in parts of the U.S. However, there is no evidence to suggest that a ban on traditional incandescent lightbulbs increases welfare. JEL Codes: D03, D12, H21, H31, L94, Q41, Q48. Keywords: Behavioral public economics, energy e¢ ciency, …eld experiments, internality taxes, lightbulbs, optimal policy, Time-Sharing Experiments for the Social Sciences. — — — — — — — — — — — — — — — — — — — — — — — — — — — — We are grateful to colleagues and seminar audiences for helpful feedback. We thank our research sta¤ Jeremiah Hair, Nina Yang, and Ti¤any Yee as well as management at the partner company, for their work on the in-store …eld experiment. Thanks to Stefan Subias, Benjamin DiPaola, and Poom Nukulkij at GfK for their work on the TESS …eld experiment. We are grateful to the National Science Foundation and the Sloan Foundation for …nancial support. Allcott: New York University and NBER. Taubinsky: Harvard University.