Working Paper: NBER ID: w20809
Authors: Shahzeen Z. Attari; Gautam Gowrisankaran; Troy Simpson; Sabine M. Marx
Abstract: There is limited evidence of behavioral changes resulting from electricity information feedback. Using a randomized control trial from a New York apartment building, we study long-term effects of information feedback from “Modlet” in-home devices, which provide near-real-time plug-level information. We find a 12–23% decrease in electricity use for treatment apartments, concentrated among individuals reporting higher willingness-to-pay for an energy monitoring system. Decrease in overall electricity use is similar among treatment apartments which received Modlets and those which declined Modlets, and does not specifically occur for outlets with Modlets. This decrease may be due to a Hawthorne or salience effect.
Keywords: electricity use; information feedback; behavioral change; field experiment
JEL Codes: D03; D12; L94; Q30; Q40; Q54
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| information feedback from in-home devices (L68) | electricity use (L94) |
| treatment apartments (R31) | electricity use (L94) |
| willingness to pay for energy monitoring systems (Q41) | behavioral changes (D91) |
| salience of information provided by modlets (D83) | behavioral changes (D91) |
| Hawthorne effects (C92) | behavioral changes (D91) |
| treatment group (C92) | behavioral changes (D91) |
| modlets (Y60) | electricity use at plug level (L94) |