Nudges are all the rage in behavioral economics and public policy applications around the world because of their potential for doing good at little or no cost. Economists, including Kip Viscusi (2018), have begun examining previously unexplored sides of nudges including some involving energy consumption. An often employed nudge in the area of conserving energy and reducing pollution is the home energy report, where after a home energy survey a household receives a message, usually monthly, that compares its energy use to that of neighbors and suggests ways to reduce electric or gas use and, in turn, its carbon footprint. Each month the consuming unit can also see how its energy use compares to its own past usage too. The subtleties of the benefits of home energy surveys and home energy reports is the subject of a recent article, "Differential and Distributional Effects of Energy Efficiency Surveys: Evidence from Electricity Consumption," available open access at the Journal of Benefit-Cost Analysis.
Nobel Laureate in Economic Sciences, Sir Angus Deaton (2018), has emphasized the importance of moving beyond only the mean differences revealed in the “gold standard” random controlled trials and recommended other statistical setups to consider effect heterogeneity including intervention distribution effects, such as how the treatment effect differs by predicted outcome quantile. The particular intervention my co-author, Dr. Galib Rustamov, and I investigated involves about 4200 customers of an Investor Owned Utility in California who voluntarily participated in a Home Energy Effectiveness Survey in January 2009 and January 2010. In addition to energy use the survey contained more than 130 questions, including household size, income, house size, and energy using/saving behavior and appliances. The comparison group contained customers who did not participate in the surveys then but would voluntarily participate in subsequent years. Our estimation procedures included quantile difference-in-differences propensity score matching regressions of two years of monthly energy consumption.
Our best estimate is that one year after beginning participation, participating customers reduced their monthly electricity consumption by about 7% on average, when compared to households who did not participate in the survey but would do so in a later program. Because of the low one-time cost of the survey (about $12 per customer), compared to an on-going annual benefit of $144 per customer in electricity savings, the nudge we studied passes a simple benefit-cost test.
Moreover, nudges are often used to deal with so-called internalities whereby they help people make better decisions as judged by their own self-interest (in this case, reduced electric bills). However, our results also show reduced externalities, which make a benefit-cost calculation more striking. Based on our estimates, the program participants for the year 2009 reduced monthly emissions by over 1500 metric tons of carbon dioxide, which lowered the social cost of electricity production by about $32,000 per month.
Finally, we examined the differential energy use effects of the Home Energy Efficiency Survey nudge by income and pre-nudge energy use. Our results showed that the program studied led to proportionately more electricity saving among low income/electricity use households than among high-use households.
Issues that we did not address but would be good topics for future research concern finding more of the details of how the low income households used proportionately less energy and taking a deeper dive into why high use/income households who may have the most slack in energy use to adjust and who are the most highly educated did proportionately less to reduce energy use.
Disclosure: In addition to his position as Adjunct Assistant Professor at the Claremont Graduate University, article co-author, Dr. Galib Rustamov, is currently an employee of Southern California Edison, where he is an Advisor for Measurement and Evaluation. The article was completed before he began his position there.
References
- Deaton, A and Cartwright, N., “Understanding and Misunderstanding Randomized Controlled Trials,” Social Science & Medicine, 210 (2018): 2-21.
- Viscusi, W.K., “Efficiency Criteria for Nudges and Norms,” Vanderbilt Law School, October 8, 2018.