How Nudges Can Help Households Internalize Energy Saving Information

When Richard Thaler and Cass Sunstein’s award-winning book “Nudge: Improving Decisions About Health, Wealth, and Happiness” was published in 2008, a broader audience began paying attention to how “nudges” can improve daily decisions. A nudge is a policymaking tool that encourages socially beneficial decisions through implicit cues or positive reinforcement. However, changing personal behaviors via nudges may also have unwanted costs: They may be costly to administer or have a negative impact on social welfare. A recent working paper attempts to estimate the increase in social welfare resulting from an energy-saving program in upstate New York. The study’s authors, Hunt Allcott and Judd B. Kessler, find that the benefits of nudging are typically overestimated but show alternatives that could be used to increase potential benefits by being more selective about who is nudged.

Nudges are typically evaluated by how much an intervention changes people’s behaviors toward a preferred outcome. Often, this calculation does not account for other costs associated with that switch. This study explores the distribution of Home Energy Reports (HERs) delivered by a natural gas company to its consumers. HERs provide information for consumers to compare their usage with that of their neighbors, and give useful tips to help save energy. The private and social gains of this kind of policy are obvious: Smarter energy usage saves money and decreases the use of a finite and polluting resource. However, the true costs associated with such a program are not clear and may include the inconvenience of changing personal behaviors, the time and financial costs of taking actions that save energy (e.g. home insulation), or even the psychological costs associated with feeling uncomfortable or guilty when comparing oneself to neighbors.

Working with an energy company in upstate New York, the researchers performed a randomized controlled experiment in which half the clients received HERs and the other half did not receive any information detailing their energy usage. Later, the authors investigated HER-receiving participants’ willingness to pay (WTP) to continue the program, assuming that respondents would take into account the non-energy costs mentioned above in their responses. This paper presents the estimates of the effects of this nudge on:

• the total amount of energy used,
• the implementation costs of providing energy-use information, and
• externalities produced by the consumption of energy.

The authors also analyzed the price effect on energy by looking at how the decrease in energy demand following the nudge may increase prices for the firm, which still has to pay the same fixed costs with less revenue from customers’ energy bills.

The model shows that variation in energy consumption produced by the HERs does not significantly alter the total social welfare outcome. There is a tradeoff in terms of social welfare between those users who have an implicit WTP that exceeds the cost of nudging—which presumably includes the cost of the HER program and its potential effect on energy prices—and those who do not. The idea is that it is more useful to identify and nudge those who want to be nudged and costly to nudge people who are not going to alter their consumption patterns.

The researchers’ model predicts that 41 percent of clients would opt in to the HER program in response to their predicted WTP (assuming no switching cost). If so, the effects of receiving this kind of information would have generated nine times more total welfare than the current program, where many people are being nudged with HERs, whether or not they want to be nudged. However, people do not act as reasonably as the models predict. For example, the same company runs a similar program in Ohio but requires people to opt in to receive the HERs. In this program, only 1.5 percent of households opt in, showing that inertia and other barriers may prevent people from deciding to receive HERs—even if they would actually respond positively to the HERs in the event that they opted in.

Therefore, the overall result on welfare of this kind of program depends on the type of consumer being nudged. Machine-learning algorithms can help to identify and target people who might have higher WTPs, or stronger preferences for environmental behaviors. They can also target individuals who stand to save more energy once they receive HERs. Both cases would increase total welfare associated with the HER program.

Nudges can help people make better decisions on a daily basis. However, changes in behavior may be costly from a welfare perspective. Understanding which drivers generate an overall increase in social welfare can transform nudges into stronger and more influential policy tools.

Article Source: Allcott, Hunt, and Judd B. Kessler, “The Welfare Effects of Nudges: A Case Study of Energy Use Social Comparisons.” National Bureau of Economic Research, 2015.

Featured Photo: cc/(Greg Hirson)