Seth Blumsack (Penn State University)
Abstract. The dominant model for the pricing of residential electric service over roughly the past century has been one in which consumers face a time-invariant price, despite high temporal and spatial variability in the economic, environmental and reliability costs of electricity consumption. This talk will outline how power grid operations translate into prices and electric rates by way of introduction to a set of behavioral experiments conducted by Green Mountain Power (GMP), Vermont’s largest electric utility. GMP has been studying ways to leverage investments in advanced electricity delivery and metering (the “smart grid”) to facilitate time-differentiated pricing for electricity, or to provide reward-based mechanisms for reducing household electricity usage. This technology can also be utilized to increase consumer awareness of their own patterns of electricity consumption and the associated economic and environmental costs. While standard economic theory would predict that paying more to consume more electricity during peak demand periods (the “stick”) should be equivalent to getting paid to forego that same consumption (the “carrot”), some insights from psychology and behavioral economics suggest otherwise. The outcomes from this set of randomized control trials suggest that: (i) sticks can be more effective than carrots at changing electricity consumption behavior, but consumers like carrots better than they like sticks; (ii) for the stick to be more effective than the carrot, consumers need a smart and transparent system to tell them when they are about to get hit and whether they are dodging the blow; (iii) without such a smart and transparent information system, neither the carrot nor the stick are particularly more effective than periodic conservation requests by the utility. We also find some evidence against the notion that consumers can be “warmed up” to the stick by first being given the carrot. The nature of the GMP experiments and others currently underway in Vermont has produced rich data sets that could be utilized to investigate a number of spatial, temporal and behavioral questions related to energy consumption in small cities and rural areas.