Information provision has also been shown to affect residential energy use and appliance choice in other contexts. Gilbert and Graff Zivin (2014) find that household electricity bills provide information that can affect consumption patterns. In a study using household-level interval billing data, they find that households reduce average daily electricity consumption by 0.6 to 1 percent in the first week after receiving a bill, an effect that evaporates as salience fades and consumers tend to revert to higher consumption patterns. Jessoe and Rapson (2014) find that an in-home energy display that provides information on real-time electricity consumption, electricity price, estimated monthly usage, and bills increases responsiveness to short-run price fluctuations. Over time, households that have these in-home energy displays tend to exhibit conservation behaviors beyond periods of high prices. Jessoe and Rapson suggest this additional effect contributes a 1 to 2 percent reduction in CO2 emissions. Davis and Metcalf (2015) use a stated choice experiment to analyze the effects of tailoring information provision in Energy Guide labels to local weather and electricity price conditions on participants’ choice of room air conditioners. They find that providing more location-specific information results in more efficient appliance choices. Although they do not have actual energy consumption data, the authors are able to construct typical energy consumption profiles for room air conditioner options offered in their experiment, and they find that the implied annual energy savings from a more informative label average about $2.14 per treated respondent. In another stated choice experiment focused on hot water heaters, Newell and Siikamaki (2014) find that augmenting Energy Guide information with ENERGY STAR labels or energy efficiency ratings increases the uptake of more energy-efficient hot water heaters.22 Studies of commercial building energy use are limited, and no study, to our knowledge, has examined the role of information on energy use in commercial buildings.
Information failures may help explain the so-called “energy efficiency gap” in commercial buildings, which account for approximately 20 percent of annual US energy consumption and CO2 emissions. Building owners may not fully comprehend what influences energy use in their buildings and may have difficulty credibly communicating building energy performance to prospective tenants and buyers. Ten US cities and one county have addressed this problem by passing energy benchmarking and disclosure laws. The laws require commercial buildings to report their annual energy use to the government. We evaluate whether the laws have had an effect on utility expenditures in office buildings covered by the laws in four of the early adopting cities—Austin, New York, San Francisco, and Seattle—and find that they have reduced utility expenditures by about 3 percent. Our view is that these estimated effects in the early days of the programs are largely attributable to increased attentiveness to energy use.
Table 7 [below] shows the calculated percentage change in utility expenditures per square foot, along with the estimated coefficients from Table 6, and the average utility expenditures per square foot for all four cities. While the coefficients for New York, San Francisco and Seattle are close to one another in size, the estimated percentage effects differ, with the most pronounced difference being between Seattle and the other two cities. [The authors] calculate a 3.3 and 3.5 percent drop in utility expenditures per square foot from disclosure laws in New York and San Francisco, respectively, and a 5.2 percent drop in Seattle. This result is attributable to the relatively low utility expenditures per square foot in Seattle, which in turn is attributable in large part to low electricity prices. During the treatment period, the average electricity price in Seattle was $50.75/MWh, in San Francisco $114.84/MWh, and in New York $148.51/MWh. This highlights a potentially interesting result—that the laws may have a larger percentage impact on utility bills in cities with relatively low average electricity prices and thus low average bills. It also highlights an important aspect of our analysis—that we are focused on the effects of disclosure laws on utility expenditures and not energy use. Ultimately, we are interested in the effects of the policy on energy use and CO2 emissions, but data on energy use are not available from our data source. The translation from energy savings to CO2 emissions reductions will also vary by region, but an analysis of these effects is beyond the scope of this paper.
by Karen L. Palmer and Margaret A. Walls
Resources For the Future (RFF) www.RFF.orgRFF Discussion Paper 15-12 | April 2015