THE IMPACT OF END-USE EFFICIENCY CORRELATION ON INDIRECT REBOUND FOR HOUSEHOLDS
Michael Blackhurst, The University of Texas at Austin, 512-471-8616,
Overview
The effectiveness of energy efficiency in the residential sector has been the subject of considerable debate. While some researchers and policy experts claim efficiency is a least cost alternative to more sustainable energy provisions, others emphasize that the efficiency increases consumer surplus, thereby increasing demand for energy services. Such behaviour is often termed the “rebound effect.” Most research on rebound focuses on the direct rebound of single energy service, often characterizing rebound as the price elasticity of demand. However, consumers are faced with an increasing array of efficiency choices, particularly for electricity end uses. These trends suggest that efficiency choice correlation may influence indirect rebound. In this research, I derive an expression for the direct and indirect rebound for two energy services: electricity and personal vehicle transportation. These end uses constitute of a majority of energy demands for typical U.S. households. Using typical prices of and budgets for electricity and gasoline, results demonstrate the potential influence of efficiency choice correlation on indirect rebound is significant and can lead to high indirect rebound effects or promote increased energy savings.
Methods
Consider a purely technical expression of the work done by electricity, C, and transportation, T. The combined energy use is E = C/C+T/T, where C and T represent the efficiency of electricity and transportation, respectively. I determine the elasticity of total energy use with respect to a change in electricity efficiency, c(E), by differentiating E and multiplying by C/C:
(details withheld) Equation 1
Substituting expenditure and pricing for work terms (C/εC=wC/pC; T/εT=wT/pT) and εc in second term (εc=C pC/ wC) results in the following expression (with relative changes expressed as elasticity’s represented by the symbol ‘’.
Equation 2
Equation 2 assumes that the elasticity of electricity work with respect to electricity efficiency approximates the price elasticity (c(C) ~ - Pc(C)), which is common in rebound research. I further assume that changes in transportation services with respect to electricity efficiency approximates the income elasticity of gasoline ((c (T) ~ I(T)), which seems reasonable given there are not ready substitutes between electricity and gasoline for most residential services.
The first term in Equation 2 includes a direct rebound component weighted by relative prices and budgets. The second term is the indirect rebound, which includes re-spending effect pc(T) and the correlation between efficiency choices for electricity devices and vehicles. As re-spending increases (pc(T) increases), the indirect rebound increases. If efficiency choices are correlated (T(εc) > 0), indirect rebound decreases. If efficiency choices are uncorrelated (T(εc) < 0), rebound increases.
Using government and literature publications, Table 1 characterizes typical ranges in price elastisty's, energy prices (pC/pT), and household expenditures (wC/wT). By varying T(εc) from - 0.5 to + 0.5, I then use sensitivity analysis to determine the influence of efficiency choice correlation relative to ranges in other rebound parameters.
Results
Base case results – which assume no correlation between efficiency choices or T(εc) = 0 – indicate that the indirect rebound effect is relatively small at 0.06, which is exclusively driven by re-spending electricity savings on transportation services. However, the two-way sensitivity analysis (in “tornado chart” format) shown below indicates that indirect rebound is most sensitive to correlations between choices in the efficiency of electricity and transportation end-uses. These sensitivities hold true across reasonable ranges in and combinations of inputs. Results suggest that positively correlated efficiency choices can even reverse rebound (indirect rebound < 0).
Table 1: Modelling ranges for indirect rebound analysis
Rebound Input / Base case / Min / MaxPc(C) / -0.5 / -0.05 / -0.8
I(G) / 0.1 / 0.02 / 0.2
pC/pT / 1.50 / 0.5 / 2.5
wC/wT / 0.7 / 0.5 / 3
T(εc) / 0 / -0.5 / 0.5
Figure 1: Two-way sensitivity analysis of indirect rebound of transportation from change in electricity efficiency.
Conclusions
Despite the apparent influence of efficiency choice correlation on long-term effectiveness, very little is known about how householders make efficiency choices across different energy end uses. On one extreme, some homeowners may consistently choose more efficient technologies independent of incentives, reducing the potential for indirect rebound within energy services. Others may may selecting a few efficient technologies to compensate for inefficient technologies or not consider efficiency at all. It is also possible that participation in an efficiency program may influence future efficiency choices, spurring additional efficiency benefits. Result shown here suggest these behavioural mechanisms are critical, but there does not appear to any research addressing these issues.
Similarly, the long-term transportation implications of more efficient housing stock are uncertain. Investment in residential efficiency has recently been relatively aggressive. While it is unclear how householders will re-spend these savings, an increase in transportation spending is feasible given transportation accounts for about 15% of typical household spending. This may mean that, for example, more efficient suburban homes are more sensitive to indirect rebound in transpiration than urbanites, who typically have lower transportation service needs and more transportation options.
These results suggest that more complete householder technology choice information is needed to better understand the long-term indirect rebound associated residential efficiency, as even slight efficiency choice correlations appear to significantly influence indirect rebound.