Living up to expectations: estimating direct and indirect rebound effects for UK households

Panel: 1. Foundations of future energy policy

This is a peer-reviewed paper.

Authors:
Mona Chitnis, Sussex Energy Group, SPRU (Science Policy Research Unit), University of Sussex, United Kingdom
Steve Sorrell, Freeman Centre, University of Sussex, Falmer, United Kingdom

Abstract

‘Rebound effects’ are various economic responses to improved energy efficiency that reduce the anticipated energy and emission savings. Most studies focus solely upon the direct rebound effects that result from increased consumption of cheaper energy services – such as driving further in fuel-efficient car. But a full accounting of the impact of energy efficiency improvements must also include various indirect rebound effects. For example, savings on fuel bills may be put towards increased consumption of other goods and services that may be viewed as ‘embodying’ emissions from different stages of their global supply chains.

This study estimates the magnitude of both direct and indirect rebound effects following the adoption of energy efficiency measures by an average UK household. It identifies the relative contribution of income and substitution effects to these results, together with the relative contribution of 12 categories of goods and services. Rebound effects are measured in terms of GHG emissions, using data from a multi-regional input-output model. The method involves estimating a system of equations for UK household expenditure from annual time-series data for the period 1964-2012 and deriving estimates of the own- and cross-price elasticities of different categories of goods and services. Energy efficiency improvements are then simulated as reductions in the price of the relevant energy carrier (electricity, gas or road fuels) and the change in consumption and emissions are calculated. The results suggest that rebound effects may be as large as ~47% for measures affecting road fuel consumption and 58-61% for measures affecting electricity and gas consumption. The rebound effects for measures affecting road fuels are mostly indirect and mostly derived from income effects, while the rebound effects for measures affecting electricity and gas are mostly direct and mostly derived from substitution effects. The results demonstrate the importance of accounting for rebound effects in policy appraisals.