Sustainability concepts that rest on the idea of resource- or energy-effici
ency improvements due to technological progress tend to overestimate the po
tential saving effects because they frequently ignore the behavioral respon
ses evoked by technological improvements. Efficiency improvements also affe
ct the demand for resources and energy, and often an increase in efficiency
by 1% will cause a reduction in resource use that is far below 1% or, some
times, it can even cause an increase in resource use. This phenomenon is co
mmonly labeled the rebound effect, which is well-known among energy economi
sts, but never attracted much attention in ecological economics. The paper
starts with the traditional neoclassical analysis of the rebound effect in
a partial equilibrium framework that concentrates on the demand of one part
icular energy service such as mobility or room temperature. II also provide
s an overview of some of the main empirical studies based on this model tha
t mostly confirm the existence of the rebound effect, but are controversial
about its actual importance. However, we have to go beyond the neoclassica
l single-service model in order to take care of the variety of possible fee
dback affecting energy use. The paper presents two important expansions of
the single-service model in order to show the potential relevance of the re
bound effect to ecological economics. First, it is shown that in a multi-se
rvices model it proves to be difficult to make general statements about the
relevance of the rebound effect. In this case, the overall effect of an in
crease in energy efficiency on total energy use depends on the on the assum
ptions about the substitutability between the services considered and the d
irection of the income effect. Second, the paper also tries to take care of
the fact that changes in resource use or energy use are frequently just 's
ide-effects' of other forms of technological progress. Especially technolog
ical change of a lime-saving nature can have a large influence on energy us
e as many time-saving devices (for example, faster modes of transport) requ
ire an increase in energy consumption that is frequently reinforced by a 'r
ebound effect with respect to time'. This effect will be especially strong
when wages are high and, at the same time, energy prices are low, as is cur
rently the case in most industrialized countries. Consequently, the paper a
lso provides a strong argument for the introduction of energy taxes. (C) 20
01 Elsevier Science B.V. All rights reserved.