the energy-economy-environment … 1995 ecn-i--94-053 the energy-economy-environment interaction and...

MAY 1995 ECN-I--94-053

THE ENERGY-ECONOMY-ENVIRONMENTINTERACTION AND THE REBOUND-EFFECT

A.P.A. MUSTERS

BIDOCPostbus 11755 ZG Petten

The Netherlands Energy Research Fo~~ndation IECNis the leading institute in the Netherlands for energyresearch. ECN carries out basic and applied researchin the fields of nuclear energy, fossil fuels, renewableenergy sources, policy studies, environmental aspectsof energy supply and the development and applicationof new materials.

IECN employs more than 900 staff. Contracts areobtained from the government and from national andforeign organizations anti industries.

ECN’s research results are published in a number ofreport series, each series serving a different public,from contractors to the international scientific world.

The l-series is for internal reports that contain resultsmainly of interest for (fellow) colleagues, principallywithin ECN but also for those employed elsewhere.Although these reports may hot be in their final form,they may be referred to.

Het Energieonderzoek Centrum Nederland (ECN) ishet centrale instituut voor onderzoek op energiegebied in Nederland. ECN verricht fundamenteel entoegepast onderzoek op het gebied van kernenergiefossiele-energiedragers, duurzame energie, beleids-studies, milieuaspecten van de energievoorziening erde ontwikkeling en toepassing van nieuwe materialen

Bij ECN zijn ruim 900 medewerkers werkzaam. D~opdrachten worden verkregen van de overheid en vaïorganisaties en industrieën uit binnen- en buitenland

De resultaten van het ECN-onderzoek worden heer.gelegd in diverse rapportenseries, bestemd voor verschillende doelgroepen, van opdrachtgevers tot d~internationale wetenschappelijke wereld.

De l-serie is de serie interne rapporten die resultaterbevat bestemd voor de ECN-collega’s, maar ook voo~vakcollega’s die elders werkzaam zijn. Deze rapporterhebben een minder definitief karakter, maar zijn werefereerbaar.

Netherlands Energy Research Foundation ECNP.O. Box 1NL-1755ZG Pettenthe NetherlandsTelephone : ÷31 2246 49 49

: ÷31 2246 44 80

This report is available on remittance of Dfl. 35 to:ECI~, General Services,Petten, the NetheriandsPostbank account No. 3977703.Please quote the report number.

Energieonderzoek Centrum NederlandPostbus !1755 ZG PettenTelefoon : (02246) 49 49Fax : (02246) 44 80

Dit rapport is te verkrijgen door het overmaken varf 35,-- op girorekening 3977703 ten name vanECN, Algemene Dienstente Pettenonder vermelding van het rapportnummer.

© Netherlands Energy Research Foundation ECN © Energieonderzoek Centrum Nederland

Frarnework of the study

In order to obtain the M.Sc. degree in Industr[al Engineering anti Manage-ment Science from the Eindhoven ~Jniversity of Techno]ogy, a researchproject of 9 months must be carried out. This report presents the results ofthis research project which was ¢onducted at the unit Policy Studies of theNetherlands Energy Research Foundation. [ am indebted to many peoplefor being abIe to accomp)[sh th[s M.Sc. thesis. I wou]d however like to ack-now]edge the particu]ar contributions made by Ir. Maarten Splinter and Dr.Cees Withagen of the Eindhoven University of Technology. [n addition, [am very gratefu[ to Ir. Piet Boonekamp, Drs. Nico van der Linden, and Ir.Tom Kram of the Nether]ands Energy Research Foundation. Their com-ments and encouragements were unva~uab]e.

Abstract

This study examines the Energy-Economy-Environment (3-E) interacfionin general and the rebound-effect in particular. The rebound-effect can bedefined as that part of the initially expected energy savings, resulting fromenergy efficiency improvements, that is lost because of the 3-E interaction.To estimate the magnitude of this rebound-effect, the MARKAL-MACROmodel bas heen used. MARKAL-MACRO fs an integrated computer model,which is created by coupling the systems engineering model MARKAL antithe macroeconomic growth model MACRO.

2ECN-I--94-053

CONTENTS

SUMMARY

INTRODUCTION

5

7

THE MYTH OF ECONOMIC GROWTH1. ] lntroduction 91.2 Economic growth as a solution to all economic problems 91.3 Ancient theories of economic growth 101.4 Modern theories of economic growth 111.5 The concept of growthmania 13

THE ENERGY-ECOblOMY-ENVIRONMENTINTERACTION 152.1 lntroduction 152.2 A paradigm shift in economics 152.3 History of the energy-economy-environment interaction172.4 Human history is energy-related 202.5 An energy theory of value 23

THE ENTROPY LAW AND THE ECONOMIC PROCESS 273.1 Introduction 273.2 Economics and mechanics 273.3 The entropy law 283.4 Entropy and order 303.5 A finite amount of low entropy 323.6 Limits to substitution 343.7 No analogy between economics and thermodynamics :36

SOLUTIONS TO THE PROBLF_M OF ECONOMICGROWTH IN A FINITE ENVIRONMENT 394.1 lntroduction 394.2 The solution of the World Commission on Environment and

Development 394.3 The solution of Herman Daly 434.4 The solution of Nicholas Georgescu-Roegen 46

THE REBOUND-EFFECT: AN INTRODUCTION5.1 Introduction5.2 The thoughts of W. Stanley Jevons5.3 The Energy Poficy polemic5.4 The effectiveness of energy efficiency improvements:

a case study

THE REBOUND-EFFECT AND THE CONSUMER6.1 Introduction6.2 Mechanical determination of energy efficiency6.3 The substitution and income effect6.4 A simple consumer behaviour model

5151

54

63

71

717273

3

The 3-E interaction and the rebound-effect

7o QUANTIFICATION OF THE REBOUND-EFFECT WlTHMARKAL-MACRO 817.1 lntroduction 817.2 Description of the MARKAL-MACRO model 81

7.2.1 The MARKAL model 817.2.2 The MACRO model 837.2.3 The limitations of the stand-alone versions of

MARKAL and MACRO 837.2.4 The MARKAL-MACRO model 867.2.5 The basic re~ations in MARKAL-MACRO 87

7.3 Calculation of the rebound-effect 88

8. THE REBOUND-EFFECT IN PERSPECTIVE8.1 lntroductlon8.2 The fallacy oi: the fifth fuel concept8.3 How can the rebound-effect be reduced?8.4 The rebound-effect in a different context

9. CONCLUSIONS AND RECOMMENDATIONS

REFERENCES AND LITERATURE

95959598100

103

105

4 ECN-I--94-053

SUMMARY

Economie growth bas become one of the most important economie topicsof the post-war years. The reason for this is the fact that many argue thateconomie growth offers a solution to all sorts of economie and social ills.The thought that economic growth is necessary for solving the problems ofunemployment, poverty, and environmental degradation is hot new. AdamSmith, for example, considered economic growth as a desirable activity. Inthe long run, however, the classical school of economie thought arguedthat the wheels of progress would stop tuming. Modern neoclassicalscholars are more optimistic in their outlook on the future. The reason forthis positivism is the inclusion of substitution possibilities anti technicalchange in their analysis of economic growth.

The problem with unlimited economie growth, however, is that it places anexcessive demand on out natural environment. Phenomena such as thegreenhouse effect and the depletion of the ozone layer are warnings thatman needs to pay greater attention to his physical surroundings. In order todeal with these phenomena, a paradigm shif~ in economics is required. Thisnew paradigm should explicitly treat the energy system, the economy, andthe environment integrative.

The acknowledgement of the energy-economy-environment (3-E) interac-tion leads to the notion that there are limits to the expansion of the econ-omie system. The entropy concept can be used to clarify this finiteness.Any organism namely reeds on low entropy resources, which it takes fromthe environment and degrades into high entropy waste. Also the economiesystem feeds on low entropy. The three sources of low entropy resourcesthat mankind bas at its disposal are the energy received from the sun, thefossil fuel reserves, and the materials stored in the earth. And since lowentropy cannot be created anti can only be used once, the possibilities forunlimited economic expansion become questionable.

As a solution to the problem of economic growth in a finite environment,three different views can be distinguished. A rather optimistic view is givenby the writers of the Brundtland report, who advocate a sustainable de-velopment. The concept of sustainable development implies a forcefuleconomic growth that does however take into account the environment.Herman Daly is less optimistic. Daly advocates a steady-state economy, inwhich there is no place for economic expansion. Nicholas Georgescu-Roegen is of opinion that even a stationary state is hot the solution.Georgescu-Roegen argues that the growth trend must be reversed into anegative economic growth.

Of these three solutions, Brundtland’s call for sustainable developmentobviously offers the most appealing future. The writers of Our CommonFuture claim that continued economic growth remains possible when it isaccompanied by efficiency improvements in the use of energy and mate-rials. The underlying thought behind this approach is that energy efficiencyimprovements make it possible to have increasing economie growth rateswithout increasing demands for energy. There is however a problem with

5


respect to this approach. It is nemely hOt certain thet energy efficiencyimprovements lead to proportionate decreases in the demand for energy.The rebound-effect is a ineens of quantffying th[s effect. The rebound-effectís a measure of the part of the [nitially expected energy savings, resultingffom energy efficiency improvements, that is lost because of the 3-E inter-act[om

The rebound-effect is hot new. A]ready in ]865, W.S. Jevons wro~e that amore efficient use of energy is hOt equivalent toa diminished energy con-sumption. But despite these wise words, there is sti|l no agreement on thequestion of whether or nota rebound-effect exists. As an example of thisdisagreement, the polemic in Energy Policy can be mentioned. An impor-tant reason for the lack of consensus is that up to now, very few attemptsare undertaken to determine the magnitude of the rebound-effect. The onlyway to resolve the debate on the rebound-effect is by ineens of a quan-titative analysis.

An est[mation of the effect of energy efficiency improvements on the de-mand for energy can be obtained with the e[d of the MARKAL-MACROcomputer model. MARKAL-MACRO [s created by link[ng the MARKALmodel and the MACRO model. 24ARKAL is a techno}ogically oriented l/nearprogramming model of the energy sector. MACRO is a neoc~assical growthmodel, which takes an aggregate view of ]ong-term economie develop-ment. Because the interactions between the energy sector and the econ-omy are incorporated in MARKAL-MACRO, the rebound-effect can be de-termined with this integrated model. The MARKAL-MACRO estimation ofthe rebound-effect on the consumption-side of the economy amounts to33%. The existence of such a rebound-effect is in agreement with economictheory. The objective of ¢onsumers is namely hOt to mìnimize energy cost,but rather to maximize utiiity.

The existence of a rebound-effect obviously has policy implicagons. One ofthe most important impl[cations [s that energy efficiency improvementscannot be seen as a new energy supply source. This means that besidesoil, ¢oal, gas, anti uranium, energy efficiency cannot be regarded a~ thenew fifth fuel.

Because of the existence of the rebound-effect, apart of the expected ener-gy savings, resulting from energy efficiency improvements, is lost. There-fore, policy measures aimed at reducing the rebound-effect should be pur-sued. One way of reducing the magnitude of the rebound-effect is by takingenergy efficiency measures that go beyond the economie optimum. Theproblem with thìs approach however is that it leads to a disruption of theeconomy.

6 ECN-1--94-052

INTRODUCTION

In bis book The Structure of Scientific Revolutions, Thomas Kuhn (1970)argues that the evolution of every mature scientific discipline is charac-terized by a pattern of long periods of normal science and short periods ofscientific revolutions. But before the first period of normal science emerges,a situation exists in which different schools of thought compete for domi-nance. Each school has its own foundation and uses its own body of con-cepts anti techniques. This uncertain period which precedes the first periodof normal science ends when one school makes that much progress as tobecome almost generally accepted. To illustrate this, Kuhn uses the historyof the theory of physical optics. Although today’s physics students are alltaught the same theory which states that light exhibits properties of bothwaves and particles, this has hot always been the case. From the beginningof time until the end of the 17th century, several theories about the natureof light namely existed side by side. It was only in the beginning of the 18thcentury that Newton’s Opticks ended this long period of controversy andformed the basis for the first period of normal science on the subject ofphysical optics.

According to Kuhn, science means research that is characterized by onecommon sc~entific approach. This body of accepted theory supplies thefoundation for further research and is portrayed in textbooks. These text-books contain the theory, applications of the theory, and experiments thatconfirm the validity of the theory. The generally accepted theory of a cer-tain scientific community is called a paradigm by Thomas Kuhn. A para-digm can be defined as a characteristic set of beliefs of preconceptions of agiven research tradition. It is composed of fout eleménts. These elementsare symbolic generalizations, common values, concrete problem-solvingtechniques, and a metaphysical component. The study of paradigms pre-pares a student for the membership of a certain scientific group. Paradigmsfocus the attention of scientists on particular subjects and they facilitatecommunication. Because all the members of a particular communityleamed the same rules and practices, there is agreement on fundamentalissues. The result of this is that a certain opposition to change exists insuch a scientific community. According to Kuhn, ’closely examined, whe-ther historically or in the contemporary laboratory, normal science seemsan attempt to force nature into the preformed and relatively inflexible boxthat the paradigm supplies. No part of the aim of normal science is to callforth new sorts of phenomena; indeed those that will not fit the box are hotseen at all. Nordo scientists normally aim to invent new theories, and theyare often intolerant of those invented by others, lnstead, normal-scientificresearch is directed to the articulation of those phenomena and theoriesthat the paradigm already supplies (1970, p. 24)’.

New and unsuspected phenomena however always take place. And whenthe existing paradigm cannot cope with a new phenomenon, a scientificrevolution begins. In this crisis situation, the rules for normal science areloosened in an attêmpt to fit the new phenomenon in some theoretical con-cept. Ultimately, every crisis is resolved in one of two ways. Either the oldparadigm proves tobe able to handle the phenomenon or a new theory

ECN-I--94-053 7


emerges as the next paradigm. The latter event is what Kuhn calls a para-digm shi~t. A¢cording to Kuhn, ’the transition from a paradigm in crisis toanew one from which a new tradition of normal science can emerge is farfrom a cumulative process, one ach]eyed by an articulation or extension ofthe old paradigm. Rather it is a reconstruction of the field from new fun-damentals, a reconstruction that changes some of the field’s toost elemen-tary theoretical generalizations as well as many of its paradigm methodsand applicatlons (1970, pp. 84-85)’.

Although Kuhn’s book is mainly concerned with revolutions in the naturalsciences, paradigm shìfts have a]so occurred in the economie science. Inthe course of time, the outlook of economics has namely changed frommercantilism to physiocracy, and from physiocracy to classlcal laissez-faire. This laissez-faire was sueceeded by Keynesianism, which in tutu wasfollowed by the neoclassical growth approach. This neoclassical growthparadigm bas heen the dominant economie theory of the post-wat years.

As a result of the economie growth of the past decades, the impact ofbutaan act[vity on the environment bas ~ncreased dramatically. The studyof this high]y comp]ex interaction between man and bis physlca] surroun-dings ca]]s foran integrative approacho Or as A]varo Umana put ir: ’sinceenvironmenta] problems do hot rail into any of the eompartments oforganized knowledge, scientists from many different discip]ines have had tostudy them joint]y, and they have pointed to the necessity of inter-disciplinary approaehes to deal with questions of thls kind. Thìs approachis needed beeause many prob]ems arise due to the interaction among dif-ferent parts of a system. For example, butaan economie activity bas asìgnificant impact in íhe global cycling of carbon, nitrogen, phosphorus,and su]phur. In most cases economie forces are an important determinantof the types and levels of pollution, as wel] as the uses to which the naturalenvironment is put (198], p. 29)’.

In order to allow for the interaction between the economy and the environ-ment, a new paradigm in the economie science is gradually emerging, Thisnew paradigm explicitly treats economie activity, energy use, and the natu-ral environment integrative. The present study on the energy-economy-environment (3-E) interaction ín genera] and the rebound-effectin particular is an exponent of th[s new paradigm in economics. Therebound-effect can be defined as that part of the initially expected energysavings, resuiting from energy efficiency ~mprovements, that is ]ost becauseof the 3-E interaction. The study begins with a description in chapter ] ofthe concept of economie growth. Chapter 2 outiines the energy-economy-environment interaction. In chapter 3, the re]ationship between economicsand thermodynamics is discussed. Chapter 4 contains three different viewson the problem of economie growth in a finite environment. In chapter 5,an introduction into the rebound-effect is given. Chapter 6 deais with therebound-effect on the eonsumption-side of the economy. In chapter 7, therebound-effect is quantified with the MARKAL-MACRO computer model.The rebound-effect is p]aced in perspective in chapter 8. Fina]]y, chapter 9presents the conc]usions and recommendations that can be drawn from thisstudy.

ECN-1--94-053

1. THE MYTH OF ECONOMIC GROWTH

Economie growth is the grand objective;it is the aim of economie policy as a whole.Sir Roy Harrod.

I. 1 lntroduction

In thia first chapter, the concept of economie growth is describedo In para-graph 1.2, the thought that economie growth can be seen as a solution toall economic problems is outlined. In paragraph 1.3, some ancient theoriesof economie growth are discussed. Some modern theories of economiegrowth are outlined in paragraph 1.4. Finally, paragraph 1.5 discusses theattitude ot= growthmania.

1.2 Economic growth as a solution to all economicproblems

Every period in the history of economie thought aan be associated with oneissue of outstanding importance. According to Jones (1975), the questìonof free trade versus protectionism was the most important economie topicin the middle of the 19th century. The issue of the gold standard and va-rious currency reform plans were the prominent economie topics at the endof the 19th century. In the beginning of the 20th century, the problem offree trade versus protectionism emerged again. The years between theWorld Wa~s were dominated by discussions concerning the big depression.In this respect, Jones argues that tbe concept of economie growth basbecome the most important economie topic in the years after the secondWorld War.

Economie growth, seen as an increase in Gross Domestic Product, basbecome a goal for economie and social policy and economie growth rateshave become measures of success for nations throughout the world. Thethought is that rapid economie growth offers a solution to all economie ills.It is, for instance, commonly urged that poverty can only be alleviated by arapid economie expansion that provides more tax revenues for welfareprograms. It is also urged that the problem of unemployment can be solvedby means of economie growth. It is even suggested that economie growthis necessary in order to have both rood and weapons.

The reason behind this gloritìeafion of the growth concept is the fact thateconomie growth postpones the discussion on an equitab]e dis’~ribution of"wealth among the population. According to Daly, ’growth takes the edgeoff distrlbutional ¢onflicts. If everyone’s absolute share of income is increa-sing there is a tendency hot to fight over re]ative shares, espe¢ia]ly sincesuch fights may interfere with growth and even ]ead toa lower absoluteshare for all (1973A, p. 22)’. Wa][ich agreed with Daly: ’growth is a substi-tute for equality of income. So ]ong as there is growth there is hope, and

9


that makes large income differentials tolerable (1972, p. 62)’. The conceptof economie growth thus enables man to recede into the backgroundtroversial issues such as the rediatribution of wealth anti income. Wìtheconomie growth, the wealth of everyone can increase. Without ir, the in-crease in wealth of one individual must be raken from another individual.Daly summarized this thought by stating that ’the way to have your cakeand eet it too is to make it grow (1973B, p. 151)’.

1.3 Ancient theories of economie growth

The thought that economie growth is the solution to all sorts of economieproblems is hot new. in 1776, An lnquiry info the Nature anti Causes of theWealth of Nations was published by the Scottish professor of morel philos-ophy and founder of the classical school of politieel economy Adam Smith.The classica[ school thought that economic growth was a desirable actlv~ty.Adam Smith claimed tt~at ’every individual is continually exerting himself tofind the toost advantageous employment for whatever capital he can com-mand (1776, p. 421)’. Because each indivídual was the best judge of bisown interest, he shou]d therefore be left free to pursue it in bis own way. Inthis way, each individual wouId hot onIy attain the best situation for him-self, but he would also serve the common good. This outcome was ob-tained when society was organized in harmony with an inherent naturelorder. According to Smith, human conduct was so carefully balanced thatthe advantage o~ one individual could not impede the advantage of all otherindividua~s. It was this belief in the naturel balance of human conduct,which led Smith to make his well-know statement regarding the invisiblehand: ’Each producer intends only bis own security; and by directing theindustry in such a manner as its produce may be of the greatest value, heintends only his own gain, and he is in this, as in many other cases, led byan invisible hand to promote an end whìch was no pari of his intention. Noris it always for the worse for society that it was no part of ir. By pursuinghis own interest he frequently promotes that of society more effeetually thatwhen he really intends to promote it (1776, p. 423)’. Unrestricted com-petition would thus lead to the best cond[tions for economic expans[on and,therefore, ultimately to the highest increase in the satisfaction of the wantsof aIl members of the community.

In the long run, however, the classical school of economie thought erguedthat there were limits to this continued increase in the common welfare.Adam Smith thought that profits tended to rail with the progress of society.The accumulation of wealth would lead to increasing competition amongcapitalists and this would ultimately reduce profits. In 1798, Thomas RobertMalthus published The Essay on the Principle of Pop~lation as it Affects thefuture Improvement of Society. According to Ma]thus, peop]e’s power ofproducing more rood was hot as great as their ability to reproduce themseloves. He explained this thought by arguing that a population tended tocrease in a geometrical progression (2,4,8,16,.), ~vhiie the supply of roodincreased only in arithmical progression (2,3,4,5,.0. The basis for thistheory of population was the law of diminishing returns. Aceording to Tur-got (1776), this law implied that a doubling of the capital invested in agri-culture would not doub]e the yleld. When an increased application of |abourtoa given piece of land began to produce a less than proportionete in-

] 0 ECH-I--94-053

The myth of economie growth

crease in yield after some time, more and poorer land would have toberaken into cultivat[on. This meant an increase in the d[ff[culty of providingrood for a growing population, and this would ultimately stop the wheels ofprogress turning. Another representative of the classical pol[t[cal economywas David Ricardo. He also drew a pessimistic picture of the future in hismost important book On the Principals of Political Economy and Taxation(1817). Ricardo believed that a progressive decline of the fertility of landwas an obstacle to an increasing wealth and population for the t:uture.

1.4 Modern theories of economic growth

Modern neoelassical theor[es of economic growth are more optimistic intheir outlook on the future. The reason for this positivism is the inclusion ofsubstitution possibilitles and technical change in their analysis of economicgrowth. It is often argued that there are no limits to economie growth be-cause it is a function of an unlimited supply of human ideas. According toJulian L. Simon (1981), for example, the problem of the scarcity ot: fossilfuels and other natural resources will not stop butaan progress, beeausethere is no natural resource problem. In í:act, the opposite is the case.Simon asserts that ’~he only scarcity of natural resources is in the sensethat it costs labour and capital to get them, though we would prefer to getthem for free (]981, p. 3)’. Simon argues that the cost ot: natural resourcesin human labour and their prices relat[ve to wages and other goods,suggest that natural resources have heen becoming less scarce over thelong run, right up to the present. If the past is any guide, Simon claims thatnatural resources will progressively become less scarce and less ¢ostly, andwill constitute a smaller proportion of our expenses in the future years.Therefore, Simon argues that special efforts need not be made to avoidusing these natural resources. Our present use is not at the expense offuture generations. Future generatìons will be faced by no greater e¢onomicscarcity than we are now, but instead will have just as large or larger sup-plies of resources to extract, despite out present use of them. Simon’s posi-tivism is based on the fact that the ultimate resource is human knowledge.A¢cording to Simon, the stock of know~edge is the main fuel to speedhuman progress: ’resources in their raw forto are useful and valuable onlywhen found, understood, gathered together, and harnessed for butaanneeds. The basic ingredient in this proeess, along with the raw elements, isbutaan knowledge. And we develop knowledge about how to use raw ele-ments for our benefit only in response to our needs. This includes know-ledge for finding new resources of raw materials such as eopper, for gro-wing new resources sueh as timber, and for finding new and better ways tosatisfy old needs such as successively using iron or aluminum or plastic inpIace of clay or copper. Such knowledge has a special property: It yieldsbenefit to people other than the ones who develop ir, apply ir, and try tocapture its benefit for themselves. Taken in the large, an increased need fornatural resources usually leaves us with a permanently greater capacity toget them, because we gain knowledge in the process. And there is nomeaningful physical limit to out capacity to keep growing forever (1981, p.346)’. According to Simon, even a growth in population wi~l probably havepositive consequences on the natural resource situation, because morepeople implies more knowledge being created.

]1


Similar thoughts were expressed by Judith Rees (1990). According toRees, resource scarcity is hot an obstaele to continued economie growth.Rees’ argumentation of this view is founded on three interrelated charac-teristics of the resource system. First, Rees argues that ’natural resourcesare product$ of the human mind; their limits ere hot physical, but are set bybutaan demands, institutions, imagination anti ingenuity (1990, p.424)’.Second, Rees asserts that ’the process of technological change is deeplyrooted in the economy, anti there ia little evidence of a slackening rare ofinnovation (1990, p. 424)’. Third, Rees argues that ’the economie systemin advanced nations has inherent adaptive mecharfisms to combat shorta-ges of specific resources when they are needed to sustain the process ofgro~rth and capital accumulafion. The potential for recycling and conser-vation, the development of new supply materials, and shifts in the outputmix of production makes the resource system highly dynamic (1990, p.425)’. Rees thus argues that the expanding economie system has huilt-inmechanisms to maintain the input flow of natural resources that is neces=sary for its ¢ontinuance.

A theory of natural resources bas thus evolved, which states that naturalresource scarcity cannot be a seriou$ long-term problem because the econ-omic system counteracts resource-related problems by substitutfon andtechnological change. Empirical justification for this thought seemed to beprovided by Bamett and Morse (1963). In Scarcity and Growth, one of themoìt important neoclas$ical analyses of resource scarcity, they argue thatincreasing natural resource scarcity would reveal itself in an ]ncreasingtrend of unit costs of these resources. Therefore, Bamett and Morse col-lected data on the unit costs of extractive products in the United Statesfrom 1870 to the end of the 1950s. These data show that the unit costs ofextractive goods have declined during the period under review. From this,Bamett and Morse conclude that natural resources have hot become morescarce ~:rom 1870 to 1959. According to Barnett and Morse, the reason forthe decline in unit costs is technological progress. They argue that ’thecumulation of knowledge and technological progress is automatie anti self-reproducing in modern economies and obeys a law.of [ncreasing retums.Every eost-reducing innovation opens up possibilities of’ application in somany new directions that the stock of knowledge, fat from being depletedby new developments, may even expand geometrically. Technologicalprogress, instead of being the adventitious consequence of lucky and high=ly improbable discoveries, appears to obey the principle that changes tendto induce further changes in the same direction (1963, p. 236)’. Barnettand Morse further argue that ’the process of growth thus generates an-tidotes to a general increase of" resource scarcity. Technological and manyother advances are causally integrated with the growth process. To theextent that there is an actual or anticipated tendency for energy costs toincrease in one industry or another, resource-saving technologies are intro-duced to avoid the cost increases (1963, p. 240)’.

Further support for this thought was given by Robert Solow (1973, 1974A,1974B). According to Solow, the prices of natural resources have hot risenover the past 50 years, compared with the prices of other things. From this,Solow draws the conclusion that ’there have sofar been counterweights toany progressive impoverishment of deposits (1973, p. 47)’. Solow arguesthat these counterweights are natural resource saving technical progress

12 E~N-I--94-053

The myth o~ economic growth

and the substitution of labour and capital for exhaustible resources. Accor-ding to Solow, there is in principle no natural resource problem: ’the worldcan, in efl:ect, get along without natural resources, so exhaustion is just anevent, nota catastrophe. At some finite cost, production can be ffeed ofdependence on exhaustible resources altogether (1974B, p. 11)’. Solowgrounds his optimistic outlook on the fact that ’in a simple aggregatemodel of a resource-using economy, one can prove that if the elasticity ofsubstitution between exhaustible resources and the other inputs is unity orbigger, and if the elasticity of output with respect to reproducible capitalexceeds the elasticity of output with respect to natural resources, then aconstant population can maintain a positive constant level of consumptionper head forever (1974B, p. 11)’.

1.5 The concept of growthmania

This positivism leads to an almost religious devotion to economic growth,which Mishan (1967) called growthmania. Growthmania is the attitude thatalways gives priority to growth. It is the attitude in economic theory thattakes the view that the technological possibilities to produce more andmore goods to satisfy people’s infinite wants are unlimited. Because theconcept ot: utility cannot be measured, the nation’s annual productionvolume bas become the indicator for economic well-being. In this context,Weisskopf remarked that ’there is hardly any disagreement among econo-mists, businessman, and politicians about the desirability of a growingGNP. There are differences about the details of national-income accountingand serious differences about the means of accomplishing economicgrowth. The desirability of overall growth for the individual and for societyis hardly ever questioned (1973, p. 241)’. An increasing GNP growth rarebas thus become a major policy goal. A decline in the rate ot: growth of theGross National Product is seen as a disaster.

Bertrand de Jouvenel (1959) called this addiction to unlimited growth laciuilisation de toujours plus. According to de Jouvenel, mankind nowadayslives in a society of more and more. Out institutions are becoming moreand more adapted to growth, and less suited to a stationary state. Outsocial system tests on the incessant development of new needs. If the everincreasing demand for goods and services would cease, Bertrand deJouvenel argues that the economy would collapse.

Max Weber (1958) called this attitude of more anti more the spirit ofcapita-lism. According to Weber, man is dominated by money-making, byacquisition as the ultimate purpose in lire. This continued accumulation ofwealth has become a goal for its own sake, rather than a means to an end.Economic growth is no longer used by man as a means for the satisfactionof his material needs. According to Weber, capitalism is identical with thepursuit of profit, and forever renewed profit. This profit making in a capita-listic system is essential, since in a capitalistic order of society, an in-dividual enterprise which does not take advantage of its opportunities forprofit-making will soon bankrupt. According to Weber, the drive towardsthe accumulation of wealth can be found in the Protestant ethic. Protestan-tism supplies the moral energy for the capitalistic entrepreneur in the econ-omic system to keep growing.

13

Although the call for economie growth is omnipresent in today’s society, inancient times the desirabi~ity oí an ever expanding eeonomy was ques-tioned. According to Weisskopf, ’the concept of growth reflects the value-attitude system of earIy capitalism be~ore and during the lndustria~ Revolu-tion. In distinction from prevìous societies where the pursuit of wealth andhard work were considered as inferior activities and as a curse, legt toslaves, women, anti inferior social groups, industrial society made the ac-quisition of wealth moral]y aeceptable and considered itas a moral obliga-tion. Economie thought justified this attitude by assuming thatacquisitiveness and the propensity to truck, barter, and exchange in orderto increase one’s wealth is a basic human propensity. Here, a unique his-torical phenomenon, the acquisitìve attitude, was interpreted as a universalbutaan inclination (1973, p. 242)’. Perhaps the attitude of more and morewas necessary in the beginning of the lndustrial Revolution, when the qua-lity o.~ lffe was rather low. Today, however, toost peop]e in the WesternWor]d are rather well-off and much of the economic growth is for luxuryitems, hot f’or necessities. Moreover, the accompanying excesslve demandthis growth places on the earth, |eads to phenomena such as deser-tificat~on, acidification, the extinction of plant and anima! species, and thegreafly accelerated exhaustion of our natural resources. Also controversia]topics sueh as the depletion of the atmospheric ozone layer and the green-house effect are re[ated to un(~mited economic expansion. Al[ these pheno-mena are wamings that man needs to pay greater attention to the impactof bis activities on the environment. Man can no longer ignore the interac-tion between the e¢onomy and the environment, and must stop the delu-sion that there are no [imits to the growth of the economic system,

14

2. THE ENERGY-ECONOMY-ENVIRONMENTINTERACTION

All fiesh is grass.Prophet lsaiah.

2.1 Introduction

At the end of the prevfous chapter, the des[rability of tak[ng into accountthe negative effects of economie activity en man’s physical surroundings isdiscussed. These effects are the result of the existence of a continuousinteraction between the energy system, the economy, anti tbe environment.This 3-E (energy-economy-environment) interaction is the subject of thischapter. The chapter begins in paragraph 2.2 with an outline of the para-digm shift in economics that is required for the integration of the energysystem, the economy, and the environment. In paragraph 2.3, the historyof the 3-E interaction is described. Paragraph 2.4 deals with the seholarswho argue that hùman history is energy-related. In paragraph 2.5, the ener-gy theory of value is discussed.

2.2 A paradigm shift in economics

$ince the beginning of this century, the re~ationship between the economyand the environment has fundamenta][y changed. At the beginning of thiscentury, man was unable te radically change his physical surroundings.Today, the impact of human activity en the environment is much greater.According te the Wor]d Commission en Environment and Development,’the scale of our interventions in nature is increasing and the physicalfects of out decisions cut across national ffontiers. The growth in economieinteraction between nations amp[ifies the wider consequenees of nationa]decisions. Many regions race risks of becoming irreversibly damaged, andthis threetens the basis for human progress (1987, p. 27)’.

Economics and the environment are thus becom[ng more and more inter-twined. Standard economic thought sometimes ignores this interaction. Theview of standard economics can be illustrated with the graph in which ina]most every introduetory textbook en economies the economie process isportrayed as a self-sustaining, circular flow between production and con-sumption. An important shortcoming of this model is that it represents theeconomy as a process that exists in eomp]ete iso]ation from the environ-ment. The economie proces~ however is net an isolated, self-sustainingproeess. The economy namely continuously alters the environment, andthe environment in turn continuously alters the economy, lllustrative in thiscontext are the words by which the French political philosopher Bertrandde Jouvenel characterized modern western man. According te de Jouvenel,’western man tends te count nothing as an expenditure, ether than humaneffort; he does net seem te mind how much mineral matter he wastes and,fat worse, how much living matter he destroys. He does not seem to realize

I5


at all that human lire is a dependent part of an ecosystem of many differentforms of lire. Because the world is ruled from towns where men are cut offfrom any form of lire other than human, the feeling ot: belonging to an eco-system is hot revived. This results in a harsh treatment of things uponwhich we ultimately depend, such as water and trees (1958, pp. 140-141)’.

By ignoring the environment, the circular flow model cannot determinewhether man’s physical surroundings can keep up with an expanding econ-omic system. The danger of this lies in the fact that the environment maybecome irreversibly afl:ected at a certain point in time. What is needed is amodel that will take info account the dependence of the economy on theenvironment and by doing so wilt protect us from the dangerous conse-quences of environmental degradation. In this model, the economy is loca-ted within the environment with which it can exchange energy and matter.

For the integration of economìcs in the environment, a paradigm shift inthe economic science is needed. Such a reorlentation is required accordingto Robert Costanza and Herman Daly, who state that ’to effect a true syn-thesis of" economics and ecology is the second toost important task of outgeneration, next to avoiding a nuclear war. Without such an integration wewill gradually despoil the capacity of the earth to support lit:e (1987, p. 7)’.The writers of the Brundtland report share the opinion of Costanza and Dalywhen they write that ’the common theme throughout the strategy for sus-tainable development is the need to integrate economic and ecologicaIconsiderations in decision making. They are, after all, integrated in theworkings of the real world. This will require a change in attitudes and ob-jectives and in institutional arrangements at every level (1987, p. 62)’.

A change in the orientation of economics is also required to make the tran-sition from the cowboy economy of the past to the spaceman economy ofthe future. In The Economics of the Coming Spaceship Earth, Kenneth Boul-ding (1973) pictures today’s economic system as the cowboy economy,the cowboy being symbolic of the illimitable plains and also associated withreckless, exploitative, and violent behaviour. Boulding ~:alls the economy ofthe future the spaceman economy, in which the earth has become a singlespaceship, without unIimited reservoirs for extraction and pollution.Boulding argues that the two types of economy can best be distinguishedwith respect to the value attached to consumption. According to Boulding,’in the cowboy economy, consumption is regarded as a good thing andproduction likewise; and the success of the economy is measured by theamount of throughput from the factors of production, apart of which, atany rare, is extracted from the reservoirs of raw materials and non-econ-omic objects, and another part of which is output into the reservoirs ofpollution (1973, p. 127)’. Now one could argue that if the sources andsinks of the earth were infinite, throughput may be used as a means ofquantifying the success of an economy. The Gross National Product is arough measure of this throughput. The earth’s surface is however limited.Therefore, Boulding argues that ’in the spaceman economy, throughput isby no means a desideratum, and is indeed to be regarded as something tobe minimized rather than maximized. The essential measure of the successof the economy is hot production and consumption at all, but the nature,extent, quality, and complexity of the total capital stock, including in thisthe state of the human bodies and minds included in the system. In the

16 ECN-I--g4-053

The energy-economy-environment interaction

spaceman economy, what we are primarily concemed with is stock main-tenance, and any technological change which results in the maintenance ofa given total stock with a lessened throughput (that is, less production andconsumption) is clearly a gain. This idea that both production and con-sumption are bad things rather than good things is very strange to econo-mists, who have been obsessed with the income-flow concepts to the ex-clusion, almost, of capitaI-stock concepts (1973, p. 127)’.

2.3 History of the energy-economy-environmentinteraction

Although standard economic thought sometimes tends to disregard theinteraction between man and the physica] surroundings on which he isdependent in al] his doings, in former days this interaction was hot ignored.In 1662, Sir Willam Petty, the earliest and most important English econo-mist who prepared the ground for the classica] system, wrote A Treatise ofTaxes and Contributions. In this work he said that ’Labour is the Father andactive princip]e of Wea]th, as Lands are the Mother (1662, vol i p. 68)’. Inthe 1750s there developed in France a body of economic theory to whichthe name physiocracy was given. The term physiocracy, derived fromGreek and litera]Iy meaning tule of" nature, was given to this stream ofeconomic thought after Dupont de Nemours’ publication Phys~ocratie ouConstltutions essentielles du Gouvernement le plus avantageux au GenreHumain (1761). The central point in the reasoning of the physiocrats wasthe search for the surplus which might be available for accumu]ation. Thissurplus was called the produit net. The physiocrats divided labour intoproductive and sterile labour. The former consisted only of ]abour whichwas capable of creating a surplus, i.e. something over and above thewealth which was consumed in order tobe capab]e of producing. AII other]abour was sterile. According to the physiocrats, the produit net was a sur-plus of tangible wea]th of physical goods. It was hot an abstract term usedas an exchange va]ue. Because of this, the physiocrats selected one sectorof the economy as the only rea]ly productive one. In agricu]ture, the dif-ference between goods produced and goods consumed was toost obvious.The produit net in agricu]ture namely was the difference between the yieldof a piece of Iand minus the rood consumed by the farmer and the amountof seed that was used. It was the most apparent kind of surplus. The phy-siocrats claimed that no surplus of value cou]d arise in exchange. [ndustrytoo created no va]ue, it only transformed objects. Agricu]ture alone cou]dyield a surplus. This surplus created by agricu]ture was the basis for Ques-nay’s famous Tableau Oeconomique (1758), in which the circulation of theproduit net between the different sociaI classes of society was described.According to the physiocrats, humans had no control over nature and thehighest wealth would be attained if economic behaviour was in accordancewith nature.

This re]iance on what is natural is also emphasized by the classical econ-omists. Classical scholars such as Smith, Ricardo, and Malthus believed inthe existence of an inherent natural order, which was superior to any man-made order. Adam Smith, for example, described the supremacy of thenatural order and the imperfections of human institutions in bis famous

ECN-i--94-053 17


Wealth of Nations in the following way: ’take away artificial preíerences andres~raints and the obvious and simple system of natural liberty will establishitself. Again that order of things which necessity imposes is promoted bythe natural inclinations of man. Human institutions too often thwart thesenatural inclinations (1776, p. 385)’. The ¢lassical school of economicthought claimed that society should be organized in harmony with thisnatural order in order to obtain the optimal conditions for a flourishingeconomic system.

In the 19th century, the physical basis of economics that was advocated bythe physiocrats and classical economists was raken a step further with thediscovery of the laws of thermodynamics. According to Gaggioli and Obert,’thermodynamics is the science which deals w[th energy and its transl:or-mations, and with the rela’donship between the properties of substances.The subject may also be ca]led physical chemistry by the chemist or heatby the physicist (1963, p. 13)’. Thermodynamics deals only with themacrostructure of matter and does not concern itseIf with events happeningat the molecular level. HistoricalIy, thermodynamics was developed beforean understanding of the internal structure of matter was achieved. Thescience of thermodynamics began with an analysis by the French engineerSadi Carnot, on the problem of how to build the best and toost efficientheat engine.

The way a steam engine ordinadly operates is that water is boiled by theheat from a fire, and the steam so formed expancls and pushes a piston,which makes a wheel go around. After that, one could complete the cycleby letting the steam escape into the air. The drawback ot: this approach isthat one has to keep supplying water. It ~s more e~ficient to condens thesteam, and pump the water back into the boiler, so that it circulates con-tinuously. Heat is thus supplied to the engine and converted into work. InReflections on the Motive Power of Heat and on Machines fitted to Developthat Power, Camot (]824) wondered what properties substances such aswater or alcohol must have to make the best possible engine. This questionto which Carnot addressed himself, constituted the beginning of thermo~dynamics.

The results of thermodynamics are ail contained in certain apparently sim-ple statements, called the laws of thermodynamics. The first law concemsthe conservation of energy. It states that energy can be changed in form,but it can neither be created nor destroyed. This means that if one forto ofenergy is changed to another forto, the same total quantity, expressed inenergy equivalents, remains after the transformation. This does hot meanthat also the quality of energy, i.e. the abil[ty to do work, remains thesame. Take for example a waterfall. The water of the Niagara rivet posses-ses potential energy. Potential energy is converted to kinetic energy when itis used to set a body in motion. So, when the water of the rivet starts fal-ling, the potentla] energy is tumed into kinetic energy. After that, this kineotic energy of the water is not lost when the water reaches the pool, but isconverted into thermal energy. This is illustrated by the fact that the waterat the bottom of the Niagara Falls is one-eight of a degree Celsius warmerthan at the top (Thirring, 19.58). But although the loss of the kinetic energyof the falling water is exactly compensated by the increase in the therma[energy of the water, the quality of the energy is altered. The ordered

18 ECN-I--94-053


motion of the flowing water can be used to do work on the b]ades of a tur-bine. With the random motion of the molecules of the water in the pool, theability to do work is much less. Consequently, the ability to do work basdiminished. Another example is a rotating wheel. When a brake is appliedto such a wheel, the mechanical energy is converted into thermal energy.So, when work is done against friction, the lost work equals to the heatproduced. The reverse however is hOt possible. It would be very convenientto be able to convert heat into work merely by reversing a process likefriction. This is however hOt possible as is described by the second law ofthermodynamics. The second law states that heat cannot completely betranst:ormed into work. Another formulation of this law is that one cannotmake heat spontaneously flow [:rota a body of low temperature to one ofhigh temperature.

Since the second law defines quality differences between types of energy, itplaces distinct restrictions on energy conversions. Although high-qualitywork can always be completely transformed into low-quality heat, this heatcan never be fully reconverted into work. Besides this, the second law per-vades a 100% efficient energy transformation, because some of" the high-quality energy will be degraded into heat. The warming up of an electricmotor when it r.uns is an example of this.

The economic relevance that thermodynamics brought to light is that mancan only use a particular form ot: energy. The second law tells us that allkinds of energy are gradually transformed into heat, and that this heat be-comes so dissipated in the end that man can no longer use it. Energy thuscan be divided into available (or free) energy, which can be transformedinto work, and unavailable (or bound) energy, which cannot be so trans-formed.

Although thermodynamics started with the study of the economy of theheat engine, standard economics did hot pay much attention to this phy-sics o[: economic value. Physical and chemical s¢ientists and biologists,however, did realize the relevance of thermodynamics as a basis for theeconomic process. The biologist and philosopher Herbert Spencer linkedthe evolutionary process to the laws of thermodynamics, because hethought that the struggle for existence was a struggle for available energyand resources. Spencer stated that ’evolution is a change from a less co-herent forto toa more coherent t:orm, as a result of the dissipation of ener-gy and the integration of matter (1880, p. 337)’.

In 1883, Sergei Podolinsky wrote Menschliche Arbeit und Einheit der Kraft.In this essay, this Ukranian biologist of Marxist persuasion consideredwhether a thermodynamic energy analysis could be fitted into the socialis-tic framework. His aim was to link the labour theory of value with the ther-modynamic analysis of the economic process. Podolinsky was quite awareof the importance of thermodynamics for economics. According toPodolinsky, the Iimits to growth of the economy are hot only to be soughtin the organization of production, but also and perhaps mainly in the physl-cal laws o~: thermodynamics.

The German chemist and blobel Prize winner Wilhelm Ostwald (1907,1911) argued that the concept and the laws of thermodynamics possess

19

The 3-E ìnteraction and the rebound-effect

the power to unify and clarify all branches of science. According toOstwa]d, ’energy is the sole universal genera]ization (]907, p. 488)’. Andtherefore, Ostwa]d argues that everything must be expressed in propertiesand relations of energy. A]so matter must be defined in terras of energy.Acc’ording to Ostwa]d, ’there is no other universa] condition among thedifferent domains of scientific phenomena than that of energy. That is tosay that whatever may happen physically, Jt is possib]e to state an equationevery time between the energies that hava disappeared and those thatnewly arrived. There is no other physical quantity to which such a genera]i-zation would apply (]907, p. 502)’. For this reason, Ostwald argues thatthe laws of thermodynamica shou]d be the basis for science as we]l as foreveryday ~ife.

In the early 20th century, the radiochemJst and Nobe] laureate Frederick$oddy (]922,]926) app]ied the laws of thermodynamics to the economiesystem. Soddy spant a considerable portion of bis career examìning therole of energy ~n the economie system anti developing alternative economietheories. According to $oddy, ’lire derives the who]e of its physical energyor power, hot from anything se]f-contained ìn living matter, and stil] ]essfrom an external deity, bui solely from the inanimate world. It is dependentfor al] the necessities of its physical continuance primarily upon the prJn-ciples of the steam-engine. The principles and ethics of butaan law and¢onvention must not run counter to those of thermodynamica (]922, p. 9)’Soddy argues that the main failure of" economics is the confusion of wealth,which is a physica] concept, with debt, which is an imaginary concept:’debts are subject to tbe laws of mathematics rather than physics. Unlikewealth, which is subject to the laws of thermodynamica, debts do hot totwith old age and are not consumed in the process of Iife. On the contrary,they grow at so much per annum, by the \vell-known mathematical laws ofinterest (1926, p. 70)’. Soddy thus argues that debt grows at ¢ompoundinterest and as a pure]y mathematical quantity encounters no limits. Wealthgrows for a while, but having a physical dimension, its growth eventuallyreaches a limit, t3ecause the growth rata of wealth is Iower than the growthrare of debt, the relationship between the two will cease to exist, and thebanking system will coJlapse. According to Soddy, the economie systemeould be restored by a 100% reserve requirements system, a policy ofmaintaining a constant price-level, anti freely fluctuating exchange rates.

2.4 Human history is energy-related

Because they have acknow[edged the fact that physical laws govem theeconomie process, some non-conformisti~ sclentists have tried to explainthe history of mankind by looking at the environment in general and theenergy provided by the environment in particular.

Regarding the relationship between man’s history and the environment ingeneral, Georgescu-Roegen (1973,1975A) pointed out that, for instance,the peoples l:rom the steppes of Asia, whose primary economic activity wassheep-raising, began their Great Migration over Europe at the beginning ofthe first millennium as an ultimate response to the exhaustion of the softCentral Asia, following a long period of grazing. Also unique civilizations,such as the Maya, vanished from the earth’s surface because their people

2O ECN-1--94-053


were unable to cope with the degradation of their environment by adequatetechnical progress. Besides this, Georgescu-Roegen argues that a]l conflictsbetween the great powers were hOt caused by conflicting ideologies. Theconflicts rather were a result of the unequa] distribution of natural resour-ces.

One of the first to emphasize the role of energy for the developmentsociety was Wilhelm Ostwald (]907,1911). According to Ostwald, allevents can be defined as transformations of energy. As a result of this,human civilization can be described by the increasing control of man overenergy. Ostwald argues that ’the progress of society is characterized by thefact that more and more energy is utilized for human purpose, and that thetransformation of the raw energies is attended by an ever-increasing ef-ficiency (1911, p. 870)’.

Uke Ostwald, Frederick Soddy ( 1912A, 1922, ] 926) believed that economicprogress was made possible because of mankind’s increasing use of ener-gy. According to Soddy, ’some conception of the part played by energy inhuman history began to take shape, and progress in the material sphereappeared hot so much as a successive mastery over the materials em-ployed for the making of weapons -as the succession of ages of stone,bronze and iron, honoured by tradition- but rather as a successive masteryover the sources of energy in nature, and their subjugation to meer therequirements of lire. The whole of the achievements of out civilization -inwhich it is difl:erentiated from the slow, uncertain progress recorded byhistory- appeared as due to the mastery over the energy of rite reachedwith the advent of the steam engine (]926, pp. 27-28)’. Soddy thus arguesthat the progress ot: out society is a result of the invention of" the steamengine. Because of this invention, the slow progress of human developmentwas dramatically speeded up. According to Soddy, ’the fact remains that ifthe supply of energy failed, modern civilization would come to an end asabruptly as does the music of an organ deprived of wind (1912A, p. 251)’.

The big Depression of the 1930s was a solid basis for the re-emergence of"a movement which started in 1918 as a group called the TechnicalAlliance. The Alliance conducted surveys of North America in which theeconomic system was described in joules rather than in monetary terms.The technocracy movement, led by Howard Scott, believed that the depres-sed economic conditions were a result of the inability of politicians andbusinessmen to manage the rapidly changing new world. The technocratsproposed to rep]ace the traditional leaders of society by engineers, who hadthe technical expertise to make the right decisions. The technocrats usedenergy as a unifying concept i:or social, political and economic analysisbecause they assumed that energy was the main determinant with respectto economic and social development. The Technical Alliance measuredsocial change in physical terms: the average number of kilocalories usedper capita per day.

Two ]ucid analyses of the role of energy in human progress were written byW. Fred Cottrel]. In ]955, this professor of sociology and govemment atMiami University wrote Energy and Society, and in 1972 he publishedTechnology, Man, and Progress. Cottrell argues that there is a relationshipbetween the energy that mankind uses and the kind of society it builds.

£CN-1--94-053 2 ]

The 3-1ì interaction and the rebound-effect

According to Cottreli, ’the amounts and types of energy employed con-dition man’s way of lire materially and set somewhat predictable limits onwhat he can do and on how society will be organized (1955, p, vii)’, Thisthought can be clarified with the example of the lndustrial Revolution. Cot-trell argues that the lndustrial Revolution was caused by the increasing useof the inanimate power of f’ossil fuels, which anormously increased labourproductivity. The result ot: the lndustrial Revolution was an enormous ìm-provement in butaan ¢ivilization. According to Cottrell, an important con-dition for civilization is the aggregation of large populations in cities, andthe differentiation of the inhabitants of these cities in specialized pro-fessions. Cottrell asserts that this kind of eivil[zation can only })een attainedwhen there exist considerable quant[t[es of surplus energy. The concept ofsurplus energy refers to the energy available to man in excess of that ex-pended to make the energy avaflable.

Another analyst of the role of energy in human history was the geologistEarl Cook. According to Cook, human progress has depended upon theincreasing control of energy for butaan purpose. In ~lan, Energ~j, Societg(1976), he set forth that centuries ago, the maritime nations Spain andHolland used the energy provided by the wind to achieve world dominance.Cook also argues that the main reason l:or the remarkably fast resurrectionof Europa affer the second World War was the utilJzation of very largequantities of abundantly avai]able and cheap energy.

With the thought that the history of man ìs energy-related, the current con-cern that the energy supply system, which is largely based on non-renew-able fossil fuels, cannot meer the demands of future generations, can beexplained. According to GeorgescuoRoêgen (1981,1986), hístory can beconsidered a sustained series of technical innovations. They took mankindfrom the cave to the Moon in a few thousand years. But, as curiously as itmay seem, only two innovations can be classified as fundamental break-throughs. The first such invention was the mastery of fire. Georgescu-Roegen argues that ’we may now regard lire as one of the toost ordinaryphenomena, yet that discovery was momentous. For fire is a qualitativeenergy conversion, namely the conversion of the chemlcal energy of com-bustible materials into calorie power. Noreover, fire creâtes ~ chain reac-tion: with just a small t’}ame we can cause a whole forest to buro (1981,p. 71)’. Man’s control over lire enabled him to warm his home and cookrood. Besides this, [t enabled man to smalt metals and bake ceramics.A¢cording to Georgescu-Roegen, it is therefore not surpdsing that the An-cient (ìreek attríbuted the bringing of íìre to man to the ímmortal demigodPrometheus. To emphasize the importance of the mast.ery of fìre for humanhistory, Qeorgescu-Roegen calls it the Promethean Technology 1. Thewood age was the result o1: this Promethean Technology 1.

As a result of the mastery of rite, wood was the main energy carrier [ormany centuries~ The beginning of the Industrial Revolution in WesternEurope however marked the end of the wood age. The problem with thePromethean Technology I namely was that because of the industria]growth the annual number of trees out down quick]y increased. Aecordlngto Qeorgescu-Roegen, this was only natural sin~e any Promethean Te~h-nology speeds up the deplet[on of the fuel wNch supports ir. The result ofthis cutting down trees was that by the second half of the seventeenth een-

22 ECN-1--94-053


tury, forest conservation measures had to be imposed by almost everyWestern European govemment. The Promethean Technology 1 was thusrunning out of its energy carrier.

According to Georgescu-Roegen, an alternative energy carrier, namelycoal, was known in Europe since the thirteenth century, but there were twoproblems associated with the substitution of wood with coal. The first dif-ficulty was that coa] bumed dirty. The second and most important difficultywas the t:act that a mine became quickly flooded when it was exploited,whereas the power sources at that time were hot forcefu] enough to drainthe mine. The solution to this problem came from the invention of the heatengine by Thomas Savery and Thomas Newcomen. Georgescu-Roegencalls this invention the Promethean Technology 2. According to Georges-cu-Roegen, ’the heat engine, exactly like rite, enabled man to perform anew qua]itative heat oonversion: the conversion of calorie power into motorpower. Moreover, also like rite, the heat origine leads toa chain reaction.With just a little coal and a heat engine, more coal and aiso other mineralscan be mined from which several heat engines can be made, which in tutuleads to more and more heat engines (]981, P. 72)’. The invention of theheat engine thus was a milestone in the history of mankind. This Prome-thean Technology. 2 nameIy enabled man to switch ffom an energy supplysystem based on wood to an energy system based on the much morepowerful energy carriers of mineral origin.

According to Georgescu-Roegen, mankind to a largo extent still lives in thePromethean Teohnology 2. But, like all Promethean Technologies, the in-vention ot: the heat origine led to an expansion which speeded up the deple-tion ot: the fuel which supported the Promethean Technology. And when-over the supporting l:uel of any Promethean Technology is almost ex-hausted, mankind’s future depends on whether or nota new PrometheanTechnology is discovered. Now Georgescu-Roegen argues that there is hotyet a new Promethean Teohnology 3, that wilI solve the present crisis, asthe invention of the heat engine solved that of the wood age. Solar energy,for example, is hot yet a viable technology. For solar energy to become thePromethean technology 3, one must be able to produce the convertors ofsolar radiation with the aid of only the energy converted by them and thIs isnot the case. According to Georgescu-Roegen, the result of all this is thatmankind is rapid]y approaching an energy crisis, due to the facî that thenew Promethean Technology is hOt yet found.

2.5 An energy theory of value

From claiming that human history is energy-related, it is a relatively smallstep to c]aiming that the value of a good is determined by the amount of"energy that is contained in ir. One of the first advocates of the energytheory of" value ~s L. Winiarski, who stated ~hat ’gold is the incarnation of"socio-biological energy (]900, p. 256)’. Winiarski suggests that gold is astandard f"or the amount of energy that is dissipate in an economie system.Ernest Solvay (]902) was a contemporary of Winiarski. Solvay argued thatthere is an equivalence between the value of" an object anti the amount ofenergy that it represents. Frederick Soddy (1912B,]922,1926) agreed withSolvay. Accord~ng to Soddy, ’as fat aa human affairs go, wea}th and avai]-

ECPt-1--94-053 23

The 3-E ~nteraction anti the rebound-ef£ect

able energy are synonymous, and the poverty or affiuence oF this planet areprimarily measured only by the death or abundance of the supply of energyavailable for its lire and werk (]912B, p. ]87)’. Soddy argues that wea]thbas its origin in useful or available energy. This imp]ies that every time anamount of wealth is created, an amount ei: energy is dissipated. Aceordingte Soddy, this seems reasonable since wealth cannot be created out ofnothing.

More recently, Howard T. Odum 1971,1974,1976,1977) argued that flowsof energy are the souree of a]l economie vaiu~. According te Odum,’butaan economie systems can bring in matedals and t:uels te supportpopu]ations and cultures. However, butaan beings are only a smal] port ofthe groot biosphere of oceans, atmosphere, mountains, valleys, land, rivers,forests, and ecological components. U]tlmately, it is net just butaan beingsand their money that determine what is important; it is uil the world’s ener-gy. It is thereFore a mistake te measure everything in money, lnstead, ener-gy shou]d be used as the measure, since only in this way we can accountfor the eontribution of nature (1976, p. 42)’. According te Odum, there is arelationship between energy and money which is called the counter-currentsystems relationship. This relationship implies that in money-based so¢ie-ties, money flows proportionate but in opposite direction te the energyrequired te produce the purchased goods and services. Odum a]so arguesthat out society is very sensitive te changes in inflowing energies: ’il a]lenergy sources are �ut off from the economie system, the dollar loses valueunti] it buys nothing, since there is no energy inflowing te produce goedsand services once storages in the system have been exhausted (1977, p.182)’.

Support for Odum’s theory was given by Robert Costanza (1980,1981).According te Costanza, energy flows are the main concern of energy analy-sìs, and an important port of this energy analysis is the quantification of theembodied energy of goods and services. For examp]e, the energy em-bodied in a dishwasher is the sum of the energy used te manufacture thedishwasher and all the energy consumed indirectly te pmduce the etherinputs of manufacturing, such as the steel and plastic needed for this com-modity. A systematic procedure te calculate the energy requirements, isthe input-output (I-O) analysis. For bis investigation, Costanza (1981) useda 90 sector energy input-output model maintained by the Energy ResearchGroup at the Llnivers[ty of lll[nois. Aooording te Costanza, the regressionanalysis resu]ts for total (direct plus indirect) energy consumption versustotal dollar output per sector indicate that a significant (r2 value of 0.99)relationship exists between embodied energy and dollar output. The onlyexceptions te this tule are the primary input seetors. According te Costan-za, the energy theory of value is rejected by mainstream economists, be-cause of the fact that energy is only ene et= a number of primary inputs tethe production process. But Costanza argues that ’[~rom a physical perspec-tive, the earth has only ene principal net input, namely solar radiation.Although very small amounts of meteoric matter a]so enter the earth’satmosphere, and deep residual heat may continuo te drive crustal move-ments, there is no stream of spacecraft carrying workers, government man-dates, and capital structures onto the planet. Thus, practically everythingen the earth can be ¢onsidered te bea direct or indirect product of pastand present solar energy. The same cannot be said for the ether primary

24 E(~N-I--94-053


factors (1980, p. 1219)’. According to Costanza, an energy theory of valueis really a production theory with all production factors costs carried backto the solar energy input.

25

3.THE ENTROPY LAW AND THE ECONOMICPROCESS

Thermodynamics is the only physical theory of universal content that willnever be overthrown. That is, heat will never pass by itself from the coldercondenser to the hotter boiler. [f a refrigerator mores heat to hotter spaces,it is only because far more heat passes from some boiler toa condenserelsewhere.Albert Einstein.

3.1 Introduction

In the preceding chapter, the energy-economy-environment interaction hasbeen outlined. An important aspect of this interaction is the possibly nega-tive impact of human activity on the environment. When taking into ac-count this negafive effect, limits emerge on the rare at which the economicsystem can expand. The concept of entropy can be used to clarify thisfiniteness. Therefore, the relationship between the entropy law and theeconomic process is the subject of this chapter. In paragraph 3.2, themechanical foundation of the economic discipline is discussed. In para-graph 3.3, the second law of thermodynamics (the entropy law) is des-cribed. The similarities between entropy and order are discussed in para-graph 3.4. Paragraph 3.5 deals with the earth’s finite amount of low entro-py. Paragraph 3.6 focuses on the limited substitution possibilities betweenthe production factors in an economy. Finally, paragraph 3.7 is concernedwith the fact that the economic process cannot be reduced to thermo-dynamic equations.

3.2 Economics and mechanics

Just like any other scientist in the first half of the nineteenth century, theleaders of the economic discipline of that time were impressed by the spec-tacular successes of the science of mechanics in astronomy. The aim ofthese leaders therefore was to build an economic science ’after the mecha-nics of utility and self-interest (1879, p. 23)’, as W. Stanley Jevons calledir. And although the economic science bas undergone profound changessince, the mechanistic view is still at the basis of economic thought. Accor-ding to Georgescu-Roegen, ’modern economists have developed, without athought, their discipline on the mechanistic tracks laid out by their forefat-hers, fiercely fighting any suggestion that economics may be conceivedotherwise than as a sister science of mechanics (1975A, pp. 347-348)’.The reason for this worship of mechanics is that economists wanted tobeable to predict the future state of the economy, just like Urbain Leverrierand John Couch Adams discovered Neptune. Leverrier and Adams dis-covered this planet, hOt by searching the sky, but by determining its posi-tion by calculations on a piece of paper.

ECN-I--94-053 27

The 3-E interact[on and the rebound-efl:ect

The three key concepts in mechanic$ are mass, speed, and position. Withthese concepts, any process can be described. In mechanics, no qua]itativebui only quantitative differences in the total (kinetic plus potentla]) amountof energy resu]ting l:rom a process are considered. According toGeorgescu-Roegen, the viewing of the economic process as a sister scienceoi’ mechanics results in the thought that a constant flow can arise from anunchanging structure. This thought was c]early expressed by A. C. Pigou,who stated that ’in a stationary state, l:actors of production are stocks, un-changing in amount, out of which emerges a ¢ontinuous flow of real in-come (3935, p. 19)’.

Because mechanics recognizes no qua]itative change but only quantitativechange, Georgescu-Roegen (1975A) argues that any economic activitymodeled afler a mechanica] process may be reversed, just like a pendu]um.Also, no laws of mechanics would have heen violated if the direction oftime was reversed and the earth was set in motion in the opposite direction.According to Georgescu-Roegen, actual phenomena in íhe real world, how-ever, do not fol]ow this story. Actual phenomena move in a definite direc-tion and involve qua]itative change. Even the money in the circular flowmodel cannot circulate back and forth within the economic process, be-cause ¢oins and banknotes ultimately become wom out and their stockmust be rep]enished from external sources. This is the ]esson of the secondlaw of thermodynamics, the entropy law. Thermodynamics bas changedthe mechanica] outlook of traditional physics. Now, it should also a]termainstream modern economics that is still based on the laws of mechanics.

3.3 The entropy law

The se¢ond [aw o~ thermodynamícs states that a process whose only netresult is to fake heat from a reservoir and convert it info work is impossible,or equiva]ently, that one cannot make heat flow by itsell: from a cold toahot object. It is also possible to state this law in another way. To i]lustratethis altemative formulation, consider a heat engine ~hat bas a boiler at tem-perature TI. A certain heat % is raken t:rom this boiler, the heat enginedoes some work W, and it then delivers some heat Q2 into a condenser atanother temperature T~. And suppose this heat engine is an ideal heat en-gine, a so caIIed reversible heat engine. In such a reversible heat engíneevery process is reversibIe, which means that, by in~:initesimal changes, wecan make the engine go in the opposite direction. This is only achievedwhen nowhere in the engine there is friction and nowhere in the enginethere is a place where the heat of the reservoirs ís in direct contact withsomething definite~y cooler or warmer. An example of an engìne cycle inwhich all processes are reversible, is the Carnot cycle~ This cycle isillustrated in figure 3.1.

28

The entropy ~aw and the econornic process

Volume

Fi9ure 3.1 The Carnot cyc~e (Fe~nman, ~966)

In the first step of the Carnet cyc)e, a gas in a cy]inder with a íriction]esspiston expands isotherma]]y at temperature T~, absorbing the heat 0.~. Theae¢ond step invo|ves an adiabatic expansion, in which the tempetaturedrops flora T~ to T~. In step three, the gas is isothermal|y ¢ompressed attemperatute T2, de]ive~ing the heat 0.2. çhe cycle is completed in step 4,the adiabatic ¢ompression. Dudng this comp~ession, the temperature risesfrom T~ back te T~. So, the gas is carried around a complete cyc[e, andduring this cycle the heat 0.~ is put in the engine at temperature TIand theheat O~ is removed frem the engine at temperature T2, ~ow, a¢cording teCarnet, the quotient of 0.2 and T~ is equal te the quotient of Q2 and T2.

(3.1)

Th[s means that if there is a single engine running between T~ and T2, thenthe result of an analysis is that ~~ is to T~ as 0.2 i$ to T» il: the engine ab-sorbs the energy O1 at temperature T1, and delivers the energy O2 at tem-perature Tz. Whenever the engine is reversible, this relationship must fol-low, AII this imp[ies that in any reversible process as much O./T is absorbedas is liberated. There is no gain or Ioss of 0./1". In classical thermodynamicsthis 0./T is called entropy, and we can say that there is no net change inentropy in a reversible cycle. The term entropy comes t:rom RudolfClausius, who derived it from a (3reek word meaning transformation orevolutfon. The symbol that usually represents entropy is $, and it isnumerically equal to the heat 0. delivered to a 1-deg~ee reservoir. Theentropy S is measured in joules per degree. So, [f we have a reversiblecycle, the total entropy o~ everything ìs hot changed, because the heatabsorbed at temperature T~ and the heat ~z delivered at temperature T2

’4-053


correspond to equal but opposite changes in entropy. And therefore the netchange in entropy is zero. Th[s law may look like the law of the conser-vation of energy, but it is hOt, because the rule only applies to reversiblecycles.

If irreversible events are included, there is no law of conservation of entro-py. An example of an irreversible event is this: If two objects with differenttemperatures T1 and Tz are put together, a certain amount of heat will flowfrom one to the other by itself. Suppose, for example, a hot stone is put incold water. Then a certain heat AQ is transferred from the stone to thewater. As a result of this, the entropy of the hot stone decreases by AQ/T1and the entropy of the cold water increases by AQ/T2. Because the terra~Q is positive and T1 is greater than T2, the change in entropy of thewhole world is positive and equals the difference of the two fractions:

This example can be general[zed into the fol]owing proposltion: in any pro-cess that is irreversible, the entropy of the whole world is increased. Only inreversible pmcesses does the entropy rema[n constant. Since no process isabsolutely reversible, there is always a gain in entropy. With thlsforeknowledge, the two laws of thermodynamics can be stated as:

First law of thermodynamics: the energy of the universe is always con-stant.

Second law of thermodynamics: the entropy of the universe is alwaysincreaslng.

Because the second law, which is offen referred to as the entropystates that the entropy continuously increases, it governs the irreversibilityof processes. The concept of irreversìbility can be explained with a ]og ofwood. When this log is burned, it can never be ’unbumed’. The buming of alog of wood is thus an irreversible process. The [ncrease in the entropythe universe is [rreversible as well.

3.4 Entropy and order

In the cla~sical thermodynamics of Carnot and Clausius, the entropy con-cept is used to describe systems on a macroscopic level. Because of theabstract definition of entropy in the classical thermodynamics, the entropyconcept bas no real meaning. Everybody knows what is meant by the phy-sical variables of mass, speed, and temperature. The entropy concept,however, only possesses a pure[y theoretical meaning. However usefulentropy may be for thermodynamics, it has no real descriptive meaning.

This changed with the statistical mechanics approach to thermodynamics.This approach describes the behaviour ot: ideal gases on the microscopielevel of atoms and molecules based on probability considerations. Take forexample the diffusion of two gases. For this purpose, a container filled with

3O

The entropy law and the economie process

an ideal gas I is considered. This container is connected to another con-tainer filled with an ideal gas 2 via a closed valve. Let the volume of thefirst container be Vl and the volume of the se¢ond container be Vz. AIso letthe two containers be iso]ated. If the valve is opened, the gases diffuse andthe result is a mixture of these two gases. The entropy change resultingfrom this process can be calculated with the following formula:

In this formula, the number of mols of the gas of: type 1 and 2 are denotedby N1 and N2. The universal gas constant is denoted by R, and the volumesof the containers 1 and 2 are represented by V~ and V2. This formula showsthat the diffusion ot: gases is accompanied by an increase in entropy. In theinJtial state of the process, in which both of the gases are separated by theva]ve between the two containers, there is a higher degree of order than inthe í~~aJ state, in which the t~vo gases are mixed. The increase in entropy isthus accompanled by an increase in the disorder of the system. ThereforeLudwig BoJtzmann argued that the concept of entropy can be used to de-scribe the order of a system. According to Boltzmann, the statistical defini-tion of entropy can be expressed by the following equation:

S -- k ¯ log D (3.4)

In this equation, the symbol k represents the so-called Boltzmann constant(1.38 * 10.29 Joule per Kelvin), and D represents a quantitative measure ofthe disorder of a system. Equation (3.4) can be illustrated with a lump ofsugar that is put in a cup of tea. The spreading out of the sugar over thewater in the cup results in an increase in disorder, and hence in an increasein entropy.

This way of looking at entropy was taken a step further by the informationtheory approach to thermodynam[¢s. In this approach, entropy is a mea-sure of the information contained in a message. The initiator of this move-ment was Claude Shannon. In his research on the capacity of com-munication channels, Shannon (1949) showed that one can define a gene-ral measure of the information contained in a message consisting of sym-bols Ei. According to Shannon, the following namely holds true:

(3.5)

In this equation, i denotes the number of symbols and Pi denotes the in-herent probability of recording symbol Ei. According to Shannon, thisamount of in[:ormatìon 1 ~an be linked to entropy by the following equation:

S:- k.l (3.6)

Consequently, the negative of entropy and information are equiva]ents.Besides Shannon, the equivalence between information and the entropyconcept was also expressed by Brillouin (1956) and Jaynes (1957).

ECN-I--94-053 31


The statistical mechanics and information concepts of entropy thus clarifythe implications of the second law of thermodynamics. The entropy of asystem is a measure of the randornness of the system. The greater thedegree of disorder, the higher is the entropy of the system. A briquette ofcoal contains an amount of ordered resources of low entropy. When thiscoal is burned, apart of the energy content of this energy carrier is con-verted into work, and the rest is converted into disordered waste heat. Thisheat no longer bas the capability to do work because of its low tempera-ture.

3.5 A finite amount of low entropy

Especially after the enormous progress of the past decades, standardclassical economists’ faith in science and technology to refute any knownlaw has become a widespread belief. And it is true that economicsapparently eludes the entropy law because living organisms, machines, andbanknotes remain almost unchanged over short periods of time. Entropicdegradation, however, pertains to long-run forces. Because these forces actextremely slowly, they are often ignored. Man is only interested in todayand tomorrow, not in what happens a thousand of years from now. Yet it isthe slow-acting forces of entropic decay that are the toost harmful in gene-tal. According to Nicholas Georgescu-Roegen, the entropy law is the majorlink between the economy and the environment. The main idea of bis bookThe Entropy Law and the Economic Process is that the entropy concept isat the basis of all economic lire. Georgescu-Roegen states that ’whateverthe economic expertise of other scientists, economists could not rare con-tinuously well in their own field without some solid understanding of theentropy law and its consequences (1971, p. ~352)L

Another explicit treatment of the importance of the concept of entropy forman was formulated by Erwin Schrödinger. In his famous writing What isLife ?, Schrödinger (1955) wondered what keeps mankind from death.Schrödinger argues that the obvious answer is by eating and drinking. Thetechnical terra for this is metabolism, which literally means change or ex-change. So, organisms exchange something through which decay is post-poned. After dismissing materials and energy, Schrödinger argues that lowentropy is what keeps us from death: ’every process that is going on innature means an increase of the entropy of that part of the world where it isgoing on. A living organism continually increases its entropy and conse-quently tends to approach the dangerous state of maximum entropy, whichis death. It can only keep aloof from ir, i.e. alive, by continuously drawingfrom its environment negative (-- low) entropy (1955, p. 72)’.

Erwin Schrödinger thus points out that an organism needs low entropywhich it sucks from the environment and degrades into high entropy~ In thiscontext, also the economic process can be described as a system in whichinputs enter in a condition of low entropy (their potential usefulness is at ahigh level) and iutputs pass out in a condition of high entropy (a low levelof potential usefulness). Besides our biological lire, also our whole econo-mic lire thus reeds on low entropy. The second law of thermodynamicsstates that any irreversible process is accompanied by an increase in entro-py. Because of this entropy law, the irreversible economic processes trans-

32 ECN-I--94-053

The entropy law and the economic process

form low entropy inputs into high entropy wastes. This makes the entropylaw, for example, the source behind the economic scarcity of energy. Wereit not for this law, the energy of a barrel of oil could be used over and overagain, by first transforming the energy content of this energy carrier intoheat, and after that transforming this heat into work, and finally transfor-ming the work back into heat.

Low entropy matter and energy cannot be created and can only be usedonce. Consequently, the sources of low entropy that mankind bas at itsdisposal are continuously used up. According to Georgescu-Roegen (!971,1975A), there are three sources of low entropy that man can use. Thesesources are the energy that can be liberated from out fossil fuel reserves,the free energy received from the sun, and the material structures stored inthe bowels of the earth. Now, one could argue that the technologies of theexploration anti extraction ol: fossil fuels have continuously improved overthe past 200 years. Besides this, one could say that the sun will continue toshine for millions of years. So, there is no need to worry about the supplyof low entropy energy. A similar optimism can be displayed for the supplyof low entropy materials. One could argue that it is foolish to think that wecan run out of matter, since the whole earth is made out of matter.

The problem, however, is that the high-quality fossil fuel resources havemostly heen extracted already, and that new discoveries tend tobe smaller,deeper, less accessible, and therefore generally more difficult to recover. Asa result of this, further technological improvements in exploration and ex-traction will ultimately become more costly to achieve. Moreover, at a cer-tain point in time, the amount of energy required to mine a ton of fossilfuels will become larger than the energy content of that ton of fossil fuels.Besides this, also the direct utilization of solar energy for human purpose ishot free of troubles. The reason for this is that there hot yet exists an ener-getically viable technology for using direct solar energy. The drawback of"solar energy in comparison with energy contained in energy carriers, suchas oil anti natural gas, is that the last are available in a concentrated form.Solar radiation, however, reaches the earth with an extreme Iow intensity.As a result of this, the direct use of solar energy in some appreciableamount would require a disproportionate amount of matter and energy. Inaddition to the critical remarks regarding the unlimited supply of low entro-py energy, one can also make objection to the thought that the supply oflow entropy matter is infinite. Although the earth is made of matter, thismatter is hot all of the kind that can be used by man to produce com-modities, such as cars and dishwashers. The reason for this is that for theproduction of those commodities, both accessible and available matter isneeded. In theory, it would be possible to manufacture commodities out ofthe metals that are dispersed in low concentrations in the water of theoceans. To collect these metals, however, would require so much low en-tropy resources that this is hot a realistic option. So, besides the amount oflow entropy energy, also the amount of low entropy matter is finite.

33


3.6 Limits to substitution

A]though there is only a limited amount of ]ow entropy matter-energy, thethought that the economy can maintain its historical growth rates is om-nipresent. The mechanism by which unlimited growth with a limited supplyof low entropy matter and energy can continue is the permanent substitu-tion of low entropy matter-energy with capita[ and labour. This thought canbe i[lustrated with figure 3.2. In this figure, the substitutíon possibil[tíesbetween a capital-labour aggregate and energy are shown.

The isoquant in figure 3.2 represents alternative combinations of capitaI,labour, anti energy to produce the same output. Consider for example aheat engine with a certain efficiency 111. With some capital expenditure, theefficiency of this engine can be increased from 111 to 1"12. As a result of thiscapital expenditure, the energy required to deliver the same amount ofservice is decreased. This substitution of energy with capital can be repre-sented in figure 3.2 by a sh~ft from point 1 to po[nt 2. The same shift is alsoobtained if the economìc output is produced with more labour and lessenergy.

In figure 3.2, the substitution of energy with eapitel or labour can continueuntil the energy requíred for the production of the output approaches tozero. In reality, however, this is impossible. The reason for this is that thesecond law of thermodynamics precludes the continued substitution bet-ween energy and other inputs.

K,L

Ffgure 3.2 Substitution graph

The energy efficiency of an engine ean be defined as the quotient of theuseful work delivered by the engine and the total energy input supplied tothe engine. Now Camot argues that the maximum energy efficiency ofwork that can be obtained by a heat engine from a temperature differenceis proporfionate only to the ratio between the temperatures, It can beshown that the theoretical maximum efficiency Emax of a heat engine can

34 ECN-I--g4-053


be calculated with the temperature of the heat source T~ (the boiler tempe-rature) and the temperature ot: the heat sink T2 in the following way:

In this formula, the temperatures T1 anti T2 are in degrees Kelvin. Emax isknown as the Carnot et:ficien~y of an energy transformation process. Alsoother cycles such as the Rankine cycle have theoretical maximum efficien-cies. The maximum efficiency of a fossil fuel electric power plant, forexample, is around 71~%. The second law of thermodynam[cs thus setsupper limits to the efficiency of economic processes such as power genera-fiono Therefore, besides the exact scìences, the entropy law also influenceseveryday economic lire. To include the entropy law in the substitutiongraph, the isoquant of figure 3.2 must be shifted to the right. This is il-lustrated in figure 3.3.

In this figure, Emin represents the energy requirement in the situation inwhich the process operates with a thermodynamically theorefical maximumefficiency

K,L

E min --~" E

Figure 3.3 Substitution graph and thermodynamics

Because of the entropy law, there are upper limits to the substitution pos-sibilities between low entropy resources anti the labour-capital aggregate.But suppose that the continued substitution of low entropy energy andmatter with capital and labour were possible, this would however hot neces-sarily imply that as a result of this substitution, the total amount of energyand matter were proportionately diminished. The substitution strategy as-sumes that capital, labour, energy, anti matter are independent inputs toproduction, which means that the availability of capital and labour does notdepend on the supply of low entropy energy and matter. In reality, how-ever, the fectors of production are hot independent of each other. Considerfor example the factor labour. An employee not only needs the energy that

35


is contained in his t:ood. He also needs clothes and a home. Besides theenergy needed for subsistence, an employee in today’s society also has aneed foran automobile to visit customers and magazine subscriptions tokeep up with the developments in his profession. Consequently, the energycontained in cara and magazines is also necessary to enable the worker toperform we]l.

This also appl[es to the production factor capital. There is an energy com-ponent to the factor eapita[ as well, as can be c[arified wíth the examplea heat recovery system. In such a system, the supply air flora the outside isheated by the exhaust air flora the insicle. This process takes p|ace in aheat exchanger. The efficiency of such a heat exchanger can be defined asthe ratio of the heat actually transferreò and the theoretically attainablemaximum heat transfer. When capital is expended to increase the surfaceof such a heat exchanger, a larger fraction of the availab}e heat can berecovered. Consequenfly, the energy efficiency has improved, However,th[s process of minimizìng the amount of energy by means of increasingthe heat exchangíng surface cannot go on forever. The reason f.or thís isthat a heat exchanger must be manu[aetured. When the size of the heatexchanger is increased, the amount of energy and matter required to pro-duce the piece of equipment increases proportionately. At a certain point,the amount of energy required to enla~ge the heat exchanger by a certainamount be¢omes greater than the addltional energy that can be recoveredby the increase in surface. Therefore, inereasing the surf.ace of a heat ex-changer beyond some point w[ll not reduce but rather [ncrease the totalamount of energy.

Another example concems insulation. Increasìng the insu~ation level of" adwelling leads to a lower annual energy demand for space hearing. How-ever, the insulation materials required for this purpose must be manufac-tured. In this manufacturing process, an amount of energy is used. Fromthe trade-off between the energy used during the manufacturing processand saved during the use of the insulation, an optimal environmental thick-ness (Bowdidge, 1990) can be determlned. This optimal environmentalinsulat[on thickness can be defined as the thickness of" insulation beyondwhi¢h no further reduction in the total amount of" energy requlred for spaceheating and manufacturing of the insulation material ean be achievecl. Theoptimal environmental insulation thickness thus denotes the maximumpotential for energy savings through insu[ation regardless of the amount ofcapital expenditure. This implies that expending capital to increase theinsulation level of a building beyond the opfimal environmental insulationthickness will have counterproductive ef.f.ects.

3.7 No analogy between economics andthermodynarnics

The concept of entropy thus can be used to explain the fact that the econ-omie system is subject to certain laws of physics. This however does notimply that all economic processes can be reduced to thermodynamica. ][t iswrong to degrade the economie science toa system of physical formulae.An example of. such an approach is given by John Bryant, who developed

36 ECN -I--94-053


a J~ormal analogy between thermodynamic equations and economic pheno-mena. According to Bryant (1985), the economic system can be defined interms of thermodynamic anti other physical laws. He argues that it is pos-sible to find equivalent properties in the economy f’or the thermodynamicconcepts of temperature and entropy. Bryant further argues that the tradecycles in the economy can be related in some way to the cycles used byphysicists and engineers to describe the working of heat engines. Bryantalso claims that physical equivalents to the concept of. inflation can bedetermined.

The problem with Bryant’s thermodynamic approach to economics is that itis an abstract exercise which has no real meaning. The economic sciencenamely bas properties that cannot be explained by natural ]aws, The samecriticism applies to the energy theory ot: value. Although energy is a veryimportant component in the production of goods and services, an in-complete picture of" the economie process is drawn if economic value isequated with the energy content of a good or service. Commodities aremore than embodied energy. Also the thought that human progress ismerely a result of the increasing use of energy for human purpose is tooone-sided. Energy is hot the only factor in the evolut[on of. mankind.

Because butaan conduct is unpredictable, no law of physics can ever ac-curately explain economic processes. No physical law, for examp~e, canexplain why a particular individual purchases brand X instead of. brand Y.The ]aws of physics, however, do set ]imits to out economie possibilities.E¢onomies is to a more or less extent constrained by the energy laws.

37

4.$OLUTIONS TO THE PROBLEM OFECONOMIC GROWTH IN A FINITEENVIRONMENT

Ideas are everywhereo but knowledge is rare.T. Sowell.

4.! Introduction

In chapter 3, the entropy ]aw bas been used to clarify the finiteness ofman’s physical surroundings. When this finiteness is raken into account,three ways of thinking about economie growth can be distinguished. Inparagraph 4.2, an optimistie view is given by the World Commission onEnvironment and Development, the writers of the well-known Brundtlandreport. They advocate a forceful economie growth, that does however con-sider the effects of this eontinued expansion on the ecosystem. A less op-timistic picture is sketched by Herman Daly in paragraph 4.3. Daiy arguesthat economie growth must cease in order to prevent it to cause irreversibledamage to the environment. In paragraph 4.4, a rather pessimist[c view isgiven by Nicholas Georgescu-Roegen, who states that the termination ofeconomìc growth is not sufficient. Georgescu-Roegen claims that the cur-tent growth trend must be reversed into a negative growth trend.

4.2 The solution of the World Commission onEnvironment and Development

What is needed is a new era of econornic growth; growth that is forcefuland at the sarne time soeially and environmentall~ sustainable.Gro Harlem Brundtland.

Since the beginning of the 1950s, economie growth has vastly improvedhuman conditions in large parts of the world. In those parts, infant mortalitybas heen falling, human lire expectancy bas heen increasing, the proportionof illiterates bas been declining, and new rood production techniques havecaused annual yields to increase. To achieve this improved quaiity of lire,industrial production has grown enormously over the past decades. Accor-ding to Rostow (]978), the annual increase in industrial production today isas large as the total índustrial production in Europe in the 1930s. And be-cause of this growth in production, also the use of natural resources basgrown tremendously. The earth is thus changing profoundly. And thisgrowth bas hot ended yet. Population will probably more than doubIefore stabilizing, thereby increasing economie activity even further.

This expansion bas a fundamental impact upon the environment, becausemany of the goods and services that have caused the economy to grow arematerial- and energy-intensiveo The negative effect of human economicactivity on the environment can be illustrated by the emergence of pheno-

39


mena, such as acid rain, the depletion of the ozone layer, and the green-house effect. But despite this explosive economlc growth, the numberpeople without rood, clothes, and a decent home is increasing. The leastdeveloped countries are becoming poorer and poorer. This also leods toenvironmental degradotion. An illustration of the way in which economicsand ecology act as antagonists is the fact that especially the paar Africannations overexploit the environment in order to pay their Western Worldcreditors. So, e¢onomic growth increasingly disrupts ecological processes,but the differences between rich and paar nations are becoming bigger, notsmaller. To reverse this process, govemments must rea~ize that econom[cdeve~opment issue$ cannot be de¢oupled from environmental issues. Theonly possible solutlon to the e¢onomic development problem and the en-vironmental problem is to treat them integrative.

This thought was behind the establishment in 1983 of the World Commis-sion on Environment and Development by the General Assernbly of theLlnited Nations. The World Commission was connected wíth, but indepen-dent of the UN and national govemments. The objectives of the Commis-s[on, presided by Gro Harlem BrundtIand, were ’to exam[ne the criticalenvironment and development [ssues and to ~ormulate realistic proposalsfor dealing with them, to propose new forms of international co-operationon these issues that will influence policies and events in the direction ofneeded changes, and to raise the levels of understanding and commitmentto actions of individuals, voluntary organizations, businesses, institutes, andgovernments (1987, pp. 3-4)’. The results and recommendations oí: theCommission were formulated in Our Cornrnon Future (1907). Ac¢ording tothe writers of" this so-called Brundtland report, the destructive [nteraction ofpoverty and environmental degradation is a waste of opportun[ties andresources. As a solution to this problem, the writers advocate a new era ofeconomic growth, in order to avert global catastrophes. But although theCommission’s overall assessment is that the international growth must be~peeded up, they stress that the environmental constraints must be respec-ted.

IEconomic growth is of course accompanied by changes in the environ-ment. A forest, for example, may be laat in one part of the ecosystem andga[ned elsewhere. The wrlters of the Brundtland report argue that this is hotnecessarily wrong if system-wide effects such as so[l erosion have heenconsidered. When the rato of use of renewable resources like foresta iswithin the limits of regeneration, the consumption of these resourcescauses na problemso For non-renewable resources such as fossil fue]s antiminerals, the story lies dit:ferent because their use reduces the stock. Thismeans that the use of these elements could lead to the process of industrialgrowth running into material resource constraints. But, according to theWorld Commission, this doos not imply that these resources should hot beused. Although non-renewabie resources are by definition exhaustible, stu-dies such as those made by Barnett anti Morse (1963), Goeller and Weín-berg (1978), and the OF’CD (1979) suggest that few minerals are likely torun out in the near future and that substitutes are available. Therefore, theWodd CommJss|on argues that the use of non-renewabJe resources will notbecome a problem now and in the future if the scarcity of a resource, theavailability of technologies fat minimizing depletion, anti the possibilitìes fatsubstitution are taken into account.

40 ECN-1--94-O,b3

Solutions to growth in a finite environment

The most important conclusion of the World Commission on Environmentand Development is that a forceful growth that does take into account na-ture’s carrying capacity is needed. This growth must be ’a process ofchange in which the expIoitation of resources, the direction of investments,the orientation of technological development, and institutional changes areall ín harmony and enhance both current and future potential to meerhuman aspirations (1987, p. 4ó)’. This process is called sustainable deve-lopment. Sustainable development c~n be defined as ’development thatmeets the needs of the present without compromising the ability of futuregenerations to meer their own needs (]987, p. 43)’.

An important part of this definition is the concept of needs, especially theneeds of those who are hot so well of. According to the World Commission,priority in the development process should be given to the needs of theworld’s poot: ’the essential needs of vast numbers of people in developingcountries for food, clothing, shelter, and jobs are hOt being met and beyondthese basic needs these people have legitimate aspirations foran improvedquaiity of lire. A world in which poverty and inequity are endemic will al-ways be prone to ecological and other crises. Sustainable deveIopmentrequires meeting the basic needs of all and giving everyone the opportunityto satisfy their aspirations for a better lffe (~I987, pp. 43-44)’. Meetinghuman needs and aspirations depends upon a forceful economic develop-ment, part[cularly in the third world. This development must be sustainab]e.The writers of the Brundtland report argue that sustainable developmentimp]ies that there are ]imits to the expansion of the economic system.These limits are a result of the way in which society is organized and thestate of science and technology. However, these limits are not fixed, so thateconomic growth remains possible.

When the limits that are imposed by the sustainable development conceptare respected, future generations will also be able to satisfy their needs.Besides the concern for the needs of the now living indigent humans, thisconcern for future generations is the second key concept in the definition ofsustainable development. According to the World Commission, ’many pre-sent efforts to guard and maintain butaan progress, to meer human needs,and to real[ze human ambitions are simply unsustainable in both the richand poor nations. They draw too heavily, too quickly on already overdrawnenvironmental resource accounts tobe affordable far into the ~uture withoutbankrupting those accounts. These efforts may show profits on the balancesheets of out generation, but out ch~ldren will inherit the lossea. We borrowenvironmental capital from future generations with no intention or prospectof repaying: They may damn us for our ~pendthrift ways, but they cannever collect out debt to them. We act as we do because we can get awaywith ir: future generations do hot vote; they have no political or financialpower; they cannot challenge out decisions (1987, p. 8)’. Therefore, thewriters of the Brundtland report advocate a sustainable development, sothat the opportunities for future generations will hot be lost.

As already mentioned, a central point in the reasoning of the World Com-mission on Environment and Development is that growth must be revivedin order to break through the negative sp[ral of" poverty and environmentaldegradation. According to LINIDO (1985), world industrial output must beincreased by a factor 2.6 when the standard o~ l~ving in developing coun-

~-053 ECN-I--94-053 41


tries is raised to the level of the industrialized countries. Qiven the projectedpopulation growth trends, a five- to tenfold increase in world production wfllbecome necessary by the time world population ceases to grow. Toachieve this, rapid economic growth is required. According to the writers ofthe Brundland report, this growth bas major consequences ~r the environ-ment. Therefore, sustainable development implies more than just economiegrowth. If development is tobe sustainable, it must change fundamentallyin order to diminish the burden that mankind places on bis physical sur-roundings.

According to the writers of the Brundfland report, especial]y a reliable ener-gy supply system that does not overpressure the environment is of majorimportance to the concept of sustainable developmenL The Commissionargues that the growth in annual world energy use bas diminished over thepast years. However, rapidIy growing populations and increasing standardsof Iiving in the deveIoping ¢ountries wil[ be accompaníed with an increasedenergy demand. An average inhabitant of an Afrìcan country, for example0uses tens of times less energy than a ¢itizen of a Western World country.To raise the quality of Iffe of the poot w[II require an enormous energy use,which cannot be drawn by the environment. This argument is especiallyvalid when the demand for energy is met by fossil-fuel energy carriers.

As a solution to this problem, the World Commission on Environment andDevelopment argues that the economic growth in the future can be madeless energy-intensive by means of energy efficiency improvements. Accor-ding to the World Commission, ’energy efficiency pollcies must be the cut-ting edge of national energy strategies for sustainable development, andthere is much scope for improvement in this direction. Modern app}iancescan be redesigned to deliver the same amounts of energy-services withonly two-thirds or even one-half of the primary inputs needed to run tradi-tional equipment. And energy efficient solutions are often cost-effective(1987, p. 14)’. A future with a diminishing energy to GDP ratio is needed.This can only be ach[eved if the world invests in energy efficiency ínsteadof building more primary supply sources. The World Commisslon arguesthat ’by using the toost energy efficient technologies and processes nowava[lable in alt sectors of the economy, annual global per capita GDPgrowth rates of around 3% can be achieved without a growing energy use.These growth rates are as least as great as can be regarded as a minimumfor reasonable development. This path requìres huge structura] changes toallow marker penetration of efficient technologies, and ìt seems unlikely tobe fully realizable by toost governments during the next 40 years (1987, p.173)’. But, according to the World Cornmission, ’the crucial point aboutthese lower, energy efficient futures is not whether they are perfectly realiz-able in their proposed time frames. Fundamental political and institutionalshifts are required to restructure investment potential in order to movea]ong these lower, more energy-efficient paths (1987, p.173)’. The Com-mission thus argues that the energy effieient scenario is the only option forthe future. This scenario is however not impossib]e, in many countries, theenergy to GDP ratio has fallen significant]y over the last years as a result ofthe introduction of energy efficient technologies. According to the WorldCommission, ’il properly managed, efficiency measures could allow in-dustrìal nations to stabilize their primary energy consumption by the turn ofthe century. They would also enable developing countries to achieve h~gher

42 ECN-l--94-053


levels of growth with much reduced levels of investment, foreign debt, andenvironmental damage (1987, p. 174)’.

4.3 The solution of Herman Daly

Anyone who believes that exponential gmwth can go on forever in a finiteworld is either a madman or an economist.Kenneth Boulding

Although they stress that economic growth must take into account theecosystem’s carrying capacity, the writers of the Brundtland report ad-vocate a forceful economic growth, as a means of alleviating poverty andsolving all other economic ills. They argue that economic output is a goodthing, as long as the environment is not overpressured. According to Her-man Daly (1973A,1973B,1977,1979), however, an ever increasing econ-omic output is not desirable at aIl. Daly argues that ’the growth paradigmhas outlived its usefulness. It is a senile ideology that should be retired intothe history of economlc doctrines (1973B, p. 152)’. According to Daly, anew paradigm in the economic science is required. This new paradigmmust correspond with the finiteness of our planet. This means that un-limited economic expansion is unwarranted. Herman Daly calls the newparadigm the steady state economy. The steady state can be defined as ’aneconomy in which the total population and the total stock of physicalwealth are maintained constant at some desired level by a minimal rate ofmaintenance throughput, i.e. by birth and death rates that are at the lowestfeasible level, and by physical production and consumption rates that areequally at the lowest feasible level (1973B, p. 152)’.

The first part of the definition of the steady state economy (constant stock)can be traced back to John Stuart Mill, and the originator of the secondpart (minimal flow of throughput) is Kenneth Boulding. In 1857, J. S. Millwrote Principles of Political Economy. In this book, Mill wrote about thestationary state in a way that is even more relevant today than in his owntime. According to MiIl, ’ir must always have been seen by political econ-omists, that the increase in wealth is hOt boundless. They must have seenthat at the end of what they term the progressive state lies the stationarystate, that all progress in wealth is but a postponement of this, and thateach step in advance is an approach to it (1857, p. 320)’. Mill argues thatthe stationary state is on the whole a better state than the progressive statein which growth is dominantly present. An egoistic ideal of life, which ischaracteristic for the social situation in the progressive state, is namely hotthe toost appealing future for mankind. MiIl argues that there is no reasonfor the fact that people who are already richer than necessary should be-come even more rich. Only in the developing nations is an increase ineconomic output still an important policy goal. Besides this, Mill arguesthat although much more people can be accommodated on the earth’ssurface, there is no reason for desiring this: ’the density of populationnecessary to enable mankind to obtain, in the greatest degree, all the ad-vantages both of cooperation and social intercourse bas, in all the mostpopulous countries, been attained. A population may be too crowded,though all be amply supplied with food and raiment. It is hOt good for aman tobe kept perforce at all times in the presence of his species. Nor is

ECN-I--94-053 43


there much satisfaction in contemplating the world with nothing left te thespontaneous activity of nature; with every rood of land brought into cul-tivation, which is capable of growing rood fat human beings; every flowerywaste or natural pasture plowed up, all quadrupeds or birds which are netdomesticated for man’$ use exterminated as his rivals for rood, everyhedgerow or super[luous tree rooted out, and scarcely a place leít where awild shrub e~ flower could grow without being eradicated as a weed in thename of improved agriculture (1857, p~ :324)’. Because of the negativeeffects of economie growth, John Stuart Mill hopes that his descendantswill be satisfied with a stationary state, be[ore they are forced te accept itwhen the earth can na langer cape with a growing economy. According teMill, such a statíonary state however does net toean a stationary develop-ment. Mill states that ’a stationary condit[on of capital and populationimplies na stationary state of human improvement. There would be asmuch scope as ever fat al| kinds of mental culture, and moral and socialprogress; as much raam for improving the art of living and much morelikelihood of its being improved, when minds cease te be engrossed by theart of getting en (1.857, p. 326)’.

Although he thought that polit[cal econom[sts must have always seen thatthe increase in wealth is net boundless, mainstream economists have cer-tainly net raken the words of Mill at heart. While admiring many of his etherpublications, Mill’s thoughts en the steady state did net struck raar withtoost of his colleagues. In Economie Theonj in Retrospective, Mark Blaug(1968), for example, rejects the statlonary state concept of MIII as stronglycoloured by bis socia[ views and hopelessly dated. Aceording te HermanDaly, however, Mill’s pleading for the stationary state is even more relevanttoday than in the ] 9th century.

Fat the second part of the steady state definition, the thoughts of KennethBou[ding (1949,1973) emerge. As already mentioned in chapter 2, Boul-ding’s closed economy of the future requires an organization of the econ-omie science which differs ffom the open economy of today. Boulding ar-gues that in the spaceman economy, the main task is net throughput butstock maintenance, and any technology which makes it poss[ble te ma[n-tain a certain stock with less throughput is an improvement. In this respect,increasing consumption rates are negative, instead of positive. Accordingte Boulding, ’il we had clothes that would net wear out, houses that did netdepreciate, and even if we cou|d maintain out bodily condition wìthouteating, we would clearly be much better of (1.973, p. 128)’.

New the origin of the steady state is traced back, the meaning of the steadystate economy can be c]arified. According te Daly, the first part of tbe defi-nition states that the steady state is a state with a constant stock of physi-cal wealth and a constant stock of people. These stocks of capital andlabeur cannot remain constant by themselves. Humans pass away andresources are used up. For this reason, Daly argues that the stocks of ea-pital and labeur can only remain constant if the rate of outflow (death andconsumpt[on) is compensated by an equally large rate of inflow (b[rth andproduction). This equilibrium te maintain total stock constant can be ob-tained either with a high rate of throughput (equal te both the rate of inflowand outflow) or with a low rare. New Daly argues that the second part ofthe definition of the steady state economy implies that the rate of through-

44

So~utions to growth in a finite environment

put shouid be as low as possible. According to DaIy, ’~or an equflibriumstock the average age at death of its members is the reciprocel of the rareof throughput. The faster the water flows through the tank, the less time anaverage drop spends in the tank. For the population, a low rare of through-put (a low birth rare and equally low death rare), means a high lire expec-tancy, and this is desirable for that reason alone. For the stock of wealth,a low rare of throughput (low production and equally ]ow consumption)ineens greater lire expectancy and durability of goods and less time sacri-ficed to production. This ineens more leisure or non-job time, tobe dividedinto consumption time, persona] and household maintenance time, culturetime, and idleness. This too seems socially desireble, at leest within limits(1973A, p. 14}’.

To minimize the flow oI: throughput, Daly argues that two variables can beused, namely the size of the stocks and the durabil[ty and the stocks. G[venthe fact that the size of the stocks remains constant in the steady stateeconomy, the durability of the stocks must be maximized so that the deple-tion of resources is minimized. According to Daly, the concept of durabilitygoes beyond the lire expectancy of a commodity. Also what happens toacommodity after its use ought to be a matter of thought. For example, thepossibility for recycling must be investigated, when talking of durability.Daly further argues that the economy’s use of resources must be modeledafter the cycles in nature. According to Daly, ’the best use of resourceswould ìmitate the model that nature bas fumished: a closed-loop system ofmaterial cycles powered by the sun. In such an economy, durability ismaximized, and the resources on earth could presumably last as long asthe sun continues to radiate the energy to rum the closed material cycles(1973A, p, 15)’. Daly thus asserts that an economic system must becreated in which all waste is recycled.

It is frequently urged that a state of non-growth will ultimately becomenecessary because of an ever increasing scarcity of naturel resources onthe depletion side of the economy. Herman Daly, however, argues thatmainly the pollution side provides the limits to a growing economy. Thelaws of the conservation of mess and energy state that matter and energycannot be destroyed. Consequently, depletion ultimately leads to pollution.Therefore, pollution provide~ another foundation for the steady state e¢on-omy. According to Daly, this pollution side bas been less studied than tbedepletion side. The reason for this is that a large part of the input side isdivided into pieces of private ownership. The output side, however, is hOtpartitioned. Therefore, the waste absorbing capacity of the environmentcan, to some extent, be used by everyone. This results in a overexploitationof the environment. Garrett Hardin (1968) calls this the Tragedy of theCommons. Herman Daly tends to cell it the invisible foot: ’Adam Smith’sinvisible hand leads private self-interest to unwittingly serve the cornmongood. The invisìble foot leads private self-interest to kick the common goodto pieces. Private ownership and private use under a competitive markergive rise to the invisible hand. Public ownership with unrestrained privateuse give rise to the ìnvisible Ibot. Public ownership with public restraints omuse give rise to the visibie hand (and foot) of the p]anner. Depletion hasbeen partially retrained by the invisible hand, whi]e pol]ution bas been en-couraged by the inv[sible foot (1973A, pp. 17-18)’. DaJy argues that it is

E~N-I--94-053 45


therefore hot surprising to find limits occurring main]y on the pollution side.

Now the necessity of the steady state has been clarified, the question ofhow it should be attained emerges. According to Herman Daly, a stationarystate can be obtained with the right technology and the right social institu-tions of control for keeping the stocks of physical wealth and people con-stant and for distributing the stock of wealth among the people. The firstmeans of assuring the future is an appropriate technology. Daly argues thatproduction technology in the steady state ought tobe aimed at improvingthe durability of commodities. According to Daly, ’maximum durabi]itymeans maximizing the time matter spends as wealth and minimizing thetime it spends as garbage. Our current technology does not aim at maximi-zing durability. It comes closet to minimizing ir, in order not to spoil themarker for rep]acement demand (1973B, p. 157)’. According to Daly, tech-nological progress should be aimed at minimizing the negative environmen-tal effects of production.

The second means of assuring the future are the social institutions of con-trol and distribution. Daly argues that ’the guiding design principle for so-cial institutions is to provide the necessary control with a minimum sacri-fice of personal freedom, to provide macrostability while allowing for micro-variability, anti to combine the macrostatic with the microdynamic (1973B,pp. 157-158)’. An example of this thought are the freely tradeable birthcertificates tobe used for maintaining a constant population level. The ideabehind these certificates is to give each person a licence to have an num-ber of children corresponding to the replacement fertility. These certificatescould be traded on a free marketo In this way, there could be stability onthe macro-level and variability on the micro-level.

Herman Daly closes his outline of the steady state economy by remarkingthat this state is hot unrealistic. According to Daly, ’the steady state is inbroad characteristics the only realistic possibility. The present economy isliterally unrealistic because in its disregard (:or nature it is attempting theimpossible. The steady-state paradigm, unlike the growth paradigm, isrealistic because it takes the physical laws of nature as its first premlse(1973B, p. 170)’.

4.4 The solution of Nicholas Georgescu-Roegen

The econornic process is entropic in all its material fibres.Nicholas Georgescu-Roegen.

Mankind has been fascinated by myths for centuries. Many myths deal withman’a foolish thought that he is superior to everything else in the universeand that his powers are unlimited. Once, man believed that the earth wasthe centre of the universe. At another time, man thought that he couldmove an object without uslng energy or tha’~ he could use the same energyover and over again. According to Georgescu-Roegen (1971,1973,1975A,1980,1981) another myth is proposed by mainstream economists. It is themyth that man will always discover new sources of energy and new ways ofexploiting them. The thought is that man, with his unlimited technologicalpossibilities, will always find a solution toa problem when the situation

46 ECN-I--94-053


becomes critical. According to Georgescu-Roegen, it is true that up to nowalmost every generation was better of than the preceding one. But it is alsotrue that each generation bas used up a greater amount of natural resour-ces. Georgescu-Roegen argues that the great mineral bonanza of the pasttwo hundred years enabled mankind to achieve a great economic growth.Especially after the second world war, an abundance of cheap crude oilmade the increase in the welfare in most parts of the world possible. AI-though technological change was essential for economic growth, thisgrowth was only possible with the incredible amounts of low entropy provi-ded by the mineral bonanza. Because of the existence of enormousamounts of natural resources of low entropy, Georgescu-Roegen assertsthat mainstream economists think that the entropy law does hot apply tothe economic discipline. According to Georgescu-Roegen, ’the most naturalrallying idea is that mankind’s entropic dowry is virtually inexhaustible,primarily because of man’s inherent power to defeat the entropy law insome way or another. The thought is that just as has happened with manyother natural laws, the laws on which the finiteness of accessible resourcesrests will be refuted in turn (1975A, p. 359)’. But, as Georgescu-Roegenargues, ’the difficulty of this historical argument is that history proves witheven greater force, first, that in a finite space there can only be a finiteamount of low .entropy and, second, that low entropy continuously andirreversibly dwindles away (1975A, p. 359)’ Georgecu-Roegen further ar-gues that substitution offers no solution to this problem, because energy inits abstract form cannot be substituted, whatever the market price.

Although he states that thermodynamics is the physics of economic value,Georgescu-Roegen is hOt completely satisfied with the way in which Car-not, Clausius and their descendants formulated the laws of thermo-dynamics. Georgescu-Roegen argues that ’thermodynamics has remained ascience concerned only with what happens to energy. It has completelyignored what happens to matter, without which there can be no piston andcylinder performing transformations of thermal energy into mechanicalwork and conversely. To be sure, thermodynamics recognizes that becausefriction is inherent in any natural production of mechanical work, part of theenergy initially available for this effect is wasted into dissipated heat, with-out producing any work. No actual engine, therefore, can convert all avail-able energy into useful work. Thermodynamics however only gives for-mulae for energy transformations, but nothing is said about the changesundergone by the matter of the piston, the cylinder, the wires, the chemicalsolutions, or the gases themselves (1981, p. 54)’. The reason for ignoringwhat friction does to matter lies in the fact that friction is a very difficuItand controversial phenomenon. In this context, Ernest Rabinowicz (1965)remarked that there are very few statements that can be made in the fieldof frlction which are hot controversial.

Thermodynamics thus acknowledges the fact that because of friction anamount of available energy is lost. But it does hot consider the fact thatfriction influences matter as well. According to Georgescu-Roegen, thisleads to the modern energetic dogma, which states that only energy mat-ters. Georgescu-Roegen argues that the energetic dogma is ’the notion thatwith a sufficient supply of energy we can mine any rock regardless of itsmineral content and also recycle ¢ompletely any material substance (1980,p. 79)’. The energetic dogma can be clarified with the statement of Har-

ECN-I--g4-053 47

The 3-F" interacfion and the rebound-eft:ect

rison Brown (1957), who claims that all that needs to be done to obtainwhatever material man desires is to add sufficient energy toa system. Themodern energetic dogma is also supported by Einstein’s equivalence E =MC2. Georgescu-Roegen however argues that ’despite the Einsteìnian e-quivalence of mass and energy, there is no reason to believe that we canconvert energy into matter except at the atomic scale in a ]aboratory andon]y for some special elements. The point is that even the formation of anatom of carbon from three atoms of helium, for example, requires such asharp timing that its probabil[ty is astonishingly small, and henee the eventmay occur on a large scale on]y within astronomica]ly hugl^Osses. Wecannot produce a copper sheet, for example, from energy alone. Al1 thecopper in that sheet must exist as copper (in pure forto or in some chemi-cal ¢ompound) beforehand. Therefore, the statement that energy is conver-tible into toost of the other requirements of lire ìs, in this unqualified forto,apt to mislead (1975A, pp. 355-356)’.

The second law of thermodynamics states that every [rrevers[ble process isaccompanied by a decrease in the quality of energy. According toGeorgescu-Roegen, just like energy, matter is subject to an irreversibledegradation as well: ’ever since my first thoughts on the entropie nature ofthe economíc process, my position bas heen that in any system that per-forms work of any kind hot only free energy but also matter arranged insome definite structure continuous]y and irrevocably dissipates (1981, pp.58-59)’. The fact that matter is important as well, inspired Georgescu-Roegen to formulate a new law, which he calls the fourth [aw of thermo-dynamics. The first law (total energy is constant) and the second law (theentropy steadily increases) were already mentioned. The third law of ther-modynamies states that the absolute temperature of zero degrees cannotbe reached, or equNa[ently, that the entropy beeomes zero at the absolutezero of temperature. According to Georgescu-Roegen, the new fourth law ofthermodynamics completes the old laws of classical thermodynamics anòis formulated as:

Fourth ]aw o.~ thermodynamics: a c]osed system cannot perform workindefinitely at a constant rare.

Just like the other laws of thermodynamics, the fourth law can be stated inother, but equivalent, formulations. One of these formulations and probablythe toost important one for economics is that complete recycling is impos-sible.

Now one could object to this formulation of the fourth law that it would bepossible to recycle al] the small particles released from a wom out auto-mobile rite, ]ust as it would be possible to reassemble all the pearls of abroken necklace in a toom. But Georgescu-Roegen argues that to reuniteall these tire particles would require an immense effort spread over a veryJong time, during which the objects that are used in this collection effort willbecome wom out as well and will therefore have tobe reassembled in tutu.This ]eads to an infinite regress. Materials thus wear out in such a way thatthe molecules origìnally belonging to these materials are gradually dissi-pated beyond the possibility of being reassembled. This means that oncedissipated, matter cannot be eompletely recycled. On]y matter that is thatis no Jonger in a desired form can be recycled. For this reason, Georges-

48 ECN-I--94-053

Soiutions to growth in a finite environment

cu-Roegen argues that only what is found in garbage and junk yards, theso-called garbojunk such as broken glass, old newspapers and worn outmotors, can be recycled. So, just like energy, matter also exísts in twodifferent states, namely available and unavailable.

Because of bis firm belief in the four laws of thermodynamics, Georgescu-Roegen argues that ’what goes into the economic process represents valu-able naturel resources and what is thrown out of it is valueless waste. Fromthe viewpoint of thermodynamics, matter and energy enter the e¢onomicprocess in a state of low entropy and come out of it in a state of high entro-py (1973, p. 37)’. Georgescu-Roegen further argues that the economicprocess neither produces nor consumes matter and energy; it only absorbsmatter and energy and throws it out continuously. This may lead to thenotion that the economy only changes valuable inputs of low entropy intovalueless waste of high entropy. Georgescu-Roegen however objects to thisthought, when he argues that ’ir compels us to recogníze that the real out-put of the economic process (or of any life process, for that matter) is notthe matedal flow of waste, but the still mysterious flux of the enjoyment oflire. Without re¢ognizing this fact, we cannot be in the domein of lire pheno-mena (1975A, p. 353)’.

The fact that the e¢onomic process transforms valuable inputs of naturelresources to va]ueless outputs of waste can be explained by a thermo-dynamic hourg]ass. Georgescu-Roegen portrays the economic process byan hourg]ass in wbich the stuff of the upper half stands for avai]able ]owentropy matter-energy. By pouring this stuff down into the lower half itbecomes unavaflab]e high entropy waste. But unlike the normal ones, thisthermodynamic hourglass cannot be tumed upside down. With this hour-g[ass, Georgescu-Roegen explains that ’in the context of entropy, everyaction, of man or of an organism, ney any process in nature, must result ina deficit for the entire system (1975A, p. 354)’. Every time, for example, acar is driven from point A to point B, the entropy of the universe increases.Another example is the mining of some copper ore. AIthough the entropyof the ore decreases every time a copper sheet is produced, this result canonly be achieved at the expense of a much greater increase in entropysomewhere else.

When the upper half of the thermodynamic hourgiass was replenishedagain and again, the transformation from low entropy inputs to high entro-py waste would hot be a reason for concern. Georgescu-Roegen howeverargues that in a finite world, there can on]y be a finite amount of low entro-py. Because of this finiteness of the amount of low entropy matter-energy,Georgescu-Roegen points out that ’the truth, however unpleasant, is thatthe toost we can do is to prevent any unnecessary depletion of resourcesand any unnecessary deterioration of the environment (1975A, p. 363)’.This is the reason why Georgescu-Roegen emphasizes that ’every time weproduce a Cadillac, we irrevocably destroy an amount of low entropy thatcould also be used for producing a plough or a spade. In other words,every time we produce a Cadil]ac, we do it at the cost of decreasJng thenumber of human lives in the future. E¢onomic development through in-dustrial abundance may be a blessing forus now and for those whoenjoy it in the near future, but it is definitely against the interest of the

ECN-I--94-053 49

The 3-E interactlon and the rebound-effect

human species as a whole if its interest is to have a lifespan as long as iscompatible with its dowry of low entropy (1973, pp. 46-47)’,

Moreover, besides the problem of the finiteness of low entropymatter-energy, continued economie growth must lead to more waste.Because of the laws of thermodynamics, Qeorgescu-Roegen argues that’we cannot produce better and bigger refrigerators, automobiles, or jetplanes w[thout producing also better and bìgger waste (1973, p. 44)’.Numerous natural scientists and economists think that the industrial pro-cesses of the future wíl] produce a neglígible amount of" waste and that thisnegligib]e amount of waste can be reeycled permanently. Georgescu-Roegen however argues that recycling requires an amount of Iow entropyresources much greater than the decrease in the entropy of what is re-cyeled. So, there is no industrial activìty free from a eonsiderable amountof pollution, just as there is no ¢ostless disposal of pollution in terms ofentropy.

Another problem that Georgescu-Roegen outlines is that of thermal pol-lution. The second law of thermodynamics states that it is impossible toachieve an efficiency of ]00% and that eventual]y a]] energy ends up aswaste heat. An implication of this second ]aw is that continued economiegrowth leads to an increase in the use of energy and this in tutu leads to anincrease in the amount of dissipated heat. Because this waste heat is animportant aspect in maintaining the thermodynamic equi]ibrium betweenthe earth anti the rest of the universe at a favourable temperature, an in-crease in the amount of dissipated heat per unit of time can alter this ba-]ance. This can have serious effects on the environment. Therefore, Geor-gescu-Roegen argues that ’since the entropy law a]lows no way to cool acontinuous[y heated planet, thermal po/]ution could prove tobe an evenmore crucial obstacle to growth than finiteness of accessible resources(1975A, p. 358)’.

Given all these probIems, Georgescu-Roegen concludes that the onlyreasonable path for the future would be to economiz.e as much as possib]e.Therefore, the most desirable state is hot a growing one, hot even a station-ary one, but a dec/ining state. According to Georgescu-Roegen, the currentgrowth must cease, anti must be reversed to achieve a negative growth. Toreverse the current growth trend, Georgescu-Roegen proposes quantitativeregulations that must result in the termination of squandering activities,such as overlighting, overheating, overcoo]ing, and overspeeding. A]so theproduction of superfluous ¢ommod~ties must be stopped. Besldes ibis,fashion must be banned and durable goods must be made more durable.Although these regulagons wi/l lead to less comfort, Georgescu-Roegendoes hot want man to return to the cave. But he does want to reverse thecurrent growth trend. Georgescu-Roegen is however sceptlc of man’s wil-lingness to cope with a negative economie growth. According to Georges-cu-Roegen, ’there is neither cynism nor pessimism in believing that even il:made aware of the entropie problem of the butaan specles, mankind wouldhOt be willing to give up its present luxurious lire in order to ease the lire ofthose humans who will lire ten thousand or even one thousand years fromnow. It is as if the human species were determined to have a short butexciting life. Let the less ambitious species have a long but uneventful exis-tence (1973, po 47)’.

50 EqN-I--94-053

5. THE REBOUND-EFFECT:AN INTRODUCTION

5.1 lntroduction

In the previous chapter, three ways of thinking about economic growth in afinite environment have been presented. These views were positive growth(The World Commission on Environment and Development), zero-growth(Herman Daly), and negative growth (Nicholas Georgescu-Roegen). Ofthese three options, the thoughts of Gro Harlem Brundtland and her con-federated are of course the most attractive. The writers of Our CommonFuture advocate continued economic growth in combination with efficiencyimprovements in the use of energy and materials. The thought behind thlsapproach is that energy efficiency improvements make it possible to haveincreasing economic growth rates without increasing demands for energycarriers, such as oil and gas. The problem with this energy efficiencyapproach to solving all energy-related problems however is that it is hotcertain that using energy more efficiently reduces the demand for itproportionately. The rebound-effect is a means of quantifying this effect.The rebound-effect can be defined as that part of the initially expectedenergy savings, resulting from energy efficiency improvements, that is Iostbecause of the 3-E interaction. In this chapter, an introduction into thisrebound-effect is given. Paragraph 5.2 contains the thoughts of the famous19th century economist W. Stanley Jevons on the subject of using energymore efficiently. In paragraph 5.3, the polemic with respect to the rebound-effect in the international journal Energy Policy is outlined. Paragraph 5.4describes a case study of the rebound-effect.

5.2 The thoughts of W. Stanley Jevons

The rebound-effect is hot new. Already in 1865 it was made clear thatusing energy more efficiently does not necessarily reduce the demand forir. In that year, W. Stanley Jevons published The Coal Question, an lnquiryConcerning the Progress of the Nation and the Probable Exhaustion of ourCoal-Mines. According to Jevons, it was possible to divide the history ofBritish industry and trade into two periods. Before the middle of the 18thcentury, Britain was a rude and half-cultivated country, with an abundanceof resources that were exported, such as com, wool, meat, and timber. TheBritish people were unskfiled. They were leamers rather than teachers com-pared with other cultures. But after 1750, all that changed. Instead ofexporters of raw materials, the British became importers; instead ofimporters of manufactured goods, they became exporters; instead ofleamers, they became teachers. According to Jevons, this progress in thecivilization of Britain was hOt due to any general intellectual superiority ofthe British inhabitants, but to the availability of natural resources.

In 1850, the British export of iron was scarcely inferior to the total produc-tion of iron of the rest of the world. W. Stanley Jevons argued that this washot a result of the quality of the British iron. Compared to the iron pro-

51


duced in Sweden, Belgium, or Austria, the British iron was of a very pootquality because of the sulphur, phosphorus, and other impurities of theBritish ore. The ressort for the British dominance was the cheapness oftheir products. According to Jevons, this cheapness depended upon extrac-ting coal from the British mines at very low costs. Coal thus was a veryimportant resource. Or as Jevons put it: ’day by day it becomes more ob-vious that the coal we so happily possess in excellent quality and abundan-ce is the mainspring of modern material civilízation. And as the sourceespecially of steam and kon, coal is all-powerful. This age has been calledthe kon Age, anti it is true that iron is the material of greatest novelty. Byits strength, endurance, and wide range of qualities, it is fitted to the ful-crum and lever of great works, while steam is the motive power. But coalalone can command in sufficient abundance either the iron or the steam;and coal therefore commands this age: the age of coaL Coal, in truth,stands hot besides but entirely above all commodities. It is the materialenergy of the country, the universal aid, the factor in everything we do.With coal almost any feat is possible. Without coal we are thrown back intothe laborious poverty of earlier times (1865, pp. vii-viii)’. With this in mind,W. Stanley Jevons argued that it was hot surprising that every year theamount of coal that was extracted from the British mines became largerand larger.

Britain was thus growing rich upon an apparently infinite supply of coal.Jevons however argued that this was a fallaey because coal ìn itself waslimited in quantity. This could be seen in the increasing difficulty to gainthe needed supplies each year. According to Jevons, ’in the increasingdepth and difficulty of coal mining we shall meer the vague but inevitableboundaries that will stop out progress. And while other countries mostlysubsist upon the annual and ceaseless income of the harvest, Britain isdrawing more and more upon a source which yields no annual interest, butonce turned to light and heat and force is gone forever into space (1865, p.307)’.

Around the middle of the 19th century, three different solutions to theproblem of supplying Britain with ¢oal in the future were considered. As afirst solution, it was suggested by many scholars that when the coal wasused up in Britain, it could be imported. But W. Stanley Jevons argued thatthe principles of trade denied such a thought: ’il: coal, as well as other rawmaterials, were t:ound abroad, in Pennsylvania, Prussia, New South Wales,or Brasil, the whole cost of freight would be a premium upon establishing asystem of coal-supported industry on the spot (1865, p. 223)’. Therefore,foreign coal could not be expected tobe the foundation of a prosperousBritish industry. According to Jevons, the fact that Britain exported coal,instead of proving an eventual possibilìty for a reversal of this current,proved its impossibility.

A~ a second option, it was argued that subsfitutes for coal could be found.After investigating the alternative sources of power of that time, Jevonshowever concluded that coal would probably never be replaced by anythingbetter. According to Jevons, ’coal is the naturally best source of power asair and water and gold and kon are, each for its own purpose, the mostuseful of substances, anti as such will never be superseded (1865, p. 142)’.Jevons acknowledged that il: the coal mines would become exhausted,

52 ECN-I--94-053

The rebound-effect: an introduction

other sources of power such as tidal mi]ls would be useful substitutes for ir.But this would only be on the principle that something is better thannothing. And it would certainly enfeeble Britain’s industrial position, com-pared with nations that would still possess large quantities of coal.

As a third solution, it was urged that the failing supply of coal could be metby using it more efficiently. It was supposed that with efficiency improve-ments the problem of scarce and costly fuel could be solved. With the aidof thermodynamics, it was possible to show that the efficiency of a goodengine in Jevons time was less than 20%. In furnaces too, only a smallportion of the energy content of the coal was actually used. Also in thedomestic use of coal, a large part the heat escaped up the chimney un-used. W. Stanley Jevons however argued that efficiency improvements inthe use of coal are hOt a solution: ’ir is however a confusion of ideas tosuppose that the more economical use of fuel is equivalent to a diminishedconsumption. The very contrary is the truth (1865, p. 103)’. According toJevons, ’ir is the very economy of use of coal which leads to its extensiveconsumption. It bas been so in the past, and it will be so in the future. Noris it difficult to see how this paradox arises. The number of tons of coalused in any branch of industry is the product of the number of separateworks anti the average number of tons consumed in each. Now, if thequantity of coal used in a blast-furnace, for instance, be diminished in com-parison with the yield, the profits of the trade will increase, new capital willbe attracted, the price of pig-iron will rail, and the demand for it will in-crease. And eventually, the greater number of fumaces will more thanmake up for the diminished consumption of each. And if such is hot alwaysthe result within a single branch, it must be remembered that the progressof any branch of manufacture excites a new activity in most other bran-ches, and leads indirectly, if hot directly, to increased inroads upon outseams of ¢oal (1865, pp. 104-105)’ Jevons thus argues that the moreenergy efficient the economic system becomes, the more will the economythrive, and the higher will the annual coal consumption be.

For almost any use of coal, the energy efficiency can be increased. As anexample of the many possibilities for saving coal, Jevons mentioned theutilization of the spare heat from waste gases of a blast-furnace by passingit through a steam-boiler. But, according to Jevons, ’no one must supposethat the coal thus saved is spared; it is only saved from one use tobe em-ployed in others, and the profits gained soon lead to extended employmentin many new forms. The several branches of industry are closely inter-dependent, and the progress of any one would lead to the progress of near-ly all. And if the economy of use in the past has been the main source ofprogress and growing consumption of coal, the same effect will follow fromthe same cause in the future (1865, p. 115)’.

According to W. Stanley Jevons, the exhaustion of the British mines wouldbe a¢companied by price rises of coal, compared with the prices of ¢oal inother countries. As a result of this, Britain’s industry and trade wouldperish. So, the increasing cost of coal would be an obstacle to further pro-gress. In a response to this problem, Jevons argued that a policy of" restric-tions to limit the extraction of coal must be imposed. According to W.Stanley Jevons, ’we have to make the momentous choice between briefgreatness and longer continued mediocracy (1865, p. 349)’.

ECN-I--94-053 53

The 3-E interaetion and the rebound-effect

5.3 The Energy Policy polemic

Although already in 1865 it was argued by W. Stanley Jevons that energyefficiency improvements do net necessarily lead te equal reductions in thedemand fat energy, there is still na consensus regarding the magnitude ofthe rebound-effect today. This can be ifiustrated by means of a descriptionof the polemic in Energy Poliey, an international journal addressing theeconomie, environmental, and social aspects of energy supp]y and utfl]-zation.

The discussion in Energy Policy regarding the rebound-effect started wìthan article by Bill Keep]n and Gregory Kats (1988), entitled GreenhouseWarming, Comparative Analysis of Nuclear and Efficiency AbatementStrategies. In this article, Keepin and Kats argue that there are two possiblepolicy responses te the problem of the greenhouse-effect, namely theadaptation response and the amelioration response. The adaptation phi[o-sophy says that greenhouse warmfng is inevìtable and that man shouldtherefore begin te adapt te ir. The amelioration response states that theprob]em can be counteracted by reducing the dependence en fossil fue]s inthe future. Keepin and Kats are advocates of the latter option, and in theirpaper they examine two et~ the toost popu[ar strategies fat ame]ioratinggreenhouse warming through substantial]y reducing the use of fossil fuels inthe future. The first ot: these strategies te fight the greenhouse-effect is bydisplacing fossil fue]s with nuclear power. The second strategy te minimizethe combustion of fossil fue]s in the future is by energy efficiency improveoments. Regarding the revitafization of nuclear power, Keepin and Kats con-clude after analyzing various scenarios that ’even a massive worldwìdenuclear power programme sustained over a period of several decades couldnet salvo the greenhouse prob]em. And even if it couId, the Third Worldcannet support a major expansìon of nuclear power en the scale that wouldbe required in an attempted nuclear salut]en te greenhouse warming (1988,p. 539)’. With respect te energy efficiency, Keepln and Kats conclude that’energy efficiency ofl~ers the greatest promise te reduce global CO2 emis-sions substantially, whiIe also ame]iorating ether problems, such as acidrain and economie ineflïciency. Energy efficiency is the slng]e toost impor-tant technological factor determining future CO2 emissions. Moreover,rather than being just a theory, this efficiency potent]al has also heendemonstrated in practice. S]nce ] 973, the energy used per unit world econ-omie output bas dec]ined by 12%, primarily in response te increasedprices. This has occurred in the absence of vigorous efforts te promoteincreased efficiency in most nations, and only gives a hint of what would bepossible in the event of a concentrated effort te implement improved ener-gy efficiency worldwide (1988, p. 550)’. According te Keepin and líats, ’inthe USA, the wor]d’s largest producer of CO2, each dollar invested ìn elec-tric efficiency displaces nearly seven times as mueh CO2 as a dollarvested in nuclear power. Even if the most optimistic aspirations for thefuture economies of nuclear power were realized today, efficiency wouldst]Il disp]ace between 2.5 and ]0 times more CO~ per unit investment(1988, p. 554)’. Therefore, Keepin and Kats conc]ude that improvements inenergy efficiency constitute an effect]re and inexpensive response te thegreenhouse warming problem, whereas nuclear power is an ineffectìve anti

54 ECN-I--94-053


expensive response. Opportunities for energy efficiency must be the cuttingedge of energy policy throughout the world.

In a response to the Keepin and Kats paper, Len Brookes (1990) wrote TheGeenhouse Effect: The Fallacies in the Energy Efficiency Solution. Accor-ding to Brookes, energy efficiency improvements cannot by themselvessolve the greenhouse-effect. Brookes argues that ’what most advocates ofthe energy efficiency route to salvation overlook, is that there is no evi-dence that using energy more efficiently reduces the demand for ir. Theefficiency of conversion of fuel to useful heat or mechanical energy basimproved by an order of magnitude in the last 100 years. Also the averagethermal efficiency of power stations in the UK has improved from less than10% in the 1920s to more than 40% today. Yet we now consume moreenergy both in total and on a per capita basis than we did then (1990, p.199)’. According to Brookes, the reason for this is the fact that efficiencyimprovements lead to price decreases, which in tutu lead to demand in-

To clarify the issue, Brookes eonsiders two scenarios, one in which energyconstitutes a constraint on the leveI of economie activity, and one in whicbit does not. Regarding scenario 1, Brookes argues that if the price of ener-gy becomes a serious economie obstac]e (as it did in 1973 and again in]979), there are two possiblû courses of action. The consuming countrycan either do nothing and accept a reduction in the level of economictivity or it can adjust by substituting capital and labour for energy. NowBrookes argues that ’compared with the option of taking no remedial ac-tion, the option of accommodating to the price rise results in the balancebetween energy supply and demand being struck at a higher price, hence ahigher level of production and consumption of energy than if there hadbeen no energy efficiency response. In other words, in the case where ener-gy price is a constraint, responding by using energy more efficiently resultsin a higher level of energy consumption than if there had been no suchresponse, anti this is true whatever the motive (1990, p. 200)’. Regardingscenario 2, Brookes refers to the research done by Sam Schurr and DaleJorgenson. In this case where the energy price is hOt a constraint on econ-omie activity, two phenomena are taking place. These phenomena are thesubstitution of energy for labour and capital with the consequent improve-ment in their economic productivity, and a parallel improvement in theproductivity of energy, i.e. a falling energy consumption per unit of econ-omie output. Unfortunately for the advocates of energy efficiency, Brookesargues that ’these phenomena are accompanied by increases in total ener-gy consumption. The reason for this is that the substitution of energy forcapita] and labour ha~ such a beneficial effect on the productivity of thosetwo inputs that their combined rate of productivity improvement exceedsthe rare of improvement of energy productivity. And it is a íru|sm thatmultifactor productivity growth exceeds energy productivity growth whi]einputs of labour and ~apital are hot fal]ing, the net effect is for total energyconsumption to increase, even though energy consumption per unit ofoutput may be fal]ing (1990, p. 200)’. So, Brookes argues that in bothscenarios energy efficiency leads to an increase in energy consumption,rather than a decrease, compared with the situation without the energyefficiency improvements. Besides this, Brookes states that in investigatingthe residentia] sector, he found a very high correlation between income and

55


energy use. According to Brookes, ’domestic consumers tend to spend aconstant proportion of their income on fue! and e]ectrlcal energy, implyingthat higher efficiencies of appliances lead to higher levels of use and in-creases in the number of appliances. More generally, purchasing powerre)eased by ]ower expenditure on existìng uses of fuel finda an out]et some-where, and in modern industria] societies it is almost bound tobe in thepurchase of goods and services, that require energy in their production ifhot on other uses of fuel itse]f (1990, p. 201)’. From a]l this, Len Brookesconc]udes that the claims of conservationists such as Keepin and Kats arehOt true. As history proves, the enormous improvements in energy efficien-cy that have taken p]ace in the course of time have heen accompanied byincreases in energy consumption, not decreases, and this is tobe expectedin the [:uture as well. According to Len Brookes, ’the present high profile o~:the topic of g]obal warming and energy efficiency seema to owe more tothe current tide of green fervour than to sober consideration of the factsand the validity and cost of so[utions (]990, p. 201)’.

Len Brookes’ paper evoked various responses in the next editions of Ener-gy Po~icy. Support for Brookes view was given by Geoffrey Greenha]gh(]990) in his paper Energy Consemation Poli¢ies. According to Greenha]gh,the oil price rises of 1973 and 1979 ended an era of cheap energy and alsomade clear to pub]fc and govemments the vital ro]e of energy in the econ-omy. As a result of these price rises, a whole range of programmes, suchas energy audits, financial incentives, and the adoption of energy efficiencyregu]ations and standards were introduced by nationa] govemments inorder to save energy, The success of these measures was usuM]y measuredin terras of improved energy intensity, seen as a reduction in the energy toGDP ratio. But, ac¢ording to Geoffrey Greenhalgh, Ït is a fundamental mis-conception to think that a change in the energy to GDP ratio gives an in-dication of the efficiency of energy use. It does hOt. The ratio merelyi[lustrates changes in the use in which energy is put in a society, and theseuses oan change quite independent]y of any change in the actual efficiencywith which energy is used (1990, p. 294)’. Greenhalgh namely argues thatthere are a number of factors that could affect the energy to GDP ratio.One of these factors is the structural change within society from industria]production to the service sector. Another factor is the shift from buikproducts with a ]ow added value to more specia[ized products wlth a highadded vaiue. Besides the misconception that the energy to GDP ratio is anindicator of the energy efficien~y of an economy, Greenhalgh argues thatthe introduct[on of more energy effic~ent appllanoes leads toa reductlon inthe running costs and this in turn will lead toa greater usage. There is ofcourse nothing wrong with that. The use of these appliances satisfies theneeds of consumers. Greenhalgh however argues that the aim of conser-vationists such as Keepin and Kats is hOt consumer benefit but a reduetionin the consumption of energy carriers. According to Greenha|gh, ’the incen-tive to improve the efficiency of energy use is ever present, driven by bothtechnical advance and econom[c ga[n. Artff[¢[al attempts to promote ener-gy conservation as a policy in its own right, deliberately aimed at reducingconsumption, may oniy distort the energy economy, lnsofar as they aresuccessíul they will only accelerate the natural rare of progress, which leadsto reduced costs of using energy. This can only tend to increase the even-tual consumption. The end result may then be quite different to what theconservationists intend (1990, p. 298)’.

ECN-I--94-053


Opposition to Brookes’ thoughts was given by three authors. One of themwas Dave Toke (1990) in lncreasing Energy Supply hot lne~itable. Accor-ding to Toke, there does exist historical evidence that using energy moreeffi¢iently reduces the demand for it. The primary energy consumption inthe United Kingdom in ] 988 namely was 3% lower than in ] 973, whereasGDP was about 31% higher. Aceording to Toke, this was a result of energyefficien~y improvements in the economy as a whole over that time periodand precisely contradicts Brookes’ assertion. Toke further argues that it iswrong to claim that there is a constant relationship between people’s in-come and their spending on energy. To clarify bis opinion, Toke uses theexample of a consumer who decides to purchase a new televis[on: ’il ef-ficiency standards produce TVs that use only half the amount of electricity,it does hOt mean that people will watch TV any Ionger, never mind twice aslong. Sure, the money released by savings on energy bilis wfll be respenton other goods and services, but only a small portion of this will be for theenergy content of such goods and services (11990, p. 673)’. To end hisobjections to Brookes’ thoughts, Toke argues that Brookes pays too muchattention to respendin9-effeets and too little attention to analyzing thesible effectiveness of political intervention efforts to reduce the energy con-sumption. Brookes dismisses energy efficiency standards as a solutíon.Ac¢ording to Dave Toke, however, energy efficiency standards can drama-tically improve the level of energy efficiency, thereby eliminating markerimbalances and misallocations of resources. Toke claims that it is not un-realistic to think that with strong political intervention measures large re-ductions in the energy to GDP ratio can be achieved. Toke further arguesthat ’although technical anti economic considerations do set limits to whatcan be done, suoh limits are stricter in the short terra than in the long terraand at the end of the day there are a number of possible energy pathswhich can produce continuing increases in GDP figures but which incurrather different levels of external costs and impacts on the quality of l~re.We can make a political decision to adopt an energy efficient path that ismore effective in combatting the greenhouse problem than the conventionalnostrums of ever [ncreasing energy supplies based more and more on nu-clear energy (1990, p. 673)’.

in a response to Dave Toke’s critical remarks on Len Brookes’ paper,Brookes (1991) wrote Confusing the Issue on Energy Efficiency. Accordingto Brookes, the fact that the energy consumption in the UK fell by 3% bet-ween 1973 and ![988 whíle economic output [ncreased by 31% does notcontradìct bis conclusions. These results are namely entirely consistent withhis claims. The period from 1973 to 1988 contained both of the oil crises.It was therefore an excellent examp[e of scenario 1[, ín which energy supp[yand price were constraints on economic activity. Besides this, Brookesclaims that the data that Toke quotes are hot well chosen. For example, 19of the 23 OECD used more primary energy in 1988 than in 1973. Brookesalso argues that Toke deliberately selected 1973 as the base year in orderto prove his point. The year 11973 namely was a peak year for UK energyconsumption. Moreover, UK economie performance between 1973 and1988 was highly atypical because muoh of the economie growth then ~vasdue to the exploitation of the oil and gas reserves from the North Sea.Brookes summarizes his crìtique by stating that ’Toke has attempted torefute a general case, supported by soholarly research, by appealing toavery special case of an atypical country in a brief atypical period in its

ECN-I--94-053 57


economic history (1991, p. 185)’. Brookes further argues that it is wrong ofToke to think that a very high correlat~on between domestic energy con-sumption and personal disposable income cannot be found. And it is equal-ly wrong to clar~fy this thought w~th ~ single hypothetical example of ener-g’y effi¢ient TV’s. According to Brookes, the high correlation was a statis-tical]y derived fact. To con¢lude bis reply, Brookes again emphasizes that’there is no economic, environmenta], or any other kind of" purpose that isbest served by maximizing energy efficiency and therefore Toke’s energyefficient Utopia is a seriousiy flawed goal (1991, p. ]86)’.

To counteraet Brookes’ comments, Dave Toke (1991) wrote Energy Ef-ficiency. In this piece, Toke again expresses the thought that Len Brookes’statement that after an oil crisis energy demand is higher in the situationwith energy efficiency improvements than if there were no efficiency im-provements is absurd. During the oil crisis, the majority of industriesnamely invested more in energy efficiency improvements than they wouldotherwise have done. The results of these investments was that the demandfor energy ended up below what it would otherwise have been, hot above.Moreover, these efficiency gains even influenced the energy economy longaffer the oil crisis ended. According to Toke, ’Brookes’ analysis is com-pletely dominated by supp}y-side considerations, a stance which leads tobizarre ana]yses in an effort to prove tbat demand-side energy measuresare always ineffective in reducing energy demand. The issue of marketimbalances is completely ignored by Brookes who lives in bis own worldwhere the demand-side of the entìre energy sector can be represented by asingle demand curve with perverse movements (1991, p. 815)’. One of thecentral points in Toke’s reasoning ~s that the market imbalances whlchprevent a successfu! introduction of energy efficiency measures can besolved by govemment regulafions. Dave Toke argues that essentially allthat Brookes’ claims say is that a combination of energy taxes and energyefficiency standards is needed to so~ve the problem.

Critical notes regarding Len Brookes’ thoughts on the fallacy of the energyefficiency solution to the greenhouse warming problem were also expressedby Horace Herring and Mike Elliott (1990) in a Letter to the Editor. Accor-ding to Herring and E}liott, Len Brookes does hot grasp the central issue inthe discussion. This issue namely is whether or hOt energy efficiency im-provements are the best means of meeting the increased energy demand.Brookes argues that in the situation where energy is a constraint on econ-omic activity, responding by increasing the energy efficiency results in ahigher level of energy demand than if there had heen no response. Herringand Elliott however argue that ’this is not truc in the short terra nor underconditions of steadily rising energy prices. For elementary productiontheory implies that lower energy eonsumption can be achieved in the longrun even with improving energy efficiency provided that the real price ofenergy rises at a steady rare over time. It is also possìble to lower energyconsumption without price rises or reducing output by ímproving energyefficiency, but only at the expense of moving toa sub-optimal economy.That is that the same levei of output is achieved but at a higher cost thannecessary (1990, p. 786)’. Len Brookes grounds bis thoughts on the con-cept of multifactor productivity growth. But, according to Herring and]iott, ’whi]e the concept of mu]tifactor productivity is usefully applied inindustry, it must be remembered that industry in the UK accounts for on]y

58 ECN-I--94-053


28% of total delivered energy use. The concept is of less relevance to theother energy using sectors (1990, p. 786)’. Moreover, Herring and Elliottargue that there is another inaccuracy in Brookes’ reasoning, namely thefact that Brookes claims that an increase in income leads to an equal in-crease in the demand for energy. Herring and Elliott argue that the energyuse in the domestic sector has not significantly changed since 1940,despite the yearly increases in income. According to Herring and Elliott,energy efficiency improvements were responsible for this. Horace Herringand Mike Elliott conclude by stating that ’conservationists are on firmerground than Len Brookes imagines. For increased energy efficiency,brought about under conditions of steadily rising prices, has the merits ofsaving energy, stimulating production, and buying time (perhaps 10-15years) forus to carefully consider what energy sources are needed to re-place fossil fuels in the 21st century (1990, p. 786)’.

In a Letter to the Editor, Len Brookes (1991) responded to the criticalremarks made by Horace Herring and Mike Elliott. According to Brookes,Herring and Elliott attempt to reject bis conclusions expressed in TheGreenhouse Effect: The Fallacies in the Energy Efficiency Solution with thepoint that economically unjustified energy efficiency improvements canreduce total energy demand. Brookes argues that this point is irrelevantsince he explicitly mentions that economically justified improvement$ inenergy efficiency do not contribute to reducing the demand for energy.Furthermore, Herring and Elliott use the fact that steadily rising energyprices can result in a reduction in total energy demand to prove Brookes’wrong. According to Len Brookes, however, also this fact does hot enfeebiehis argument. Another point by which Herring and Elliott try to rejectBrookes’ opinion, is that energy efficiency improvements may result inshort-term reductions in the total demand for energy. Brookes replies tothis critical remark by arguing that the greenhouse effect is surely not ashort-term problem. Brookes ends bis Ietter to the editor by remarking that’ir is truly astonishing how tenaciously the faithful stick to their belief inenergy efficiency as a cure for all ills. Herring and Elliott’s letter is no ana-lysis worth the name: it is simply a series of affirmations of beIief by thefaithful (1991, p. 187)’.

A third opponent to Brookes’ thoughts was Michael Grubb (1990). In Ener-gy Efficiency and Economic Fallacies, Grubb agrees with Brookes that ifenergy becomes a serious economic obstacle (scenario 1), improving ener-gy efficiency will hot reduce the demand for it. In the much more usualsituation where energy is not an obstacle to economic activity (scenario 2),historical data at least up to 1973 indicate that although energy produc-tivity improved, the labour and capital productivity improved much faster.As a result of this, efficiency improvements were more than offset by thenew uses for energy. From this, Brookes concludes that energy efficiencyimprovements have and will always increase the attractiveness of energyfor new uses. Consequently, improvements in energy efficiency cannotreduce energy demand. In a response to Brookes’ thoughts on scenario 2,Grubb writes that ’while the historical analysis is fine, the final step inBrookes’ arguments contains at least two major flaws. The first flaw is toassume that future economic reactions will parallel those of the past. Thelast decades have shown great changes in the pattern of energy and econ-omic development. A long era of steadily declining fossil fuel prices will

ECN-I--94-053 59

_ II I I I III I ~ r ..............


never return, and there is much evidence that some important end-uses ofenergy are approaching saturat[on. Also the pattem of economie develop-men(. [s shifting towards less mater~al- and energy-intensive goods. There-fore, general statements about the energy foundation of future economiegrowth, largely based on pre-lg73 data, are questionable (lgg0, p. 783)’.With respect to the second shorteoming in Brookes’ analys~s, Brookeswrites that ’a more serious flaw is to confuse the role of naturally-occurringefficiency improvements with the effects of deliberate attempts to minimizeenergy consumption when price and availability are not constraints. In al|the period up to 1973, improved energy efficiency was primarily a conse-quence of other pursuits, for example the drive to colonize new markets, byintroducing automated processes in which the energy costs were lower thanthe labour costs. It is not surprising that this led to demand growth. But ifenergy efficiency is the goal, rather then the means, then the driving forceis entirely different, and so are the macroeconomlc implications. Considerfor example a sector where the energy-using activity is relatively unrespon-sire to price. In the course of naturel developments there may be |ittle in-centive for efficiency to improve, irrespective of technical potentiaL Yetthese are precisely the sectors which are likely to o[:fer the greatest energysavings. Conservation policy een choose to focus upon such areas, inclu-ding those dominated by marker failures, where the implicit price falls fromincreased efficiency have litt]e impact on activity ]evels (1990, p. 783)’.Qrubb thus argues that there are many areas in the economy in whiehenergy efficiency improvements can save energy because the price of ener-gy is of minor importance. But there is more. IBrookes argues that energyefficiency improvements lead toa higher purchasing power because ofIower expenditures on existing uses of energy. Thls purchase power tìndsan outlet somewhere in the purchase of goods and services that requireenergy. But, according to Michael Grubb, total energy ¢osts are generally afew per cent of GDP and this respending effect will also be of th[s order.Moreover, Grubb disagrees w[th Brookes’ thought thet dornestic consumerstend to spend a constant proportion of their income on gas and electrícity.Ac¢ording to Michael Grubb, a LIK F’amily Expenditure Survey name[yshowed that poot households spend a greater portion of their income onenergy than average households even ìn absolute terras, because pootpeople do not have the financial means to insulate their homes and buyenergy efficient appliances. Grubb argues that the major problem of theenergy efficient solution is not the rebound-effect but the fact that withrespect to energy efficiency serious marker imperfections exist. Thesemarker failures result in an inefficient use ot: resources and higher thanneeded costs to society. As reasons for this inefficiency, Qrubb mentionslack of knowledge, know-how, and technical skills, separation of expen-diture and benefit, limited capital o[ten arising ffom external restrictions oncapital budgets, rapid payback requirements, lack of interest in peripheraloperating costs, and legal and administrative obstacles. According toMichael Grubb, ’the resu]t of all these factors is to create a pervasive im-balance between investments in supply and investments in end-useefficiency. Therefore, policies airned at removlng or clrcumventing thesemarket obstacles and installing efficient and cost-effective technologies,save energy and bring both env[ronmental and economie benefits. Thìs isnot wishful thinking but the a]most universal conclusion from those whohave studied the realities of the economie imbalance between supply anddemand (1990, p. 785)’.

60 ECN-I--94-053


Len Brookes (1992) responded to Michael Grubb’s ~riticism by writingEnergy Efficiency and Economic Fallacies: a Reply. Brookes starts bis re-sponse by stating that he will confine himself to scenario 2 (when energy ishot a constraint) since Grubb agrees w[th him on scenario 1. Regarding thesecond scenario, Grubb states that Brookes’ conclusions are wrong if im-proved energy efficiency is itself the goal, rather than the means to someend. Grubb claims that in such circumstances real energy savings andeconomie benefits will accrue at the macroeconomi¢ level. Brookes rejectsthis thought because microeconom~c measures cannot simp]y be extendedto macroeconomic results. According to Len Brookes, ’comparing theeconomics of different options for reducing economy-w[de or even world-wide emissions of greenhouse gases due to fuel-using activities is a macro-economic exercise that calls for macroeconomic treatment. The results ofadopting an energy efficiency solution can only be judged at the macro-economic level (1990, p. 390)’. Furthermore, Brookes dismisses Grubb’sclaims because of the fact that Grubb is reversing the debate, by setting themaximization of energy efficiency as a goal in its own right. According toBrookes, it ìs wrong to state that the starting point of the discussion shouldbe the acceptance of maximum energy efficiency as the goal. Brookesargues that ’ir would be reasonable to set the maximization of economieoutput as a goal; or the minimization of environmental damage; or themaximization of output subject to there being improvements at a given rarein defined environmental parameters. The maximizafion of energy efficien-cy does not result in any of these admirable goals and it would bias optionsin a damaging way to pretend otherwise. It would also divert attention fromtruly uset:ul subsidiary goals, like setting targets for reductions in green-house gas emissions (1992, p. 392)’. Besides this, Brookes also arguesthat when Grubb’s goal is maximizing energy efficiency per se, energybecomes a scarce resource. This implies that Grubb’s version of scenario 2~s in fact an exarnple of scenarìo 1, with energy efficiency taking the placeof the OPEC. Another point of controversy is the statement by Brookes thata high correlation exists between personal dísposable íncome and residen-tial energy expenditure, which means that people respend the money thatis saved by efficiency improvements on energy. Michael Grubb argues thatthis is not true because poor people spend a ]arger proportion of theircome on energy than rich people. According to Len Brookes, this fact isinvalid for this d[scussion. The research by Brookes was a t[me sedes re-gression of total residential expenditure on energy upon total UK disposableincome, whiIe Grubb’s argument merely il[ustrates behaviour changesacross socioeconomic groups at a certain point in time. Brookes ends hisresponse by agreeing with Grubb that a lack of knowledge on the part ofthe consumer is a source of economie inefficiency. Brookes argues that ’ilconsumers are to make the right choices when faced with a constraint, saya forto of carbon[ferous fuel rationing or taxat[on, [t is important that theyshould be informed ones. A comprehensive approach to labelling in all itsmanifestations (showing the consumption level of appliances for example)is an important means to that end (1990, p. 392)’.

In bis Reply to Brookes, Michael Grubb (]992) put aside Len Brookes’comments. A large part of Brookes’ paper is about the accusation thatGrubb promotes energy efficiency as a goal rather than as a means tosome end. According to Grubb, thia is a strange accusation sin~e a largepart of the paper Energy Efficiency and Economie Fallacies is concerned


with the conditions under which efficiency improvements become economi-cally attractive. To clarify bis point of view again, Grubb argues that ’poli-cies for promoting energy efficiency should be pursued until the marginalcost of squeez[ng more out of efficiency equates to that of new supply(1992, p. 393)’. Another matter of dispute is whether the rebound-effectshould be viewed in a microeconomic or macroeconomic context. Accor-ding to Michael Grubb, the macroeconomic context is important butBrookes is using generalized macroeconomic results in ways that are ir-relevant in the context of this discuss[on. Take for example a fridge. Accor-ding to Grubb, competent microeconomic analysis will prove that the priee-elasticity of fridge use is ra’ther low. This means that if 1 buy a much moreefficient fridge, I wil] not react to the decreased operating costs by leavingthe fridge door open more often. So, according to Grubb, this effect isnegligible. Grubb further argues that competent economic analysis can alsoprove that of the money that is saved in this way, on]y a small fraction willbe respent on energy, implying that this effect is small as well. Grubbc]oses his reply by asserting tbat ’Brookes neglects the central issue,namely the extent to which the measures intended to help close theefficiency gap wilI result in reaIízed energy savíngs. That is nota questíonthat can be answered by attempting to raise macroeconomic generaliza-tions to the status of precluding microeconomic realities or by misrepresen-ting those who are trying to clari~y the issues ìnvolved (1992, p. 393)’.

In Energy Efficiency Fallacies: the Debate Concluded, L. G. Brookes (1993)responded to Grubb’s reply to Brookes’ repIy to Grubb’s reply to Brookes’paper The Greenhouse Effect, The Fallacles in the Energy EfficiencySolution. According to Len Brookes, Grubb has hot mentioned newarguments which support his claims. In the first place, Grubb denies that hethinks that energy efficiency is an end, instead of a means to an end. But,according to Len Brookes, in bis reply Michael Grubb states that energyefficiency should be improved until the marginal ¢ost of efficiency equatesto that of new supply. This thought proves Grubb’s prepossession. Accor-ding to Brookes, ’il we are looking at the issue economicalIy, the objectivefunction tobe maximized is not the energy efficiency but the efficiency ofthe economy as a whole. If consumpfion of certain types of fuel, for exam-p]e, is causing probiems, the toost direct and economica~ly soundest so[u-t[on is to constrain consumption of that fuel by some form of rationing orby taxing it, and than reoptim[ze the objective functlon of the economicefficiency as a whole subject to the new constraint. Neither Grubb nor anyother like minded writer bas put forward cogent arguments to show why itis right to concentrate all the adjustment on energy-efficiency improve-ments (1993, p. 346)’. Besides this, Brookes argues that Grubb offers nosupport in bis reply for his claim that Brookes’ macroeconomic approachto examining the problem is invalid. It is not sufficient to mention a singleanecdotica~ microeconomic example of the purchase of a refrigerator tocounter a serious macroeconomic argument. To end bis reply, Len Brookesasserts that he does hot want to condemn the foIlowers of the energy ef-ficiency movement. Brookes is hot opposed to energy efficiency improve-ments and he enjoys the benefits of them at both the microeconomic andmacroeconomic leveL Brookes admits that he likes having appfiances withlower operating costs and he welcomes the higher levelì of national incomethat follow from the economy-wide improvements in energy productivity.But, according to Len Brookes, all this does hot imp]y that improvements

62 ECN-I--94-053


in energy efficiency can be seen as a means to reducing total energy con-sumption and as a solution to the greenhouse warming problem.

After Brookes’ piece, it is said in an editorial note that this was the finalpiece in the debate on the rebound-effect. It is namely clear that the va-rious authors cannot come to an understanding on the issue. This is at thesame time the most apparent conclusion that can be drawn from thispolemic. Those who believe in a large rebound-effect are of opinion that,for instance, increased car efficiency will lead to an increased demand fortravelling and an increased demand for other goods and services, whiehcontain a considerable amount of energy. Those who do hot believe in theexistence of a large rebound-effect think that increased car efficiency willhOt lead to more travelling and an increased energy demand resulting fromthe respending of the money that is saved by the increased car efficiency.The whole question thus remains unsolved and highly speculative. Thereason for this is the fact that no real attempts are undertaken to ca]culatethe magnitude of the rebound-effect. The only way to reach agreement onthe issue of whether or hot a large rebound-effect exists is by means of aquantitative analysis.

5.4 The effectiveness of energy efficiencyimprovements: a case study

The energy required for domestic use represents an important part of thetotal energy consumption of a country. A considerable amount of this do-mestic energy consumption is used for the heating of dwellings. Accordingto van Rossum (1991A,1991B), space hearing accounted for 65% of thetotal Dutch residential energy consumption in 1985. It is theret:ore hot sur-prising that governments see energy conservation by means of reducingthe energy required for space heating as an important means in meetingtheir environmental policy goals. For example, in the National Environmen-tal Policy Plan of the Netherlands which appeared in 1989, energy conser-vation through increasing the thermal efficiency of dweIlings is one of themain options to reduce the emissions of carbon dioxide, lnsulation is themost important means of changing the thermal efficiency of homes. Forthis reason, the increase in the thermal efficiency of an existing home(retrofitting) by means of insulating the cavity walls is the subject of thiscase study.

In any dwelling, heat transfer between the inside and the outside takesplace by three separate agencies. These agencies are çonductlon, convec-tion, and radiation. When heat goes through solid materials, the term con-duction is used. A measure of the rare at which conduction happens is thecoefficient of thermal conductivity. The thermal conductivity of metals isvery high. The ¢onductivity of aluminum for example amounts to 210W/mK at 0°C (van Buuren and de Jong, 1990). Non-metals have muchlower thermal conductivities. The lowest thermal ¢onductivities are mea-sured by materials that contain a considerable proportion of voids. Thesevoids are small bubbles of gas or air. Examples of such materials are mine-tal wool and plastic foams.

EeN-I--94-053 63

The 3-E interaction anti the rebound-effect

Now ¢onsider the walls of a dwelling. In the case of heat transfer by meansof conduction, the heat that passes from one area of the wall to the othercan be described by the fo]lowing equation:

This emp/rica] relation is only valid in a steady state situation, in whieh theheat eapacity of the wall is neglected. In equation (5.1), the symbol Qconddenoïes the heat in W/m2 íhat passes through the material, the symbol d

denotes the thickness of the material in metres, and ~XT represents thetemperature difference between the two surfaces in degrees Kelvin. Thesymbol X in equation (5.1) denotes the coefficient of thermal conductivity.The dimension of this coefficient is in W/mK. In English speaking countries,the symbol k instead of Z is used to indicate the therma] conductivity coef-ficient.

The thermal conductivity of air is very low, namely 0.020 W/mK at 0’C(van Buuren and de Jong, 1990). When a layer of air has an apprec[ablethickness, this X-value however only has a pure]y theoretical meaning. Thereason for this is a process called convection. Convection is the transfer ofheat by the movement of the molecules of a liquid or a gas. An example of~onvection is the transmission of heat from the inside air to the internal wallof a dwelling and flora the external wall to the outside air. The movementof the liquid or gas molecules can be caused by an external agency, su¢has a pump in the case of liquids or the wind in the case of gases. This iscalled forced convection. The movement of the molecules can also takeplace simply because apart of the liquid or gas is warmed up. This war-ming up results in a change in the density of the gas or liquid. As a resultof this, the heated molecules start moving. This is called free convection.

Convection can only occur through a medium. For radiation, no medium isrequired. In fact, heat transfer by means of radiation is most favourableacross a vacuum. Radiation heat transfer appears whenever there are twosurfaces at different temperaturés faeing each other, or if there is a singlesurfaee fa¢ing an open space. Radiation proceeds at the speed of light,which is about 300,000 kilometres per second. Radiant heat is a form ofelectromagnetic radiation. This means that it is related to phenomena suchas infra-red light, visible light, and X-rays. An example of radiation is thehearing up of the external walls of dwellings during day-time when the sunrails upon them. Another examp]e is that of the heat which is radiated froma fireplace on to the internal walls of a house.

Although complex equations are derived for the heat transfer by means of¢onvection and radiation, within the building industry the two are frequentlylumped together in the following empirical relation:

64 ECNd--94-053


Just Iíke equation (5.1), this equation is only valíd in a steady state. Inequation (5.2), ~conv/rad denotes the heat in W/mE that is transferred,and AT represents the temperature difference. The symbol o~ denotes theheat transfer coefficient, which is expressed in W/m¢K. This heat transfercoefficient is composed of two components, name]y a convective terra0~oonv and a rad[ative terra O:r~~. The magnitude of 0~~onv depends on thespeed of the air flow. Ac¢ording to Boeke and van de Kraaij (lg77), avalue for O.co~v of 2 W/rocK can be used inside a building. Outdoors, thefollowing approximate relation holds:

U.~o~u = 6.2 + 4.2 " u (5.3)

In this equation, the symbol v denotes the air velocity in metres perse¢ond. Equation (5.3) illustrates that an air flow with a velocity of, forexamp[e, 3 metres per second corresponds with a vaIue for 0too~v of around19 W/m2K. Regarding the radiative terra, it can be said that a value of 5W/m2K (Jaeckel and Okken, ]978) can be used for normal temperatures.

blow consider the heat transfer from a heated dwelling to the environment.The total heat flow through a wall can be e×presseò by the followingequation:

1(~tot = 1 .-. d 1 " ~ T

(~~

(5.4)

Equation (5.4) can also be written as:

Qtot -- k ¯ k T (~.5)

In this equation, the so-called k-value is expressed in W/m2K.

To quantify the effect of cavity wall insulation on the demand for naturalgas, first the k-values of the walls of a dwelling in the situation withoutinsulation and with insulation must be determined. After that, the differencein the k-values of these two situations can be used to determine the annualgas savings. For the calculation of the k-values, a traditional double brickwal1, consísting of two layers of 105 mm thick brickwork with a X-value of0.7 W/mK (Jaeckel and Okken, 1978) is considered. In between the twobrick layers, there is an air cavity of 60 mm that can be filled with in-sulation materials. On the inside, the wall is covered with ]5 mm of gyp-sum plaster with a therma! conductivity of 0.9 W/mK (Jaeckel and Okken,1978). For the inside of the more, a heat transfer coeff[cient ~i of 7 (= 2 +5) W/m2K can be used. When there is a wind of 3 metres per second blo-wing against the walls, the heat transfer coefficient for the outside 0t~ be-comes 24 (= 19 + 5) W/m2K.

ECN-1--94-053 65


When the air cavity is not filled with insulation materials, the k-value of thewall becomes:

0.105 0.105 0.015+ 0.10 + + + --0.7 0.7 ~ 24

k = 1.663 (5.6)

In this equation, the 0.10 m~K/W (Jaeckel and Okken, 1978) representsthe total resistance of the air cavity.Now the air cavity is filled with mineral wool with a coefficient of thermalconductivity of 0.035 W/mK (Bowd[dge, 1990). The k-value ~n this situa-tion becomes:

1 1k = 1 + ~d~ + 1 1 0.105 0.06 0.105 + 0.015 ÷ 1

7 + ó7 +o.-ô~-~ + o.---7- o.--W- 2~"

k = 0.451

The difference in the k-values without and with insulation then becomes1.212 (~ 1.663 - 0.451) W/m~K. This difference can be used to calculatethe annual gas savings in the following way:

‘SG = A ¯ 24 ¯ 3600 o DD . ‘sk (5.8)Ho~I

In this equation, A denotes the inside measurement of the wall sufface, DDdenotes the average annual degree day value, H denotes the lower hearingvalue of a normal cubic meier of natural gas, and 1i denotes the boilerefficiency.

For the average amount of degree days, a value of 3000 degree days peryear (Macdaniel, 1980; Jaeckel and Okken, 1978) can be used when theinside temperature is set at 18"C. The lower hearing value of a cubic meterof gas amounts to 31.65 MJ. When the interior surface of the dwelling is120 m2, and the boiler efficiency is 0.85, the annual gas savings become:

,50 = A "24 "3600 "DD .‘5 k = 120 ’24 "3600 "3000 . 1.212H ¯ ~1 31.65"10~ - 0.85

‘50 = 1401.28 Nm3 per year (5.9)

The insulation of the cavity walls of this dwelling thus leads to annualsavings of around 1400 Nm~ gas. These savings agree well with other cal-culation$ of the natural gas savings resulting from cavity wall insulation,such as the studies by Jaeckel and Okken (1978) and Macdaniel (1980).

66 ECN-I--94-053

The rebound-effect: an introcluction

According to the Consumers Association (1990), the lifetime of the mineralwool that is used for the insulafion of the cavity walls can be set at 20years. This means that 28,000 (= 20 * 1,400) Nm3 of natural gas will besaved over the lifetime of the insulation material. This is also the traditionalengineering esfimate of the energy savin9s resulting from cavity wa]l in-sulation. In reality, however, the total energy savings will not amount to28,000 Nma of natural gas or equivalently to 886,200 Megajoules (=28,000 * 31.65 MJ), because apart of the energy savings is lost. The rea-son for this is the fact that cavity wall insulation is a very profitable energyconservation option.

According to the Consumers Association (1990) and Ybema and Okken(1-992), the costs per square metre of cavity wall insulation amountsguilders. The costs of insulating the whole dwelling thus become 2,520 (=120 * 21) guilders. The annua~ gas savings resulfing from these insulat/onmeasures are 1_,400 Nm~. When the price per Nm~ of natural gas is set at0.50 guilders (Krachtkroniek, 1994), the annual amount of money that issaved becomes 700 (= 1-,400 * 0.50) guilders. This money will be respenton various goods and services. All of these goods and services contain anamount of energy. Consequently, respending the money saved by thesulation measures, results in a ]?art of the energy savings being lost. DePaauw and Perrels (1993) studied the energy content of goods and serviceson a fairly detai]ed level. The results of their analysis indicate that the ave-rage energy content of goods and services amounts to 6.5 Megajoules perguilder spent. Now assume that the money that is saved by the insulationmeasures is respent on the average good or service with an energy contentof 6.5 MJ/guilder, and assume that this average energy content will hotchange over the next 20 years. Further assume that the price changes overtime of this average good or service are the same as the price rises of natu-tal gas. Then the rebound-effect in this particular situation can be deter-mined.

The annual gas savings are 1400 Nm~. This corresponds with 700 guilders.When the house-owner pays the 2520 guilders for the insulation measuresat the beginning of the 20-year lifetime of the insulation materials, themonetary savings in the first year will become -1820 (= 700 - 2520) guil-ders. This implies that besides saving 700 Nm~ of natural gas, the house-owner w[ll not be able to spend 1820 gu[lders on goods and services, con-taining an amount of energy. From this, the gas savings in the first yearcan be calculated. These savings namely amount to 56,140 MJ. This num-ber is the addition of 44,3]0 (= ]400 * 31.65) MJ for the annual gassavings and 11,830 (= 1820 * 6.5) MJ for hot being able to spend 1820guilders on the average good or service. In the other 19 years, the house-owner will save 1400 Nm3 gas and 700 guilders. The annual energy savingthen become 39,760 MJ. This is an addition of 44,310 (= 1400 * 31.65)MJ for the annual gas savings and -4550 (= -700 * 6.5) MJ for respendingthe money saved by the insulation measures on the average good or ser-vice. The total energy savíngs over the lífetime of the insulation material of20 years then become 811,580 (= 1 * 56,140 + 19 * 39,760) MJ.

The installat~on of cav[ty wall insulation will thus lead to energy savingsthat amount to 811-,580 MJ. The engineering estimate of these energysavings higher, namely 886,200 MJ. The difference in the calculation of the

ECN-1--94-053 67

The 3-E interact[on and the rebound-effect

energy savinga between the engineering estimate and the real energysavings is caused by the rebound-effect. This feedback-effect can be de-fined as that part of the initially expected energy savings that is lost be-cause of the energy-economy-environment interactiOno In formule:

rebound-effect = lost energy sauingsexpected energy savings

(~.1o)

In this equation, the expected energy savings represent the traditional en-gineering estimate of the energy savings, resulting flora energy efficiencyimprovements. This engineering approach does net take into account thefact that energy efficiency improvements lead te respending effects. As aresult of this, the engineering approach tends te overestimate the net ener-gy savings resulting from imprevements in energy efficiency. In this par-ticular case of cavity wall insulation, the engineering estimate of the energysavings is 866,200 MJ. Because of the respending of the money that issaved by the insulation measures, in reallty only 811,580 MJ of energy issaved. The lost energy savings thus amount te 74,620 (= 886,200 -811,580) MJ. With these lost energy savings and the expected energysavings, the rebound-effect can be calculated in the foIlowing way:

lost energy savings 74,620 = 0.084rebound-effect = expected energy savings 886,-~Ô’Ô(5.~1)

This means that more than 8% of the expected energy savings is lost be-cause of the feedback from the engineering measures te the economy.

Besides the fact that apart of the initially expected energy savings is lostbecause of respending effects, the net energy savings resuiting from cavitywall insulation are in reality even smaller. The reason for this is the fact thatthe mineral wool f.hat is used for the insulation purpose must be manufac-tured. Also the raw materials that are needed te produce the insulationmaterial must be mined and transported te the fa’¢tory. Moreover, the in-sulation material must be distributed flora the factory floor te the locationwhere it will be used. All of these steps require an amount of energy. Inaddition te this direct energy requirement, there is an indirect energy re-quirement in the production of the mineral wool. This indire~t energy re-quirement, for example, involves the energy needed for the construction ofthe plant and the production processes. According te Boustead and Han-cock (1979), the energy involved in the direct manufacture of the mineralwool insulation material amounts te 14 MJ/kg. New assume that another14 MJ/kg is required, either directly or indirectly, in the rest of the entireprocess. Then the total energy requirement for the mineral wool becomes28 MJ/kg. In this case study, a traditional double brick wall with an aircavity of 60 mm is considered. It is assumed that the total surface of thewalls amounts te ]20 m2. This means that 7.2 (= 120 * 0.06) m3 of mine-rel wool is needed te insulate the cavity walls in this particular case. Accor-ding te Bowdidge (1990), a characteristic density of 28 kg/m3 can be usedfor the mineral wool that is used for cavity wall insulation. The 7.2 m~ mi-neral wool that is used in this case thus corresponds with 201.6 (= 7.2 *28) kg insulation materiaL New recall that the energy content of 1 kg ofmineral wool amounts te 28 MJ. Then, the amount of insulation materlal

68 ECN-I--94-053

The rebound-efI:ect: an introduetion

that is used to fill the air cavity in this particular case represents an energycontent of 5645 (-- 28 * 201.6) MJ. So, when besides the energy that islost because of the respending of the money that is saved by the insulationmeasures, a]so the energy content of the insulation material is raken intoaccount, the lost energy savings become 80,265 (= 74,620 + 5,645) MJinstead of 74,620 MJ. This in turn implies that the magnitude of therebound-efl~ect also changes:

rebound-effect = lost energy savings =_80’265 = 0.091 (5.12)expected energy sa~ings 886,200

The rebound-effect thus becomes somewhat larger than in the situation inwhi~h the energy content of the insula~.ion materia] is hot considered.

Now one cou]d argue that the price increases of natural gas and the ave-rage good or service wil] probably not remain the same. And one cou]dalso argue that the average energy intensity of goods and services willprobably not remain constant at 6.5 MJ/guilder in the ¢ourse of time. Thishowever does hot change the idea behind the quantification of the rebound-effect. Besides th!s, it is very difficult to predict the price of a Nm~ of natu-tal gas and the average energy intensity of goods and services in the fu-ture. It is, for example, frequently urged by numerous authors that the aver-age energy intensity per guilder spent will decline in the future, as a resultof the shift from the energy-intensive industrial society to the society inwhich the energy-extensive service sector is dominant. Other authors, suchas Gilmer (1977) and Gershuny (1977,1978), however argue that the ser-vice sector itself is rather energy-intensive and that the energy to GDP ratiowill hot decline in the future. Therefore, a constant average energy intensityof 6.5 MJ/guilder is used throughout the planning horizon in the calculationof the rebound-effect in this case study.

ECN-I--94-053 69

6. THE REBOUND-EFFECT AND THECONSUMER

6.1 lntroduction

In the preceding chapter, an introductíon into the rebound-effect bas heengiven. In Ibis chapter, the rebound-effect is outlined for theconsumption-side of the economy. This rebound-effect results from im-provements in the energy efficiency of appIiances. In the context of therebound-effect, the terra app]iance should be interpreted in a generic sense,in order te include all energy-using durables. This means that besides ap-pfiances such as TV’s and dishwashers, also durables such as cars andcentraI hearing systems must be considered. Paragraph 6.2 dlscusses theshortcomíngs of the traditional engineering estimation of the effect of ener-gy efficiency improvements of appliances en the demand for energy. Inparagraph 6.3, the substitution and income effect are expIained. Peragraph6.4 contains a simple consumer behaviour model.

6.2 Mechanical determination of energy efficiency

The discussion of the consumption-side of the rebound-effect starts withthe energy demand mode]s of the late 1960s end early 1970s. Aíter tbc oilcrisis et: 1973, serieus doubts were expressed regarding the accuracy withwhich these models predicted the consumer demand for energy. An impor-tent shortcoming of the then existing models namely was thet they wereuneble te quantify the effects of energy efficiency improvements of applian-ces en the total demand for energy. The oil price hike resulted in an aware-ness of the ìmportance of energy conservation. The pre-oil crisis medelswere only capable ~f predicting demand under business-as-usual con-ditions, in which energy efficiency trends were extrapolated from historicalpatterns. During end a~er the oil price hike, however, the historical pattemregarding energy efficiency was fundamentelly altered.

As a solutíon te the dissatisfactíon with the energy demand models thatexisted before the oil crisis, models were developed that could capture theeffects of energy efficiency improvements en the total demand for energy.Because a large part of the efficiency improvements were a resuIt of chan-ges in the engineering characteristics of appliances, it was argued that onlyso-called engineering models could accurately predict the impact of con-servation measures en the energy demand.

But also the engineering approach does net correctly quantify the impact ofconservation measures en the demand for energy. The reason for this isthat the determination of the impact of energy conservation on total energydemand by engineering models is essentially, as J. Daniel Khazzoom(1980,1986,1987,1989) callsprovement of appliance efficiency by some percentage, engineering modelspredict that the energy demand of that eppliance wfil drop by an equalamount. So, when the average appliance efficiency increaseS by 1%. en-

71

I I ] II ~ ......

The 3-E interaction anti the rebound-effect

gineering models prediet that as a result of ibis, energy demand will drop to99.01% (=I/I.01) of its original leve]; when the average efficiency goes upby 2%, demand is predicted to drop to 98.04% (=111.02) of its originalleve]; and soon.

At fìrst sight, this thought appears tobe plausible. And foran engineer in aresearch institute, a change ]n energy efficiency should lead to an equal butopposite change in the demand for energy. The engineer namely assumesthat the same amount of service will be províded by a more efficient ap-pliance. Bui while this thought may be truc in an isolated research environ-ment, it cannot simply be extended to society. One cannot ignore the im-pact of engineering measures to improve the energy efficiency on society,as if these conservation measures live a lire of their own:

The reason for this is that besideì the engineering side, there is also aneconomie side to the prob]em. The purchase of a more energy efficientappli~nce namely bas impIications for the price per unit of service that theappliance deIivers. For example, investing in energy-efficient Iighting wil}have consequences for the price per unit of il]umination. Also buying amore efflcient air-conditioner will ]ead to a change in the price per unit ofcooling. Another examp]e is the fact that the purchase of a more fuel-effi-cient ear wil] lead toa change in the price per vehicle-kilometre. Becauseof these changes in the price per unit of service, consumers wil] react bychanging the way in which they spend their money.

Since traditional engineering models neglect the fact that changes in ap-pliance efficiency also have an economie component to them, the traditio-nal engineering approach exaggerates the energy savings that result fromimprovements in energy efficiency. More realistic models should take ]ntoaccount the fact that energy efficiency influences demand in the same wayes, for instance, income or the price of fuel. To achieve this, an economiecomponent must be added to the engineering models to allow for a feed-back effect (the rebound-effect). When this effect is taken into con-sideration, economie forces that are the result of efficiency improvementscan be determined. These economie forces result in e modification of theinitially predicted energy savings.

6.3 The substitution and income effect

Consumers will react to changes in the energy efficiency of applianees.This can be clarified by ineens of an example, eonceming the purchase ofa more efficient applianee, such as an air-conditioner. It is assumed thatthe purchase of this applianee will result in a deerease in the price per unitof cooling, which includes both the energy costs and capital costs. Thisineens that besides an amount of energy, the consumer will also save anamount of money.

When the price per unit of service of an appliance decreases, the appliancewílI become less expensive relative to other goods and services. This maylead to an increased demand for the service that this appliance delivers. Ineconomic theory, this type of rebound-effect is eailed the price or substitu-t[on effect. Empir[cal evidence for the faet that efficiency improvements

72

The rebound-effect and the consumer

lead to demand increases seems tobe provided by Clifton T. Jones (1_993),who studied the effect of increased motor vehìcle efficiency on the demandfor travelling. The results of Jones’ investigation namely ìndicate that about30% of the energy-saving benefits associated witb fuel efficiency improve-ments of cars are offset by an increased travelling demand, which is initia-ted by these efficiency improvements. Other evidence of this type ofrebound-effect was given by Adams and Rockwood (1983), who conc]udefrom monitoring house-owners who insu)ated tbeir dwe]]ings that a 20%increase in the thermal efficiency of residential bui]dings on]y yie]ds 8%energy savings in gas-heated homes anti 14% savings in electricaily-heatedhomes. Thís feedback-effect is a result of increases in the temperaturesetting and other behavioural changes, Simflar results are a]so obtained byHirst and White (]985). After analyzing specific e]ectricity and gas billingdata for severa! years, they conc[ude that between 5% and 25% of the ener-gy savings due to retrofittíng is taken back in terras of increased comfort.

The substitution effect however does not tell the who]e story. A decrease inthe price per unit of service of an appliance name]y hot only increases theservice demand of that app[iance, but it may also lead to an íncreaseddemand for other goods and services. This type of rebound-effect is ca]]edthe income effect iD economie theory. The income effect stems from thefact that the purchase of a more efficient appliance leads to an amount ofmoney beíng saved. This money, which can be seen as additional income,wil] ultimate]y be respent on various goods and services, all containing anamount of energy. An example of this type of rebound-effect is the foct thatthe money that is saved by the lower running costs of the air-conditioner isrespent on increased car use. A[so the case study of chapter 5, in whichthe money that is saved by cavity wall insulatìon is respent on averagegoods and servìces, is an example of the income effect. According to dePaauw and Perrels (]993), the average amount of energy per guilder ex-penditure amounts to 6.5 MJ/guilder. Therefore, this type of rebound-effectis hot of a negligible magnitude.

Besides these two causes, there may also exist another possible cause forthe rebound-effect. This cause refers to the fact that energy efficiency im-provements may lead to a reduction in the demand for energy carriers,such as oil and gas. As a result of this, the pressure on energy supply mayrail and the prices of oil and gas may drop, thereby ìnducing an increaseddemand for these energy earriers. The problem with this rebound-effect isthat the effect of a decreased demand for oi~ and gas on the prlce of theseenergy carriers is difficult to predict. Therefore, this type of rebound-effectwill hot be considered in this study.

6.4 A simp]e consumer behaviour mode]

As a result of the substitution and income effect, energy efficiency improve-ments do hot lead to proportionate reductions in the demand for energy.The underiying reason for this rebound-effect is the fact that consumers dohot want to minimize energy costs but rather want to maximize utility. Thiscan be illustrated with the following consumer behaviour model In thismodel, the purchase and utilization of an energy-using appliance such as

ECN-I--94-053 73


an air-conditioner is descríbed for a consumer who must replace his oldappliance whìch bas gone out of order.

When an air-conditioner is purchased or replaced, a consumer bas theoption of choosing from a number of different types of air-cond[t[onlng. Inthis consumer behaviour model, two types of air-conditioning which bothdeliver the same amount of service are considered. Generally, a consumerwho is well informed about the purchase price, energy efficiency, and ser-vice level of different types of applíances wfll choose the appliance typewhich best balances the added purchase costs of higher efficiency againstthe reduction in operating costs. Therefore, the two types of air-conditio-ners that are considered are a less energy efficient alr-conditioner with alower purehase price (air-conditioner I) and a more efficient aír-conditionerwith a higher purchase price (air-conditioner 2). This tradeofl: between thelower operating costs for the more efficient appliance and the lower capitalcosts for the less efficient appliance is an important element in the model.

In the model, the fo]}owing notat[on will be used:

SU = service units.MU = money units.EU = energy units.

xY

flow ot: service provided by the appliance (air-condltioner) in [SU].flow of service provided by all other goods in ISLI].

price per unit of service of the appliance in [MU/SU].price per unit of service of all other goods in [MLI/SU].

MQ

total income of the consumer inpurchase price of the appliance in IMU].

price of energy in [ML1/EU].efficiency, measured in units of deiivered service per unit of energyconsumed in [SU/EU].

A consumer desires the services provided by the air-conditioner X and ellother goods Y, which can be summarized in the utility function U (X, Y).These services cannot be consumed infinitely, since the consumer has alimited amount of money at his disposal. The budget constraint that theconsumer feces is that the purchase price of the appliance plus the price ofthe service delivêred by the appliance and all other goods may hot exceed

Px.X+Pr.Y+ Q <-M (6.1)

A rational consumer will choose that combination of X and Y which yieldsthe highest utility under the given budget constraint. The amount of serviceof X and Y that is demanded in the optimal situation depends on the typeof utility function that is used. In this model, a CES (Constant Elasticity ofSubstitution) utility function is utilized. The CES utility function is given by:

74 ECN-I--94-053


~ (6.2)U (X, Y) -- [ e~.X-P ÷ (]-~)’Y-P ]--~

To calcuIate the demand f’or the service of X and Y, consider the Lagran-gian with multiplier X:

L = (c~.-X-P + (1 -et).y-v) p +k.(M_Q_px.x_pr.y)(6.3)

With this Lagrengian, the first=order conditions can be derived:

Lx = -~ ¯ (g’X-P + (1 -«).Y-P) P " -P ’ ~ " X-P-~ - ~’Px = 0P

or:

-- (6.4)1 . (et.X-P + (1-et)’Y-~) ~P

and:

Lr__ ___1 .(~.X-p ÷ (1-~)’Y-P) p ’-P "(]-~) " Y-P-~ -% "Pr =0P

or:

1 . (c~.X_p + (l_et).y_p) p -p (I-00 Y-P-~ k P~,P

(6.5)

Now divide the first-order condition (6.4) by the first-order condition (6.5).This results in:

e~ ¯ X-P - 1 Px (6.6)( 1 -~) ¯ Y-P-~ Py

This equation can a[so be written es:

X-p - ~ Px . ( I - (~ ) . y-p- ~ (6.7)

or:

(6.8)X=(

Now substitute the X from equation (6.8) into the budget constraint (6.1) toobtain the demand for Y:

ECH-I--94-053 75


M_Q= y.( px,( 1 -et , Px )-’~J-~+~ +Pv) (6.10)

M-Q

In a simi]ar manner, the demand for X can be derived. Recall equation(6.6). This equation can be rewritten as:

y-,o - ~ = PY . ~ . X-~ - 1 (6.12)P~ 1-o~

y _- ( Pv . ~ )-’~~ ~ ¯ X (6.13)Px 1-~

Now substitute the Y from equat~on (6.13) into the budget ¢onstraü~t toobtain the demand for X:

M-- Q+Px’X+P~," « )--Fr’~ . X (6.~4)Px 1-~

M-Q=X" Px + PY ( P,,. ~ )-’r-~) (6.]5)Px ~-~

M-Q

P~ 1-~(6.]6)

To determine the rebound-effect in this consumer behaviour model, recallthe two air-conditioners. Air-conditioner ] represents the less efficient ap-pliance with the lower purchase pdce. This piece of equipment is a newversion of the old appliance which bas gone out of order. Air-conditioner 2represents the more effìcíent appliance wíth the higher purchase price. Thefollowing values are chosen for the parameters in this model:o~ 0.1 in [-1.p 0.5 in [-].Pe 15 in [MU/EU].Q~ = :1000 in [MU].Q2 = 1200 in [MU].M :1 ooooe~ 0.5 in [SU/EU].e2 0.75 in [SU/EU].

76 ECNol--94-053


The price Px per unit of service for the air-conditioners is determined by theenergy costs:

Pe (6.17)Px=__

For air-conditioner 1, the efficiency amounts to 0.5 units of service per unitof energy. The efficiency of air-conditioner 2 is 50% higher, namely 0.75[SU/ECI]. With these efficiencies and a price per unit of energy of 15 MoneyUnits, the price per unit of service for the air-¢onditioner$ can be cal-culated:

15Px, -- -- -- -- -- 30 [MU/SU/ (6.18)e1 0,5

px2 = P~

]5 = 20 IMU / S~ (6°]9)e2 0.75

The efficiency of all other goods Y is 3 /SU/EU~. This means that one unitof energy is required for three service units of all otber goods. The price perunit of Y is partia]]y determined by energy costs and parfia]]y by non-energy costs. The price Py ~s:

py = Pe + 15 + 5 = I0 IMU / SU] (6.20)

With these numbers, the demand for X and Y can be calculated for situa-tion 1 in which the less energy efficient air-¢onditioner ~ is purchased antifor situation 2 in which the more effi¢ient air-conditioner 2 is purehased:

X1 =M- Q~ 10000 - 1000

X~ = 75 (6.21)

M- Q1 10000 - 1000

p~ + p;<, ,( ~ -=. P«, )-ï4-~,~ ~o +3o .(Pv 0.1 10

Y~ = 675 (6.22)

ECNq--94-053 77


c~ . Pr )--~-~

X2 -~ 99.23

10000 - ]200

20 ÷ 10,(-6-$ .~.ô

(6.23)

M- g)~ - 10000 - 1200Y2 -- 1

0.9 20Pr+Px~ ( 1 -(~ " Px~ )-’F’;~ 10 +20’( 0.1

681.54 (6.24)

Witb these demands, the energy use can be calculated for the two situa-tions:

energy use for X in situation 1-- --X1 = __75 = 150 EU (6.25)% 0.5

energy use for Y in situation I --__Y1 = ~675 = 225 EU%, 3

(6.26)

total energy use in situat~on 1 -- 150 + 225 -- 375 E~/ (6.27)

X2 99.23 (6.28)energy use for X in situation 2 = -- -- ~ = 132.31 EUez 0.75

energy use for Y in situation 2 = Y2 = 681.54 = 227.18 EU(6.29)

total energy use in situation 2 = 132.31 + 227.18 = 359.49 EU(6.30)

Now recall thst the ~ir-¢onditioner which has gone out of order was of theless energy efficient type I. This air-conditioner must be replaced wíthe[ther a new type I (less e[ficient, lower purchase price) or a type 2 (moreefficient, higher purcbase price} appliance. To determine the preference forsir-conditioner 1 or sir-conditioner 2, the demands X~, Y~, X» and Y~ mustbe substituted in the CES utility function:

_1 1

U~ = ( a ¯ X(~ + (I - ~.) - Y~-~ ) "~ = ( 0.1 ¯ 75-0.5 + 0.9 - 675-0.5 )-’o5"

468.75 (6.31)

78 ECN-I--94-053

The rebound-effect and the consurner

U2 = (c~ ¯ X2-p + (1 -o,) ¯ Y~-P) "~ : (0.1 ¯ 99.23.0.5 + 0.9 ¯ 681.54-°’s)-’OTg

U2 : 504.69 (6.32)

Since U2 is greater than ~~, the more efficient air-conditioner 2 is preferred.

In situation I, the demand for air-conditioning amounts to 75 units of ser-vice. To de]iver this arnount of service, ]50 (= 75/0..5) Energy Units arerequired. Because the total amount of energy required for a]l other goods Yarnounts to 225 Energy Units, the total energy use in situation I is 375150 + 225) Energy Units. ]n situation 2, the efficiency of the alr-conditionerhas increased by 50% flora 0.5 to 0.75, cornpared with air-conditioner I.The traditional engineering approach now assurnes that as a result of thisenergy efficiency improvement, the energy use for air-conditioning in situa-tion 2 wi]l drop from 150 Energy Units to 100 (: 75/0.7.5) Energy Units.Because the traditional engineering approach also assumes that the energyuse for al] other goods rernains 225 Energy Units, the engineering predie-tion of the consumer’s energy use in situation 2 is 325 (: I00 + 225) Ener-gy Units and the expected energy savings arnount to 50 (: 375 - 325)Energy Units.

In reality, however, the energy use in situation 2 amounts to 359.49 EnergyUnits. The reason for the fact that the engineering estirnate of the energysavings is higher than the actual energy savings is that the traditional en-gineering approach does not consider the interaction between the energysystern and the econornic systern in its analysis. More energy efficient ap-pliances reduce the price per units of service they deliver. As a result ofthis, an arnount of money is saved. This rnoney will be respent on an in-creased demand for eir-conditioning and a]l other goods. This results inenergy savings which are smaller than those projected by the traditionalengineering approach. ]n this particular case, the lost energy savingsamount to 34.49 (-- 359.49 - 325) Energy Units.

With these lost energy savings of 34.49 Energy Units and the expectedenergy savings of 50 energy units, the rebound-effect can be calculated:

rebound-effect lost energy savings 34.49.... 0.690expected energy savings 50

(6.33)

The rebound-effect in this particular case thus amounts to 69%. This meansthat almost 70% of the engineering estimate of the energy savings is lostbecause of the feedback from the engineering measures to the economy.No real meaning must however be attached to the rnagnitude of therebound-effect in this consurner behaviour rnodel, since the values adoptedfor the pararneters in the rnodel are not estimated by empirical research.The consumer behaviour rnodel has merely heen used to clarify the under-lying thought behind the calculation of this feedback-effect. A more realisticdetermination of the rebound-effect can be obtained with the aid of theMARKAL-MACRO model

ECN-I--94-053 79

7. QUANTIFICATION OF THEREBOUND-EFFECT WITHMARKAL-MACRO

7.1 Introduction

In chapter 6, the rebound-efllect and the consumer has heen described. Inthis chapter, an attempt is made to determine the magnitude of thisrebound-effect on the consumption-side of the economy. An estimat~on ofthe effect of energy efficiency improvements on the demand for energy canbe obtained with the aid of the MARKAL-MACRO model. Paragraph 7.2contains a description of this MARKAL-MACRO computer model. Para-graph 7.3 presents the calculation of the rebound-effect.

7.2 Description of the MARKAL-MACRO model

Since the rebound-effect is a result of the energy-economy-environmentinteraction, it can only be determined with a model that integrates thesethree aspects. The MARKAL-MACRO model is such an integrated model. InMARKAL-MACRO, two models are formally linked together. These modelsare the systems engineering model MARKAL and the long-term macro-economic growth model MACRO.

7.2.1 The MARKAL model

The MARKAL (acronym for MARKer ALlocation) model was developedbetween 1976 and 1981 by the members of the Energy Technology Sys-tems Analysis Programme (ETSAP) under the auspices of the InternationalEnergy Agency (IEA) (Fishbone, 1983; Rowe and Hill, 1989). ETSAP wasinitiated in order to provide the IEA with an analytical tool for evaluatingenergy technologies within the context of national energy systems. TheMARKAL model has been used for assessments of the national energysystems for toost 1EA countries. It bas also been used as an aid for energyplanners in developing nations such as Brasil, China, and Indonesia. At alower level of aggregation, MARKAL has heen used to analyze regionalenergy systems in Canada and community energy systems in Sweden.

MARKAL is a technologically oriented linear programming model of anenergy system. In any energy system, primary energy sources are firstextracted and then ~ransformed by various supply technologies into a num-ber of energy carriers, such as oil and gas. These energy carriers in tutuare ultimately used by a vast number of end-use technologies for the pur-pose of satisfylng human demands for particular energy services. In orderto accuralely represent such an energy system, the MARKAL model is builton the concept of a RES (Reference Energy System). The RES is a flow-chart indicating all possible routes from each source of primary energythrough various conversion processes to all service demands for energy.

81

L


The flowchart can also incorporate the emissíons resulting from activities inwhich energy is transported or converted. With the aid of the ReferenceEnergy System, the MARKAL model can identìfy those routes anti tech-no}ogies that best satisfy the stated policy goa}s.

Since MARKAL is a data-driven model, tbc resu]ts of the MARKAL runsdepend heavily on the qua]ity of the input data. To bui]d the model, thefollowing fout brood categories of data are required:

Technologies and their characteristics. In MARKAL, a technology is de-scribed by technical and economie data, such as the efficiency of a pro-cess, the useful lifetime of the technology, the investment costs per unitof capacity, the operating and maintenance costs of the technology, andthe datum of avaflability o~: a new technology. The scale of the teeh-nologies incorporated in the model may be either large or small. Anintegrated coal-gasification combíned-cyc[e electricity generation stationis an example of a large technology. Energy-efficient lighting and electriccars are examples of small-scale technologies. Technologies and theircharacteristics represent toost of the ìnput data in a MARKAL data-base.Primary energy sources. Primary energy may be expressed as oil andgas wells, coa] and uranium mines, anti biomass raw materlal. A sourceof primary energy is normal]y represented by a graph indJcatlng the po-tenfial for extraction and the assoclated extraction costs. The MARKALmodel a]so incorporates import and export of primary energy sources.Useful energy demands. The demands for useful energy do hot refer tospecific fuels. They rather refer to the service that a fuel render~. Exam-ples ot: useful energy demands are automobile passenger-kflometres,units of thermal comfort of dwellings, and tons of steel. The service de-mands for energy are divided into several sectors, such as ìndustry,transportation, and residential. These sectors may in tum be further dis-aggregated into subse¢tors.Environmenta[ constraints. Since the negative effects of human activ[tyon the environment are more and more acknowledged, emission reduc-tion strategies are becoming increasingly important. Therefore, environ-mental ¢onstraints that limit the emissions of carbon dioxide, sulphurdioxide, and nitrogen oxides may be incorporated ~n t4ARKAL.

In any computer model, the va]ues for the parameters are either supp]iedas input data (exogenous) or ca]cu[ated by the model (endogenous). InMARKAL, exogenous quantities are for instance prices of specific fuels,useful energy demands per sector and subsector, and the unit costs oftechnologies. Examples of endogenous parameters are the quantity of eachspecific fuel that is used in each time period, anti the utilized capacity ofeach technology in each time period. The endogenous parameters arecalled variables. In the MARKAL model, these variables are tied together bya number of logical relationships, which are called the constraints of themodel. The most important constraints refer to the satisfaction of usefuIenergy demands, the limits of the capacity of technologies, and the maxi-mum allo~vable emissions.

MARKAL is solved by means of dynamic linear programming. This impliesthat in MARKAL the values of a]l model variables are determined simul-taneously in such a way that the constraints are satisfied and the objective

82 ECN-I--94-053

Quantification of the rebound-effect wRh MARKAL-MACRO

function in mlnimlzed. The objective function in MARKAL is the total netpresent value of the energy system’s costs throughout the planning horizon.

7.2.2 The MACRO model

The MACRO model is a neoclassical maeroeconomic growth model, whichtakes an aggregate view of long-term economie developments (Manne andRiche]s, Jg92). MACRO in¢orporates one economy-wide production func-tion. The factors of production for the model are capital, labour, anti in-dividual demands for useful energy. These input factors enter into the pro-duction function, which is a nested CES function. The nest/ng scheme ofthe production function is [(K,L), EI. This particular nesting bas capital andlabour combining Cobb-Douglas fashion and the two together ¢ombiningCE$ fashíon with energy. Beeause the production function bas a nestednonlinear forto, the MACRO model is charaeterized by smooth substitution.With its nonline~rity, a small prlce change of one of the input la�tors leadsto a small change in the mix of inputs. This outcome could hOt be obtainedby linear progr~mming models, such as MARKAL. With these models, aso-called penny-swRching effect is namely often en¢ountered. This meansthata small change in a variable will either lead to almost no effect or elsetoa very large effect.

The economy’s output in MACRO is used for investment, consumption, aodinter-industry payments for the cost of energy, lnvestment is used for ca-pita! accumulation. In each period, the net [ncrease in the stock of cap[talis determined by investment minus depreciation. The long-term growth rareof the economy in MACRO is determined primarily by the value adopted forthe growth of the Iabour force and its productivity. The combination ofthese two factors is expressed in labour efficiency units.

MACRO is solved by means of dynamic nonlinear optimization. In eachtime period, the MACRO model selects among the possible combinations ofinvestment, consumption, and energy cost payments in order to maximizethe objective function of maximum discounted utilìty of consumption.MACRO is dynamic in the same sense as MARKAL, which means thatMACRO uses lookahead features for optlmizing the objective function overthe entire planning horizon. This is called perfect foresight.

7,2.3 The limitations of the stand-alone versions of MARKALand MACRO

There are two different approaches towards the modeling of theenergy-economy-environment interaction. These competing mode]ing tech-niques, which are called the top-down approach ond the bottom-upapproach, reach very different conclusìons and recommendations on sub-jects which concern the 3-E interaction (Krause, 1993; Wflson ond Swi~her,1993). This cao be illustrated by means of the answer to the fundamenta!question in the greenhouse debate of how much it would cost to reducegreenhouse gas emissions. Both in the top-down and bottom-up approach,the costs of reducing carbon dioxide emissions are determined by �om-

ECN-I--94-053 ~3


paring scenarios in which there is an upper-bound set on greenhouse gasemissions with a business-as-usual scenario. The difference between thetwo competing views is that in the top-down approach carbon dioxide re-duction measures will increase the costs with which the energy supply sys-tem delivers the needed energy services. In the bottom-up approach, onthe other hand, carbon dioxide reduction measures result in energy coststhat are lower than in the bus[ness-as-usual situafion. As a result of this,very different po[icy measures are being put forward by the two ap-proaches.

In the top-down approach, the relations between the energy system, theeconomy, and the environment are mainly studied from a macroeconomicperspective. Examples of top-down analyses are the studies by Manne andRichels (1990) anti Nordhaus (1990). Also the MACRO model has amacroeconomic orientation. The name top-down refers to the highaggregation leve] of this approach. The most important input assumptionsfor top-down models are pro~ections of GNP growth rates; price forecastsfor capital, labour, and energy; the e]asticity of price-induced substitutionbetween the three input factors; and the rate of non-price-induced changesin the energy intensity per unit of economic output.

The strength of the top-down approach is that the interactions between theenergy system and the economic system are fairly accurately represented.Doubts however can be expressed regarding the accuracy with which theenergy sector itself is described. Because of the macroeconomic orientationof top-down assessments, the energy sector is namely modeled by only afew equations. This implies that most of the energy technologies are notincorporated. Another important shortcoming of the top-down approach isits empirica] foundation. In top-down mode]s, for example, the future de-mand for energy is forecasted on the basis of historie re]ationships betweenenergy use, energy prices, and econornic output. This is abackwardqooking approach. The problem with such an approach is thefact that historic relationships do hot explain how things work. For thisreason, the extrapolation of historica] data into the future is on]y re]iab]e ifthe system írorn which the data are obtained does hot change too muchover time. In the changing wor]d of today, using historical data to predictfuture deve]opments in the energy system is hot very reliab]e. By assumingthat everything wi]l be the same in the future as it was in the past, newenergy technologies that can provide the same service with less energy arehot included. Aiso saturation effects in the demand for energy services arehot raken into account in top-down assessments.

Because of the highly simplified description of the energy sector and thebackward-looking approach, top-down modMs tend to systernatically over-estimate future demands [or energy. Another implication of the short-comings o1: the top-down approach concems the estimat[on of the cost ofcarbon dioxide reduction strategies. The models that are utilized in top-down assessments determine the impact of carbon díoxíde reduction strate-gies by calculating the effects of carbon taxes on overall economic perfor-mance. This means that carbon dioxide reduction measures are seen asopportunity costs. As a result of the overestimation of future energy de-mands and the limited technological options for reducing carbon dioxide

84 ECN-l--94-053

Quantification of the rebound-effect with MARKAL-MACRO

emissions, maeroeconomic modeIs tend to overestimate the costs or car-bon emission reductions.

In the bottom-up approach, the relations between the energy system, theeconomy, and the environment are analyzed from an engineering perspec-t[ve. Examples of bottom-up approaehes are the studies by Lov[ns andLovins (]991) anti Williams (1990). Also the MARKAL model has an en-gineering orientation. The name bottom-up refers to the fact that this ap-proach begins its analysis with the consumer demands for energy services,such as hearing, cooling, anti lighting. The focus is thus on technologiesthat provide energy services, rather than on energy carriers, such as oi! andgas. In bottom-up models, energy technoIogies that can provide the desiredenergy services are included. Also alternative state-of-the-art teehnologiesthat de]iver the same amount of service with lower energy requirements areincorporated. The thought behind this is that the exìsting energy tech-nologies have fiNte lifetimes, and that energy efficient state-of-the-art tech-nologies can rep~ace these existing technologies when they are written of.The most important input assumptions for bottom-up models are the costsand energy consumption of existing technologies and their state-of-the-artaltematives; the penetration rate of alternative technologies; the priees ofspecific fuels; and the consumer demands for various energy services.

The main strength of bottom-up mode]s is that a very accurate and exten-sire representation of existing and alternafive technologies in the energysystem is used. A very important shortcoming of the bottom-up approachis that the economie feedback effects resu]ting from changes in the energysystem are hOt inciuded. For example, the effect of the introducfion ofmore energy eíficient techno]ogies on the demand for energy services ishot raken into account. The models used in the bottom-up approach thusonly deal with demands for energy services and technologies that can deli-ver these services. They are hot capaNe of quantifying the effect of chan-ges in the energy sector on the other sectors of the economy or on econ-omie output.

With respect to carbon dioxide reduction strategies, the bottom-up ap-proach only calculates the costs of carbon dioxide emission reductions asenergy expenditures, and hot as opportunity costs, In fact, many bottom-upmodellers argue that these energy expenditures are negative because alarge part ot: the more energy efficient technologies are profitable. This inturn implies that it is the business-as-usual scenarios that entail the oppor-tunity costs. When carbon dioxíde reductíon strategies save money, notcarrying out these strategies will result in a lower than possible overalleconomie performance. The bottom-up approach thus reaches very favour-able conclusions on the ability of energy technologies to mitigate the globalwarming probIem. The problem with bottom-up assessments however isthat by neglecting economic feedback-effects, the bottom-up approaehunderestimates the costs of carbon dioxide reduction strategies.

85


7.2.4 The MARKAL-MACRO model

It is thus difficult to compare the outcomes of top-down and bottom-upanalyses. The reason for this is that the language of the two approaches israther different. This problem is partially caused by the fact that the twomodeling techniques originate from different academie disciplines. Anotherpart of the problem is caused by the faet that the two modeling approachesare not asking the same question. The models were namely developed fordifferent purposes. Because the two approaches were designed for differentreasons, each of them possesses certain distinguishable qualities. However,both top-down and bottom-up models also sulfer [rom serious short-comings. To overcome these limitations of the stand-alone versions oftop-down models such as the MACRO model and bottom-up models suchas the MARKAL model, the MARKAL-MACRO model has been developed(Hamilton, 1992; Kypreos, 1992). With MARKAL-MACRO, an attempt ismade to reconcile the macroeconomlc and the systems engineering ap-proach. The integrated model unites MACRO’s aggregate economic viewtogether with MARKAL’s detailed description of the different technologies inthe energy system.

With MARKAL-MACRO, the interaction between economic activity, theenergy demands associated with this activity, and the energy system thatmust supply these demands can be investigated. As a result of this,MARKAL-MACRO can be seen as a tool to enhance the understandingbetween the engineer, the economist, and the policy maker. In figure 7.1,an overview of the MARKAL-MACRO modeI is given.

MARKAL- MACRO¯ Energy Sources |. Technology¯ Environmental

Constraints

~:’""«=~1MARKAL 1~4

] ConsumptionEnergyUSeful

Demand~iL MACRO

EnergyCost

L Investment

Figure 7.1 The MARKAL-MACRO model

In order to keep the need for structural changes in the two origina] modelsto a minimum, only two types of linkage are introduced. These are therepresentation of flows of energy from MARKAL into MACRO, and pay-ments for these flows from MACRO into MARKAL.

The stand-alone version of MARKAL is demand-driven¯ This means that inMARKAL the demands for useful energy are specified exogenously. [n thelinked model, energy supplies, energy costs, and useful energy demandsare interdependent. They are determined endogenously within MARKAL-MACRO.

86 ECN-I--94-053


Aggregate energy eosts are generated in MARKAL. Besides fuel costs,these energy costs are composed of capital costs and operating and main-tenance costs for all supply and conversion technologies. Together withconsumption and investment, the energy costs represent claims upon thegross output generated by the MACRO production function.

MARKAL-MACRO is written in GAMS (a Genera]ized Algebraic ModelingSystem). GAMS is a high level modeling language used te build computermodels and support the development of these models. Besides this, GAMSalso increases the reliability .and flexibility assocìated with the use of themodels. The mode]ing language has been deve]oped by the Analytic Sup-port Unit of the World Bank (Brooks, 1988). GAMS has been especiallydesigned te handle both linear and nonlinear complicated computermode]s. The syntax of GAMS closely resembles the row-oriented style ofthe constraint equations formulated in MARKAL-MACRO. GAMS provides a¢onvenient interface te nonlinear optimizers. The MARKAL-MACRO modelis solved using the optimizer MINOS (Model Incore Nonlinear OptimizationSystem). MARKAL-MACRO operates in a processing environment which is~entred around an integrated analysis support tool. This tool is calledMUSS (Markal User Support 8ystem). MUSS is a user-friendly interfaceincorporafing the features of a relational database, spreadsheet, filemanagement, and graphics presentation system.

7.2.5 The basic relations in MARKAL-MACRO

MARKAL-MACRO is a model which is created by coupling the MARKALmodel and the MACRO model. This integration of MARKAL and MACRO isbased upon the concept of ene economy-wide production function. Themain advantage of this approach is that a direct link can be establishedbetween a physìcal process analysis and a macroeconomic growth modelThe main disadvantage is that the representafion of the whole economy byonly ene production function is a gross simplification of reality. InMARKAL-MACRO, aggregate output during period t is determined by enenested CES (Constant Elastícity of Substitution) productíon functíon. Thisproduction function is of the following specific form:

Y~ = [ a . Ktr’’’~ . Ltp’(1-~) + ~bi. Dü ]-~(7.1)

In this function, the symbol Kt denotes the capital stock in period t, thesymbo] Lt denotes the labeur force in period t, and the symbol Di,t denotesthe demand for useful energy type i in period t. Aggregate output in periodt is represented by the symbol Yt. This particular CES production functioncontains a capita]-]abour aggregate and an aggregate of usefu] energy de-mands. At the top level, the capital-labour aggregate may be substitutedfor the energy aggregate. At the bottom level, there is a unity e]asticity ofsubsfitution between capital and labeur. This structure implies that capitalanti labeur may be directly substituted for each ether, tìor example, threughthe automation of [abour-intensive tasks. The higher the wages become,the more attractive ít becomes for employers te switch over te automated

ECN-1--94-053 87


pmduction. At the bottom level, also each category of useful energy de-mand may be substituted for the other.

The ease or difficulty of substitution between the capital-labour aggregateand the energy aggregate in MARKAL-MACRO is determined by the valueadopted for the ESLIB (Elasticity of SUBstitution) parameter. This para-meter is related to the exponent p of the produetion function through thefollowing equation:

~ (7.2)ESUB = o = --1-p

Aggregate output in MARKAL-MACRO is thus a funetion of capital, labour,and different types of useful energy demand. This aggregate output is usedfor three purposes, namely for consumption, for building up the stock ofcapital, and for inter-industry payments for energy ¢osts. This relafion canbe specified through the followlng equation:

(7.3)

In this equation, Ct denotes consumption in period t, [Vt denotes in-vestment in períod t, and the energy cost payments ín period t are denotedby ECt.

In NARKAL-MACRO, the energy sector is defined as an intermediate sec-tor. Therefore, the Gross Domestic Product equals consumption plus in-vestment. In formula:

GDP~ "= Ct + IVt (7.4)

MARKAL-MACRO is a non-linear computer model. It uses the eriterion o~:the maximum ~um of the discounted logarithm of aggregate consumptionto obtain the optimal solution.

7.3 Calculation of the rebound-effect

Energy efficiency improvements do not necessariiy lead to proportionatedecreases in the demand for energy. The rebound-effect is a means ofquantffying this effect. The rebound-effect can be defined as that part of theinitially expected energy savings, resulting from energy eflìciency improve-ments, that ~s lost because of the energy-economy-environment interaction:

rebound-effect = lost energy savingsexpected energy sauings

(7.5)

Because feedbacks from the energy system to the economy are incor-porated in the MARKAL-MACRO model, the effect of energy efficiencyimprovements on the demand for energy can be estimated with this Ilnked

88 ECN-1o-94-053

Quantification of the rebound-effect with/V~RKAL-MAcRO

energy-economy-environment model. Bui first consider the example offigure 7.2.

SITUATION A: NO EFFICIENCY IMPROVEMENT + NO DEMAND CHANGE

0.5200 " 1 O0

cenversion effF wJthout D without

SiTUATION B: EFFICIENCY IMPROVEMENT (25%) + NO DEMAND CHANGE

0.5 x 1.25 = 0.625160 ~ 100

conversJon eff

SITUATION C: EFFICIENCY IMPROVEMENT (25%) + OEMAND CHANGE

0.5 x 1.25 -- 0.625176 ~" 110

conversion effF with D with

F = Final energy use D ,= useful energy demand

Figure 7.2 Example

In this figure, the conversion from final energy use F te usefu~ energy de-mand D is shown for three different cases, Situat[on A of figure 7.2 de-scr]bes ihe base case. In ibis base case, there is a uaefu] energy demand ofI00 energy units. The conversion efficiency in situation A is 0.5. Since thefinal energy use is the quot[ent o~: the useful energy demand and the con-version efficiency, the final energy use in situation A becomes 200(= 100/0.5) energy uníts.

blow the conversion efficiency is increased by 2.5% flora 0.5 te 0,625(= 1.25 * 0.5). Situation B gives the traditional engineering response te thisefficiency improvement. The engineering approach only takes into accountthe effect of the efficiency improvement en the tìnal energy use. Since itdoes net take into account the effect et: energy efficiency improvements enthe demand for energy, the useful energy demand in situation B rema[ns100 energy units. With this usei:ul demand of 100 energy units and a con-version efficiency of 0.625, the flna~ energy use in situation B becomes 160(= I00/0.625) energy units.

In sJtuation C, also the impact of the price-induced energy efficiency im-provement en the useful energy demand is considered. ~rice-induced ef-ficiency improvements oan be defined as efficiency improvements thatreduce the prìce per unit of useful energy demand. For example, the pur-chase of a more fuel-efficient car will Iead te a reduction in the pdee perautomobile passenger-kì{ometre. As a result of this price reduction, theuseful energy demand will very lìkely increase. In situation C, the usefuldemand increases flora 100 te ] 10 energy units. Because the energy ef-ficiency is 0.625, the fìnal energy use in situation C becomes 176(= ] 10/0.625) energy units.

89


New the rebound-effect can be quantified. The expected energy savingscan be calculated from the difference in the final energy use in the basecase (situation A) and the situation in which no behavioural response ispermitted (situation B), Theref’ore, the expected energy savings in thisexample amount te 40 (= 200 - 160) energy units. Since the real energysavings ~:an be calculated from the dlfi:erence in the fìnal energy use in thebase case (situation A) and the situation in which a behavioural response ispermitted (situation C), the real energy savings are 24 (= í~00 - 176) ener-gy units. The lost energy savings can be calculated from the differencebetween the expected and the real energy savings. The lost energy savingsin this situation become 16 (= 40 - 24) energy units. The rebound-effect isthe quofient of the lost energy savings and the expected energy savings.The rebound-effect in thís example thus becomes:

lost energy savings =__16 = 0.40 (7.6)rebound-effect =expected energy sauings 40

In order te quantify the rebound-effect with MARKAL-MACRO, two runs arete be carried out. These runs are identical except for ene thing. In ene ofthe two runs, there is a very low upper-bound set en price-induced energyefficiency improvements en the consumption-side of the economy, such asnew cooling techniques, energy-efficient lighting, and improved thermalefiïciency oí dwellings. This means that no price-induced energy-efficiencyimprovements are possib]e in this particular run. This MARKAL-MACROrun without energy efficiency [mprovements corresponds with situation A ofthe example. In the ether run, no upper bounds are set en efficiency irn-provements. This MARKAL-MACRO run with energy efficiency improve-ments corresponds with situatìon C of the example. The results of the tworuns with the MARKAL-MACRO model for the Nether[ands in the year 2020are shown in figure 7.3.

final energy useful energyuse demand

no efficiency 2353.53 2605.74improwmonf

F without D without

with efficiency 2340.00 2613,10improvemontF with D with

Figure 7.3 MARKAL-MACRO results en the consumption-side

In this figure, final energy use F is expressed in Petajoules. Useful energydemand D is a mix of various energy service demands expressed as anindex number. The two runs of figure 7.3 are performed under a 50% CO2reduction constraint and with a value for the overall elasticity of substitutionESUB of 0.25. The CO2 reduction constraint is used because without thisconstraint, very few energy efficiency improvements wou~d be profitab]e.This means that practica]ly no energy efficient technologies would be pre-

90 ECN-1--94-053


sent in the MARKAL-MACRO run with the price-induced energy efficiencyimprovements. Consequently, the comparison of the runs with and withoutprice-induced energy efficiency improvements would lead toa comparisonof two nearly identical MARKAL-MACRO runs. With the data of figure 7.3,the rebound-effect can be calcu]ated in a simi]ar way as in the example.The results of this ca[cu]ation are shown in figure 7.4.

SITUATION A: NO EFFICIENCY IMPROVEMENT + NO DEMAND CHANGE1.107

2605.742353.53 conversion eff ~ D withoutF without

SITUATION B: EFFICIENCY IMPROVEMENT + NO DEMAND CHANGE1.117

~ 2605.742333.41 conversion eff

SITUATION C: EFFICIENCYIMPROVEMENT+ DEMANDCHANGE1.117

2340.00 conversion eff ~ 2613.10F with D with

F = final energy use D = useful energy demand

expected energysavings -- 2353.53-2333.41 20.12real energysavings = 2353.53-2340.00 = 13.53Iost energy savings = 20.12- 13.53 = 6,59

Iost energy savings 6,59Rebound = expected energy savings = 20.12 = 32.8 %

Figure 7.4 Calc~lation of the rebound-effect

In this figure, the final energy use and useful energy demand of situation Aare obtained from the MARKAL-MACRO run without energy efficiency im-provements. The conversion efficiency of 1.107 in situation A is the quo-tlent of useful energy demand (2605.74) and final energy use (2353.53) ofthat particular run.

The final energy use and usefuI energy demand of situation C are obtainedfrom the MARKAL-MACRO run with energy efficiency improvements. Theconversion efficiency from final energy use to useful energy demand insituation C is calculated in a similar way as in situation A.

For situation B, no MARKAL-MACRO run is carried out. The reason for thisis the fact that the final energy use and the useful energy demand for thissituation can be obtained from the two MARKAL-MACRO runs that areused for the situations A and C. Since in situation B no behavioural respon-se is allowed, the useful energy demand of situation B equals the usefulenergy demand of situation A, namely 2605.74. The conversion efficiencyof situation C is 1.117 (= 2613.10/2340.00). Because in both situation Banti situation C price-induced efficiency improvements are included, theconversion efficiency of situation C can be used for situation B as wel].Now the final energy use of situation B can be ca]cu]ated by dividing theusefu] energy demand of 2605.74 by the conversion efficiency of 1.117.This results in a final energy use in situation B of 2333.41 Petajoules.

The fina] energy use and useful energy demand of situation B can also bederived fo][owing another path. Situation B represents the situation Mth

ECN-I--94-053 91


price-induced efficiency improvements but without the behavMuraI respon-se to these changes. Since the useful energy demands are specified exoge-nously in MARKAL, this stand-alone engineering version of the MARKAL-MACRO model exactly fits this description. To determine the final energyuse and useful energy demand for situation B with the MARKAL modelwould however requke a great effort. That is, in order to use the MARKALmodel for situation B, the values for the use[ul energy demands that aredetermined endogenously in the MARKAL-MACRO run of sitüation A mustbe led into the MARKAL model [or 39 types of useful energy demand andfor 5 time periods.

With the data of figure 7.4, the expected energy savings, the real energysavings, and the lost energy savings can be ca]cu]ated. These energysavings are of course hot actual realizations but figures calculated by theMARKAL-MACRO model. The expected energy savings refer to the en-gineering estimate of the energy savings, in which no feedback effects aretaken into account. The expected energy savings amount to 20.12 (=2353.53 - 2333.41) Petajoules. The real energy savings refer to the situa-tion in which there are feedback-effects from the energy efficiency im-provements to the demand for energy. The real energy savings are 33.53(= 2353.53 - 2340.00) Petajoules. From the difference between these twofigures, the lost energy savings can be calculated. The lost energy savingsamount to 6.59 (-- 20.12 - 13.53) Petajoules. With the lost energy savingsand the expected energy savings, the rebound-effect calculated byMARKAL-MACRO can be determined:

rebound-effect lost energy savings ó.59.... 0.328expected energy savings 20.12

(7.7)

The MARKAL-MACRO results thus indicate that there is a rebound-effect of32.8% in this particular case. This means that almost 33% of the initiallyexpected energy savings is lost because of the energy-economy linkage.

The rebound-effect results from price-induced energy efficiency improve-ments pulling up the useful energy demand. In MARKAL-MACRO, there aretwo reasons for the fact that the useful energy demand in the run with ener-gy efficiency improvements exceeds the level of useful energy demand inthe run without efficiency improvements. To clarify this, consider the ag-gregate production function that is used in the MARKAL-MACRO model:

From this production function, the dual function which relates the price ofoutput to the pr[ces of capital, labour and useful energy demands can bederived:

(7.9)

92


With these equations, the demand for useful energy type i under proffimaximizati0n can be determined in the following way:

max! profit -- max! W= max! Py’Y- PK’K- PL’L-~PD,’Dz (7.10)

~0

~.~ì- ~1Pv " ( Y ~ " bz " Dï- ~ ) - Pr~, -- 0

y~ _~ " bi . D~_~ =Pv

y1 - ~ , bi , Pv

Pq (7.11)Di= y.(__ )-°b~ ¯ Pç

With this equation, the two causes for the increase in the demand for usefulenergy resulting from energy efficiency improvements can be explained.The first cause is the fact that price-induced energy efficiency improve-ments lead toa decrease in the price per unit of useful energy demand,relative to the price per unit of eapitM and labour. For instance, cavity wallinsulation is a very profitable example of an energy efficiency improve-ment. Consequently, the installation of insulation will lead toa reduction inthe price per unit of thermal comfort. When the relative price of one of theproduction factors changes, the production function allows us to adapt bysubstituting more of the cheaper production factor in place of the moreexpensive production factors. So, when useful energy becomes cheaper,capital and labour will be substituted for useful energy, and the useful ener-gy demand will increase. This can be illustrated with equation (7.11). Whenthe price PD~ per unit of useful energy demand of type i in this equationdecreases, the demand Di for useful energy of that type will increase.

The second cause is the fact that price-induced efficiency improvements inthe MARKAL-MACRO model lead to higher economie growth rates. Totalproduction Y in the year 2020 in the run with the energy efficiency im-provements amounts to 905.890 billion guilders. In the MARKAL-MACROrun without the energy efficiency improvements, total production is slightly

93


lower, namely 904.094 billion guilders. A higher level of production is ac-companied by a higher useful energy demand. This a]so can be i]]ustratedw[th equat[on (7.1 ~). When tota] production Y increases, the useful energydemand increases as welL

There a]so exists a third possib]e cause for the rebound-effect. This feed-back-effect results flora the fact that widespread energy efficiency improve-ments may depress the dementi of energy, thereby reducing the prices ofenergy carders, such as oil and gas. These price reductions in rum mayincrease the demand for the energy carriers. ]n MARKAL-MACRO, thisrebound-effect is hot considered, because for exemple the pdce of a barrelof oil is hot endogenously determined in the model, but specified exoge-nously.

The rebound-effect ca]culated with the MARKAL-MACRO model thusdicates that almost 33% of the expected energy sav[ngs is Iost becauìethe economic feedback from engineering rneasures to increase the energyefficiency. Great care however must be raken when this result is extrapo-]ated fram the MARKAL-MACRO model to the real wor]d. Just like anyother modeI, MARKAL-MACRO is by definítion a símpIification of realíty.Recall, for example, the aggregate production function of MARKAL-MACRO:

(7.12)

In this production function, there are substitution possibilities between thecapital-[abour aggregate and the aggregate of useful energy demands.These substitution possibilities are govemed by an overall eiasticity of sub-stitution ESUB. In MARKAL-MACRO, energy is substituted for capita] and]abour when the pdce per unit of useful energy is decreased. Consider forexample energy-efficient lighting. When the introduction of this type oflighting results in a reduction in the price per unit of illumination, the priceper unit of the aggregate of usefu] energy demands is decreased as we]l. Asa resu/t of this pdce reduction of the energy aggregate, there will be substi-tution flora the capitaI-labour aggregate to the useful energy aggregate.The increased demand for the useful energy aggregate is then decomposedinto increased demands for the different types of useful energy in separate.This is a very crude way of describing the substitution possibilifies betweencapital, labour, and energy in an economy. Besides this, the economy inthe MARKAL-MACRO model is represented by only one sector, in whichthere are no possibilitie$ for international trade. This is a gross simp]ifica-tion of reality as we]l. Therefore, the MARKALoMACRO calculat[ons of therebound-effect must be carefully interpreted when they are used to estimatethe magnitude of the rebound-effect in reality.

94 ECN-l--94-053

8. THE REBOUND-EFFECT IN PERSPECTIVE

8, ~. Introduction

In the previous chapter, the magnitude of the rebound-effect has beenest[mated with the MARKAL-MACRO model. In this chapter, the rebound-effect is placed in perspective. Paragraph 8.2 outline~ the consequences ofthe existenee of a rebound-effect for the so-called fifth fuel concept. Para-graph 8.3 deals with the queation of how the rebound-efíect can beduced. In paragraph 8.4, the fact that the rebound-effect besides to energyal$o appl[es to ~he other production factors is discussed. The existence ofthe reboundeffeet in other fields of research is outlined in this paragraph aswell.

8.2 The fallacy of the fifth fuel concept

Over the past 20 years, the interest in energy efficiency bas vastly grown.The reason for thisis the fact that many think that improvements in energyefficiency are the solution to all aorta problems. These problems range fromreducing the dependence on imported fuels, through avoiding ìnvestment innew power plants, to combatting the threat of global warming. Also natio-hal govemments believe that promoting energy efficiency is an importantelement in meeting their environmental policy goals.

An important advocate of energy efficiency is Klaus Meyer-Abich. In Ener-gieeinsparang als neue Energiequelle, Meyer-Abich argues that energyefficiency improvements can be considered a new source of energy: ’Ene-rgieeinsparung ist neben den fossilen Brennstoffen und den Kernenergie-trägem eine der pro,en Energiequellen der Zukunfl. Der in der Nachfragenach Energie zum Ausdruck kommende Bedarf kann nicht nur durch Bren-nstoffe (fossil oder kernenergetisch), sondern in grol~em Umfang genausogut oder besser durch Energieeinsparung gedeckt werden. Die Möglich-keiten zur Nutzung der Energiequelle Energieeinsparung treten deshalbirnmer mehr in den Vordergrund der energiepolitischen Diskussion (1979,p. 17)’. And Meyer-Abich continues: ’Man kann denselben blutzen, derbisher durch den Einsatz bestimmter Mengen von Energíeträgem erzíeltoder gewährleistet wurde, auch durch einen geringeren Einsatz an Ener-gieträgern hervorbringen, wenn zusätzliche Einsparungsmal~nahmen getrof-fen werden, lnsoweit Energieträger durch Einsparungsma~nahmen substi-tuiert werden können, ist die Einsparung von Energie funktional eine Ener-giequelle. Da~ IEnergieeinsparung im Rahmen von Energieversorgungssys-temen funktional eine Energiequelle ist und als solche verstärkt genutztwerden kann, ist energiewirtschaftlich und energiepolitisch auch in der Öf-fentlichkeit eine der erfreulicheren Perspekfiven der letzten Jahre. Der An-tell der nicht durch Verzichte erkauften und unter dem Gesiehtspunkt desgleichen Nutzens mit dem Energieträgerangebot konkurrierenden Energie-quel]e Energieeinsparung an der Energieversorgung der Bundesrepub]ikkann in den achtziger und neunziger Jahren durchaus die Gröl3enordnungdes Einsatzes von Kernenergie und Steinkohle erreichen und den derBraunkohle übersteigen (1979, p. 28)’.

ECN-1--94-053 95


T. Anderson agrees with Meyer-Abich. In Letting Consemation Compete forEnergy Dollars: a Policy Imperatiue, Anderson (1990) argues that energyefficiency improvements are becoming increasingly attractive as an alter-native te meer the growing world energy demand. According te Anderson,’energy conservation should be analyzed as both an energy resource in itsown right and a means of reducing costly uncertainty about future energyneeds. Conservation should be offered the opportunity te compete en equalterms with more conventional options for the capital that authorìtiesallocate specifically for energy investments (1990, p. 343). Anderson ar-gues that a new approach te energy planning is required. This new ap-proach should exp]icitly treat energy efficiency as a new energy supplysource.

Similar thou9hts are also expressed by Steward Boyle (1989) in More Werkfor Less Energy. According te Boyle, energy efficiency works: ’adding upthe advantages of energy efficiency invariably leads te the conclusion thatspending money en efficiency measures, rather than en new sources ofsupply, is much more cost-effective (1989, p. 40)’. Boyle asserts that thereason that net al! options for efficiency improvements are utilized is thatuti]ities have an inertia with respect te energy efficiency improvements.Utìlities namely never had te consider the pessibility of energy savings asan alternatlve te the increased energy supply from new power stations.

A]so the Association for the Conservation of Energy sees in energy conser-vation a new supply source. The director of this association Andrew Watten(1987) namely argues that besides ei], gas, coal, and uranìum, energyefficiency improvements can be regarded as the new fifth fuel. Warren ex-plains this thought by the following reasoning: ’eustomers do net requireelectri¢ity or gas per se, but rather they require energy services, such asheat, light, and mechanical power. These servíces can be met by the com-bination of supply expansion or demand reduction, whichever meets theirneeds for energy at the lowest possible cost te society. This can come netonly from new power supply sources but equally from energy efficiencyinvestments (1987, p. 522)’, And because Watten argues that investmentsin energy efficiency often provide very good rates of return, a situationmust be created in which these energy efficiency investments are evaluatedby the same criteria as these appiied te investments in traditional energysupply sources. Warren further argues that in policy models that describeenergy supply systems, energy efficiency should be explicitly introduced asa new fuel source.

Seeing energy efficiency improvements as a new $ource of energy obvious-ly bas policy implications, ene of the most important implications is thatthe advocates of the fifth fuel concept argue that the energy needed duringthe time that is required te make the structural change from an energysupp]y system based largely en fossil fuels te an energy supply systembased en renewabIes can be supplied by energy conservation. According teWeissmahr (1993), for example, the end of the fossil fuel era isapproaching, but there is still no new economieally viable energy techonology based en renewables in sight. Confronted with this problem, somescholars have drawn the conclusion that the era of economie growth hasended. Weissmahr however argues that the contrary is the truth: ’the futurewill be prosperous, precisely because so mueh energy bas heen wasted in

ECP{-{--94-053

The rebound-effect in perspective

the past. In the next 30 years about 70 per cent of the fuel used for hearingand cooling houses, and 40 per cent of the oil used in transportation canbe saved and redirected to industry for doing work. This programme willachieve two results at the same time. It will create new rnarkets for allproducts used in well insulated housing and fuel et:ficient transportation,and at the same time hold the price of oil constant because total dernand inindustrial countries will remain stationary or decline (1993, p.43)’. For thisreason, Weissrnahr argues that energy efficiency improvernents will be thedriving force behind the transition period kom fossil fuels to renewables.

S[m[lar thoughts are expressed by Klaus Meyer-Abich, who wrltes the fol-lowing about the tirne provided by energy efficiency improvernents to "rnakethe transition to an energy system based on renewables: ’durch eine Ein-sparungspolitik kann das eigentl[che, langfrlstig gleichwohl anstehendeProblern -nämlich das der Zeit, die für einen ~bergang auf praktisch nichterschöpfbare Energiequellen zur Verfügung steht- erheblich entschärf~ wer-den. Denn eine Pol[t[k, die s[ch die Entwicklung energiesparender Tech-nologien zum Ziel setzt, kann gerade diejenigen technischen Fortsehritteinduzieren, deren Möglichkeit gegen das Erschöpfungsargument und diedaraus gezogenen Konsequenzen geltend gemacht wird, und kann dadurehdie giobalen Vorräte im Vergleich zu einer ungebrochenen Steigerung ihrerAusbeutung strechen, wodurch wiederurn die nötige Zeit für die Entwick-lung und Erprobung der Technologien zur Nutzung von praktisch níchterschöp~baren Energiequellen gewonnen wird (]979, p. 39)’. So, alsoMeyer-Abich argue$ that the tirne available to develop an energy systembased on renewables can be extended by means of energy efficiency irn-provernents.

Because of the slackening of our fossil fuel reserves, the viewing of energyefficiency irnprovernents as a new fifth fuel is very appealing. Tiae onlyproblem of this fifth fuel concept, however, is that it is not true. As a resultof the rebound-effect, energy efficiency improvements eannot be viewed asa new energy supply source. Recall for example Andrew Warren’s (1987)pleading for energy efficiency as the new energy supply source. In ôavingMegabucks by Savtng Megawatts, Warren argues that investments in ener-gy efficiency are o~en very profitable, and that therefore ut[lities must in-vest in energy conservation instead of energy supply capacity. Warren’sassertion that energy efficiency irnprovernents often provide attractive in-vestment opportun[ties is true. For example, [nsulating the cavity wafls is avery lucrative way of irnproving the energy efficiency of a dwelling. And itis also truc that as a result of the insulation measures, a smaller annualamount of natural gas is needed to del[ver the sarne amount of therrnalcomfort. But this does not necessarily irnply that the speed with which ourfossil fuel reserves are depleted is proportionately reduced. The reason forthis is the feedback-effect from the efficiency rneasures to the eeonorny.When energy efficiency irnprovements resu]t in Megabucks being savedbecause less Megawatts are required, these Megabucks will be respent ongoods and services, thereby creating a Mega-rebound-effect. This rebound-effect takes away part o1: the inifially expected energy savings. Conse-quently, referring to energy efficiency as the fifth fuel is rnisleading. Thereis no such thing as a fifth fuel.

ECN~I--94-053 97


8.3 How can the rebound-effect be reduced?

Engineering measures to improve the efficiency of use of energy set inmotion economic forces that may diminish the initially expected energysavings. This effect is called the rebound-effect. The rebound-effect can bedefìned as that part of the initially expected energy savings, resulting fromenergy efficiency improvements, that is Iost because of the 3-E interaction.Thís definition ímplíes that the bigger ís the rebound-effect, the smaller arethe energy savings resulting from the efficiency improvements. Therefore,the rebound-effect should be reduced.

One approach aimed at diminishing the magnitude of the rebound-effect, isby adopting energy efficiency measures that go beyond the economic op-timum. This can be illustrated with the simple consumer behaviour modelof chapter 6. [n this model, the purchase and utilization of an air-conditio~net was described for a consumer who had to replace his old air-conditio-net which had gone out of order. The consumer had to choose between twoappliances. Air-conditioner 1 represented a new app]iance of the same typeas the air-conditioner which had gone out of order. The purchase prìce ofth[s app]iance was 1000 Money Un[ts and its efficiency was 0.5. Air-cond[-tioner 2 represented a more efficient appliance with an efficiency of 0.75and a purchase price of 1200 M0ney Units.

In the consumer behaviour model, appliance 2 was chosen because thecombination of" only a slight]y higher purchase price and a much higherefficiency led toa decrease in the price per unit of cooling, compared withappliance 1. This price decrease per unit of cooling led to an amount ofmoney being saved. In the model, this money was respent on an increaseddemand for air-conditioning and other services. As a result of this respen-ding, 69% of the expected energy savings was lost.

Now consider an air-conditioner of type 3. Thia air-conditioner la a state-of’-the-art appliance with an efficiency of 0.9. Because of the better materialsand components used in their construction, state-of-the-art appliances aremuch more expensive to buy than comparable appliances with averageefficiencies. The purchase price of air-conditioner 3, for example, amountsto 2000 Money Units. With the aid of the formulae derived in chapter 6, thedemand for air-conditioning and other services can be determined for situa-tion 3:

M- Q3 10000 - 2000

Px~ + Pv " ( 1 - ~ P,~ 0.9 16.67

X~ = ]03.24 (8.1)

98 ECN-I--94-053

The rebound-effect in perspeetive

M - 09 _- ]0000 - 2000

PY + Px~ " ( ] - ~ " Px~ )--i--;-~ 10 + 16.67 . ( 0.9 ]6.67(~ p~, 0.1 10

627.93 (8.2)

In these two equations, Px3 represents the price per unit of service for air-conditioner 3. Px3 is the quotient of the price per unit of energy of 15Money Units and the efficiency of the air-conditioner of 0.9.With these demands for air-conditioning and other services, the energy usein the situation when air-conditioner 3 is purchased can be calculated:

energy use for X in situat~on 3 = Xa 103.24 = 114.71 EU (8.3)% 0.9

Y3 627.93energy use for Y in situation 3 = 209.31 EU3

(8.4)

total energy use in situat~on 3 = 114.71 + 209.31 = 324.02 EU (8.5)

In situation 1, the energy use for air-conditioning was 150 Energy Units andthe energy use for other services was 225 Energy Units. The total energyuse in that situation thus amounted to 375 Energy Units. In situation 3, theefficiency of the air-conditioner has increased by 80% from 0.5 to 0.9,compared with the air-conditioner in situation 1. The traditional engineeringapproach now assumes that as a result of this energy efficiency improve-ment, the energy use for air-conditioning will drop from 150 (= 75/0.5)Energy Units to 83.33 (= 75/0.9) Energy Units. Because the engineeringapproach also assumes that the demand for other services remains thesame, the engineering prediction of the energy use in situation 3 becomes308.33 (= 83.33 + 225) Energy Units. The expected energy savings can becalculated from the difference between the energy use in situation 1 andthe engineering prediction of the energy use in situation 3. The expectedenergy savings become 66.67 (= 375 - 308.33) Energy Units.

In reality, the energy use in situation 3 is not 308.33 but 324.02 EnergyUnits. The difference between these two numbers, which is a representationof the lost energy savings, is caused by the feedback-effect resulting florathe purchase of the state-of-the-art energy efficient air-conditioner 3. Thelost energy savings in this case thus amount to 15.69 (= 324.02 - 308.33)Energy Units. Now the rebound-effect in situation 3 can be calculated withthese lost energy savings of 15.69 Energy Units and the expected energysavings of 66.67 Energy Units:

rebound-effect = lost energy savings =__15"69 = 0.235expected energy savings 66.67

(8.6)

ECN-I--94-053 99


So, the rebound-effect in this situation be¢omes 23.5%. This is considerablysma[ler than the rebound-effect in the case when a[r-conditioner 2 waspurchased. The rebound-effect in situation 2 namely emounted to 69%. Thereason for thi~ is the fact that state-of-the-art appliances are so expensíveto buy that the lower energy costs per unit of service cannot make up forthe higher purchase price of these appliances. This rneans that less moneywill be saved than in the economically optimal situation. When less moneyis ~aved, less money wilI be respent, and the rebound-effect will be smaller.

The purchase of appiiances with the highest energy efficlencies thus basvery beneficial consequences. First, the higher is the efficiency of thepfiance, the lower is the energy use per unit of service that the appIiancedelivers. Second, the higher is the efficiency of the appliance, the lower isthe rebound-effect. There is however a probIem with this state-of-the-artapproach towards reducing the magnitude of the rebound-effect. The pur-chase of a very energy efficient appliance namely may lead to a sub-optimal economie decision. To illustrate this, recall the consumer beh~viourmodel in which the two types of air-condìtioners were compared on a utili-tarian basis. Total utility U1 in the situation when appliance 1 was pur-chased amounted to 468.75. Total utility U~ in situation 2 was 504.6g.Because Uz was greater than L[~, the more energy efficient air-conditioner 2was purchased. Now consider the state-of-the-art air-conditioner 3. Withthe demand for air-conditioning X~ and other services Y~, total utility insituation 3 can be determined:

(í!3 -- 477.61 (8.7)

Total utility in situation 3 thu~ amounts to 477.61, which is considerabiysmaller than total utility in the situation when an air-conditioner of type 2 ischosen. This impiies that the purchase of an air-condltioner of type 3 leadstoa ]ower than possible utility under the given budget eonstraint. More ingeneral, it means that a reduction in the magnitude of the rebound-effectby adopt[ng state-of-the-art energy efficiency measures can only be ob-tained at the expense of a more or less disruption of the economy.

8.4 The rebound-effect in a different context

Up to now, the rebound-effect has heen examined for energy. In thi$ finalparagraph, the fact that the rebound-effect also applies to the other produc-tion factors will be clarified by rneans of an example concerning theproduction factor labour. Besides this, an illustration wiil be given of theexistence of the rebound-effect in other fields of research.

The rebound-effect for the production factor labour can be illuatrated withthe historical example of the change in labour productivity in the 19th cen-tury. Since the industrial revolution, there has heen a labour-saving bias oftechnical change which was a result of the invention of new machines and

100 ECN-I--94-053

The rebound-effect in perspective

processes which replaced labour with capital and energy. The consequenceof this substitution was that the labour productivity, i.e. the output pro-duced per employee or per worker-hour, increased. According to Sonen-blum (1983), for example, the labour productivity in the U.S.A. increasedby a factor 2200 in the course of the fgth century. This increased labourproductivity worried many scholars of that time (Khazzoom; 1980,1986).They thought that a doubling of labour productivity would lead to an equalreduction in employment. Also Karl Marx drew this pessimistic conclusion.Marx thought that the diminished demand for labour would lead to massunemployment. This unemployment in tutu would cause the capitalisticsystem to collapse.

In rea]ity, however, the massive unemployment and consequent destructionof capitalism did hOt occur. The reason for this was the fact that the in-creased [abour productivity led to an increased economic growth at thattime. In fact, Fabricant (1983) even argues that the increases in labourproductivity have been the dominant source of economie growth in theU.S.A. over the past two centuries. The increased demand for goods andservices associated with this economie groWth led to an increase in thedemand for labour which more than offset the decrease in labour demandassociated with the improvements in 1abour productivity. And the net resultwas that employment increased, rather than decreased.

The same also applies to the increased utilization of computers and robotsin today’s manufaeturing processes. These production aids enable produc-tion workers to achieve a higher productivity, thereby making some wor-kers idle. The efficiency improvements however lead to an increased econ-omic growth. This means that while some jobs may be lost, others aregaine& And although the expert opinions on this subject are hot exactly ofthe same tenor, the net result is probably positive.

The rebound-effect thus also applies to the factor labour. But, there is adifference with the rebound-effect that results from energy efficiency im-provements. The labour rebound-effect namely bas positive instead of ne-gative consequences. The rebound-effect represents that part of the initiallyexpected saving that is lost. The result of lost energy savings is an in-creased depletion of out fuel reserves (unless use is made of renewables).Lost Iabour savings result in less unemployment. So, the rebound-effectwith respect to the factor labour should be maximized instead of minimized.

Besides the fact that there is a feedback-effect with respect to the produc-tion factors in an economie system, a rebound-effect is also encountered inother domains. This can be clarified by means of the example of the intro-duction of wide-bodied aircraft (Spare, 1990). In the middle of the 1960s,the predicted increase in air-transportation for the 1970s presented difficul-ties for this industry. The thought was that if the increased demand forair-transportation had to be met by a larger number of small planes such asthe Boeing 707, there would not be enough airline space available and theairports would hot be able to cope wlth the increased activity. In a responseto this threat, a jumbo-type aircraft was designed. On 22 January 1970,Pan Am began service with its first wide-bodied Boeing 747 on the NewYork-London route. The introduction of these large airplanes thus was ex-pected to lead to a decrease in the number of daily take-offs and landings

ECN-I--94-0.53 l 01

The 3-~ interaction and the rebound-effect

at airports. In reality, however, quite the opposite occurred. The efficiencyimprovemen’~s that were a result of the introduction of the larger airplanesnamely led, in a competitive marker, to lower air fares. Because of theselower air fares, people started flying further and more often. And the de-crease in the number of take-offs and landings resulting from the use of thebigger planes was more than offset by the increase in the number of take-offs and landings resulting from ’~he inqreased demand for air-travelling.

102 ECN-I--g4-053

9. CONCLUSIONS ANDRECOMMENDATIONS

The following conclusions can be derived from this study:1. The rebound-effect exists, This implies that apart of the initially ex-

pected energy savings, resulting from energy efficiency improvements,is lost because of the energy-economy-environment interaction.

2. The rebound-effect is comp[etely consistent with economic behaviour.Consumers do hot want to minimize energy costs, they r~ther want tomaximize utility.

3. The MARKAL-MACRO model is feasible to determine the magnitude ofthe rebound-effed. The MARKAL-MACRO estimation of the rebound-effect on the consumption-side of the economy emounts to 33%. Caremust however be raken when the MARKAL-MACRO results are extra-polated from the model to the real world.

4. The MARKAL-MACRO rêsults [ndicate that the rebound-effect is largerthan what the energy efficiency optimists assert, but smaller than whatthe energy efficiency pessimists claim.

5. The rebound-effect can be reduced by adopting energy efficiency mea-sures that go beyond the economie optimum. This reduction can how-ever only be obtained at the expense of a more or less distortion of theeconomic system.

6. There is no fi[th fuel, Because of the existence of the rebound-effect,energy efficiency improvements cannot be viewed as a new energysupply source.

The following recommendations can be mede:1. Additional research is required on the rebound-effect. The focus of that

research should be on a more accurate estimation of the magnitude ofthe rebound-effect and on measures to reduce the rebound-effect.

2. More research should also be carried out on the subject of the ener-gy-economy-environment interaction in general. The entropy conceptcan be seen as a useful tool in this context.

3. The MACRO component of the MARKAL-MACRO model should befurther developed as to allow for a better representation of the e¢on-omy.

4. The economic feedback-effects resu]ting from profitable reductions inpo]|ution (the Pol]ution Prevention Pays nostrum) shouid be examined.

ECN-I--94-0~3 ~ 03

REFERENCES AND LITERATURE

References

R.C. Adams and A.D. Rockwood (1983): Impact oflmproved BuildingThermal Efficiency on Residential Energy Demand. Pacific NorthwestLaboratory, Richand.

T. Anderson ( 1990 ): Letting Conservation Compete for Energy Dollars:A Policy lmperative. Sustainable Energy Choices for the 90s, ConferenceProceedings, 343-350.

H.J. Barnett anti C. Morse (1963): Scarcity and Growth. The JohnsHopkins Press, Baltimore.

M. B]aug (1968): Economic Theory in Retrospect. Richard D. Irwin,Homewood.

A.W. Boeke and P. A. van de Kraaij (1977): Climate Control. DelftUniversity of Techïology, Delft (in Dutch).

K.E. Boulding ( 1949): lncome or Welfare? Review of Economic Studies, 17,77-86.

K.E. Boulding (1973): The Economics of the Coming Spaceship Earth. InH.E. Daly (Ed.), Toward a Steady-State Economy, Freeman, SanFransis¢o, 121-132.

I. Boustead and G. Hancock (1979): Handbook oflndustrial EnergyAnalysis. ~Ellis Horwood Publishers, Chichester.

J.R. 13owdidge (1990): Environmental ]nsulation Thickness. Energy and theEnvironment, l, 307-317.

S. Boyle (I989): More Work forLess Energy. New S¢ientist, 5 August1989, 37-40.

L. Brillouin (1956): Science and lnformation Theory. Academic Press, NewYork.

L.G. Brookes (1990): The Greenhouse Effect, The Fallac~es in the EnergyE~ciency Solution. Energy Policy, Match 1990, 199-201.

L.G. Brookes ( ] 99 ] ): Confusing the Issue on Energy Ef~~ciency. EnergyPolicy, Match 1991, 184-186.

L.G. Brookes (1991): Letter to the Editor. Energy Policy, March 1991, 187.

L.G. Brookes (1992): Energy E~ciency and Economic Fallacies: A Reply.Energy Policy, May 1992, 390-392.

105

The 3-lE interaction and the rebound-effect

L.G. Brookes (1993): Energy El’ficiency Fallacies: The Debate Concluded.Energy PoIicy, Apríl 1993, 346-347.

A. Brooks (1988): GAMS: A User’s Guide. The World Bank,Washington D.C.

H. Brown, J. Boner and J. Weir (1957): The Next Hundred Years. VikingPress, New York.

G.H. Brundtland (1987): Our Common Future: The World Commission onEnvironment and Development. Oxford University Press, Oxford.

J. Bryant (1985): Econornics, Equilibrium and Thermodynamics. In W. vanGooi and J.J.IZ. Bruggink (Eds.), Energy and Time in the Economic andPhysical Sciences. North Holland, Amsterdam.

B. van Buuren and J.A. de Jong (1990): Physics. Educabook, Culemborg(in Dutch).

S. Carnot ( 1824): Reflections on the Motive Power of Heat and on Machinesfitted to Develop that Powen Macmillan, London.

Consumers Association (1990): Energy Conservation in her second Youth.Consumers Guide, September 1990, 510-523 (in Dutch).

E.F. Cook (1976): Man, Energy, Society. Freeman, San Fransisco.

R. Costanza (1980): Embodled Energy and Economic Valuation. Science,210, 1219-1224.

R. Costanza (1981): Embodied Energy, Energy Analysis and Economics. InH.E. Daly and A.F. Umana (Eds.), Energy, Economic$ and theEnvironment, Westview Press, Boulder, 119-145.

R. Costanza and H.E. Daly (1987): Towards an Ecological Econornics.Ecological Modelling, 38, 1-7.

W.F. Cottrell (1955): Energy and Society. McGraw-Hill, New York.

W.F. Cottrell (1972): Technology, Man and Progress. Merrill, Columbus.

H.E. Daly (1973A): lntroduction. In H.E. Daly (Ed.), Toward a Steady-StateEconomy, Freeman, San Fransisco, 1-32.

H.E. Daly (1973B): The Steady-State Economy: Toward a PoliticalEconomy of Bìophysical Equiiibrium and Moral Growth. In H.E. Daly (Ed.),Toward a Steady-State Economy, Freeman, San Fransisco, 149-174.

H.E. DaIy (1977): Steady State Economics. Freeman, San Fransisco.

H.E. Daly (1979): Entropy, Growth and the Political Econorny of Scarcity.In V.K. Smith (Ed.), Scarcìty and Growth Reconsidered. The Johns HopkinsUniversity Press, Baltimore. 67-94.

106 ECN-I--94-053

References and literature

Dupont de Nemours (1761): Physiocratie ou Constitutions essentielles duGouvernement le ptus auantageux au Genre Humain.

S, Fabricant (1983): The Productivity-Growth SIowdown: A Review of theFacts. In S.H. Schurr, S. Sonenblum and D.O. Wood {Eds.), Energy,Productivity and Economic Growth. Oelenschlager, Gunn and Hain,Cambridge, 47-70.

R.P. Feynman, R.B. Leighton and M. Sands (1966): The Feynman Lectureon Physics: Mainly Mechanica, Radiation and Heat. Addison-Wesley,Reading.

L.G. Fishbone, G. Giesen, G. Goldstein, H.A. Hymmen, K.J. Stocks, H. Vos,D. Wilde, R. Zolcher, C. Balzer and H. Abilock (1983): Uaer’s Guide forMARKAL. Brookhaven National Laboratory, Upton.

R.A. Gaggioli and E.F. Obert (1963): Thermodynamica. McGraw-Hill,New York.

N. Georgescu-Roegen ( 1971 ): The Entropy Law and the Economic Process.Harvard University Press, Cambridge.

N. Georgescu-Roegen (1973): The Entropy Law and the EconomicProblem. In H.E. Daly (Ed.), Toward a Steady-State Economy. Freeman,San Fransisco, 37-49.

N. Georgescu-Roegen (1975A): Energy and Economic Myths. SouthernEconomic Joumal, 3, 347-381.

N. Georgescu-Roegen ( 1975B): Dynamic Models anti Economic Growth.World Development, 11,765-783.

N. Georgeacu-Roegen ( 1979): Comments on the Papers by Daly andStiglitz. In V.K. Smith (Ed.), Scarcity and Growth Reconsidered. The JohnsHopkins Press, Balt[more, 95-105.

N. Georgescu-Roegen (1980): Matter: A Resource Ignored byThermodynamica. In L.E. St-Pierre and G.R. Brown (Ed.), Future Sourcesof Organic Raw Materials: CHEMRAWN 1, Pergamon, New York, 79-87.

N. Georgescu-Roegen (1981): Energy, Matter and Economic Valuation:Where Do We Stand? In H.E. Daly and A.F. Umana (Eds.), Energy,K¢onomics and the Environment, Westview Press, Boulding, 43-79.

N. Georgescu-Roegen (1986): The Entropy Law and the Economic Processin Retrospective. Eastern Economic Journal, 1, January/March, 3-25.

J.l. Gershuny (1977): Post h]dustrial SocMtg: the Myth of the SerukeEconomy. Futures, 2, 103-114.

J.l. Gershuny ( ] 978): After lndustrial Society: The Emerging Self-ServiceEconomy. Macmillan, London.

ECN-I--94-053 107

II II II I


R.W. Gilmer (1977): Services anti Energy in US Economie Growth. lnstitutef’or Energy Ana~ysis, Oak Ridge.

H.E. Goeller and A,Mo Weinberg ( 1978): The Age of Substitutability.American Economie Review, December, 1-11.

G. Greenhalgh (1990): Energy Conservation Polieies. Energy Policy, April1990, 293-299.

M.J. Grubb (1990): Energy Efficiency and Economie Fallacies. EnergyPolicy, October 1990, 783-785.

M.J. Grubb (1992): Reply to Brookes. Energy Policy, May 1992, 392-393o

L.D. Hamilton, G. Goldstein, J. Lee, A. Manne, W. Marcuse, S. Morris andC.-O. Wene (1992): MARKAL-MACRO: An Overaiew. Brookheven NationalLaboratory, Upton.

G. Hardin (1968): The Tragedy ofthe Commons. Science, December 13,1243-1248.

H. Herring and M. El]iott (1990): Letter to the Editor. Energy Policy,October 1990, 786.

E. Hirst and D. White (1985): lndoor Ternperature Changes After Retrofit:lnferences Based on ElectriÇity Billing Data for Nonparticipants andParticipants in the BPA Weathefization Program. Oak Ridge NationalLaboratory, Oak Ridge.

G. Jaeckel and P.A. Okken (1978): Economie Aspects ofthe Insulation ofDwellings. University of Groningen (in Dutch).

E.T. Jaynes (1957): Physical Review 106, 620.

W.S. Jevons (1865): The Coal Question. Macmillan, London.

W.S. Jevons (1879): The Theory ofPolitical Economy. Macmillan, London.

H.G. Jones (1975): An [ntmduction to Modern Theories of EconomieGrowth. The Camelot Press, Southampton.

C.T. Jones (1993): Another Look at US Passenger Vehicle Use and theRebound-Effect from lmproved Fuel Efficiency. The Energy Journal, 4,99-110.

B. de Jouvenel (1958): A Philosophy of lndian Economie Development.Navajibad Publishing House, Ahmedabad.

B. de Jouvenel (1959): Organisation du Travail et l’Amenagement del’Existence. Free University Quarterly, 7, ô.

108 ECN-1--94-053


B. Keepin and G. Kats (] 988): ûreenhouse Warming: ComparativeAnalysis of Nuclear and Efficiency Abatement Strategies. Energy Policy,December 1988, 538-561.

J.D. Khazzoom (1980): Economie lmplications of Mandated EfficiencyStandards for Household Appliances. The Energy Joumal, 4, 21-40.

J.D. Khazzoom (1986): An Econometrie Model lntegrating ConseroationMeasures in the Resìdential Demand for Electricity. JA1 Press, Greenwich.

J.D. Khazzoom (1987): Energy Savings Resulting from the Adoption ofMore Efficient Appliances. The Energy Journal, 4, 85-89.

J.D. Khazzoom (1989): Energy Savings from More Effìcient Appliances: ARejoinder. The Energy Joumal, 1, 157-166.

Krachtkroniek (1994): Prìces of Energy. May 1994 (ìn Duích).

F. Krause ( 1993): Cutting Carbon Emissions: Burden or Benefit?International Project Sustainable Energy Paths, El Cerrito.

T.S. Kuhn (1970i: The Structure ofScientific Revolutions. University ofChicago Press, Chicago.

S. Kypreos (1992): The MARKAL-MACRO Model, Links and PotentialExtensions. PSI, Villigen.

A.B. Lovins and H.L. Lovins (] 99 ] ): Least-Cost Climatic Stabilization.Annua] Review of Energy and the Environment, 16.

T.D. Macdaniel (1980): lnsulation Guide. K]uwer Technische Boeken,Deventer (in Dutch).

T.R. Ma]thus (1798): Essay on the Principles of Population.

A. Manne and R. Richels (]990): Carbon Dioxide Emission Limits: anEconomie Cost Analysis for the U.S.A.. The Energy Journal, 2, 51-59.

A.S. Manne and R.C. Richels (1992): Buying Greenhouse lnsurance: TheEconomie Costs of Carbon Dioxide Emission Lirnits. MIT Press, Cambridge.

K.M. Meyer-Abich (1979): Energieeinsparung als Neue Energiequelle. CarlHanser Ver/ag. Munehen.

J.S. Mill (1857): Principles ofPolit[cal Economy, Vol 2. J.W. Parker andSon, London.

E.J. Mishan (1967): The Costs of Economie Growth. Staples Press, NewYork.

W. Nordhaus (1990): Count Before You Leap. The Economist, 7 July 1990,19-22.

109

The 3-E interaction and the rebound-effect ~ .....

H.T. Odum ( 1971 ): Environment, Power and .Society. Wiley-lnterscience,New York.

H.T. Odum (1974): Energy, Ecology and Economics. Ambio, ô, 220-227.

H.T. Odum and E.C. Odum (1976): Energy Basis for Man and Nature.McGraw-Hill, New York.

H.T. Odum (1977): Energy, Value and Money. In C.A.S. Hall and J.W. Day(Eds.), Ecosystem/’4odeIling in Theory and Practice, WiIey-lnterscience,New York, 174-196.

OFCD (1979): lnterf~tures: Facing the Future. Paris.

W. Ostwald (1907): Modern Theory ofEnergetics. The Monist, 4, 481-515.

W. Oatwald (191 ~ ): Efficiency. Independent, 71,867-871.

K.F.B. de Paauw anti A.H. Perrels (1993): The Energy [ntensity ofConsurnption Paekages. Netherlands F,nergy Research Foundation.ECN-C-93-043 (in Dutch).

W. Petty (1662): A Treatise of Taxes and Contributions.

A.C. Pigou (1935): The Economics ofStationary States. Macm[flan, London.

S. Podolinski (1883): Menschliche Arbeit und Einheit der Kraft. Die NeueZeit, 1.

FoQuesnay (1758): Tableau Oeconomique.

E. Rabinowicz (1965): Friction and Wear ofMaterials. Wiley, New York.

J. Rees (1990): Natural Resources, Routledge, London.

D. Ricardo (18~.7): Principles of Political Economy and Taxationo

T.F.M. van Rossum (199 lA): Deoelopments in the Electricity Consumptionof Households. University of Utrecht, November 1991 (in Dutch).

T.F.M. van Rossum (199113): Developments in the Natural GasConsumption ofHouseholds. Unìversity of Utrecht, November 1991 (inDutch).

W.W. Rostow (1978): The Wodd Econotny: History and Prospect. Universityo~: Te×as Press, Austin.

M.D. Rowe and D. Hill (1989): Estimating National Costs of ControllingEmissions from the Energy System, A Report of the ETSAP Project.Brookhaven National Laboratory, Upton.

lE. $chrödinger (1955): What is life? University Press, Cambridge.

] ] 0 ECN-I--94-053

Referenees and literature

S.H. Schurr (1982): Energy Ef]ìciency and Productive Efficiency: SomeThoughts based on American Experience. The Energy Jouma], 3, 3-14.

C.E. Shannon and W. Weaver (1949): The Mathematical 7heorg oflnformatlon. University of l]linois Press, Urbana.

J.L. Simon (1981): The Ultimate Resource. Princeton University Press,Princeton,

A. Smith (1776): An lnquiry ~nto the Nature and Causes of the Wealth ofNations.

F. Soddy (1912A): Matter and Energy. London.

F. Soddy (1912B): Transmutations, the Vital Problem o[ the Future.Scientia, 11, 199.

F. Soddy (1922): Cartesian Econom~cs. Henderson, London.

F. Soddy (1926): Wealth, Viriual Wealth and Debt. George Allen andUnwin, London.

R.M. Solow (1973): Is the End of the World at Hand? Challenge,March/April, 39-50.

R.M. Solow (1974A): lntergenerational Equity and Exhaustible Resources.Review of Economic Studies, 29-45.

R.M. Solow (1974B): The Economics of Resoumes or the Resources ofEconomics. American Economic Review, May, ]- 14.

E. Solvay (1902): Note sur des Formules d’lntroduction a l’EnergitiquePhysio- et Psycho-Sociologique. H. LamartJn, Bruxelles.

S. Sonenblum (1983): Overview and Commentary. In S.H. Schurr,S. Sonenblum and D.O. Wood (Eds.), Energy, Productivity and EconomicGrowth, Oelgeschlager, Gunn and Hain, Cambridge, 3-46.

P.H. Spare (1990): The Fifth Fuel: The Debate Continues. Energy World,April 1990, 177.

H. Spencer (1880): First Principles. Appleton, New York.

Thirring (1958): Energy for Man: from Windmills to Nuclear Power. Harperand Row.

D. Toke (1990): lncreasing Energy Supply not lnevitable. Energy Policy,September 1990, 671-673.

D. Toke (1991): Energy Efficiency. Energy Policy, November 1991, 815.

ECN-I--94-053 111

The 3-E interaction and the rebound-efl:ect

A.F. Umana (1981): Toward a Biophysical Foundation for Economics. InH.E. Daly and A.F. Umana (Eds.), Energy, Economics and theEnvironment, Westview Press, Boulding, 21-41.

UNIDO (1985): lndustry in the 1980s: Structural Change andlnterdependence. New York.

J.R. Ybema and P.A. Okken (1992): Technology Characterizations2000-2040. Netherlands Energy Research Foundation, ECN-C-92-063 (inDutch).

H.C. Wallich (1972): Zero-Growth. Newsweek, January 24, 62.

A. Warren (1987): Saving Megabucks by Saving Megawatts. EnergyPolicy, December 1987, 522-528.

M. Weber (1958): The Protestant Ethic and the Spirit of Capitalism. HarperCollins, London.

W.A. Weisskop~ (1973): Economic Growth versus Existential Balance. InH.E. Daly (Ed.), Toward a Steady-State Economy, Freeman, SanFransisco, 240-251.

J.A. Weissmahr (1993): An Evolutionary Economic Theory for aSustainable Future. International Joumal of Globa[ Energy lssues, SpecialIssue on Energy Analysis, 1, 39-43.

R.H. Williams (1990): Low-Cost Strategies for Coping with Carbon DioxideLimits. The Energy Joumal, 3, 35-59

D. Wilson and J. Swisher (1993): Exploring the Gap: Top-Down versusBottom-Up Analyses of the Cost of Mitigating Global Warming. EnergyPolicy, Match 1993, 249-263.

L. Winiarski (1900): Essai sur la Mechanique Sociale: l’Energie Sociale etles Mesurations. Revue Philosophique, 49, 256-287.

112 ECN-I--94-053


Literature

L.A. Arnoux (1985): The Value of Energy/The Energy of Value. In W. vanGooi and J.J.C. Bruggink (Eds.)0 Energy anti Time in the Economic andPhysical Sciences, Nortl»Holland, Amsterdam, 343-379.

R.U. Ayres and 1. Nair (1984): Thermodynamics and Economics. PhysicsTody, November 1984, 62-71.

W. Beckerman (1972): Economist.s, Scientists and EnvironmentalCatastrophe. Oxford Economic Papers, May 1972, 327-344.

K.K. Boulding (1976): Economics for Good or Evil. Technology Review,JuIy/August 1976, 5.

L.G. Brookes (1978): Energy Policy, the Energy Price Fallacy and the Roleof Nuclear Energy in the UK. Energy Policy, June 1978, 94-106.

J.J.C. Brugglnk (1985): The Theor~l of Economic Growth andThermodynamic Laws. In W. van Gooi and J.J.C. Bruggink (Eds.), Energyand Time in the Kconomic and Physical Sciences, Norlh-Holland,Amsterdam, 135-145.

P.P. Christensen (1987): Classical Roots for a Modern Materials-EnergyAnalysis. Ecological Modell[ng, 38, 75-89.

C.J. Cleveland ( 1987): Biophysical Economics: Historical Perspectíve andCurrent Research Trends. Ecological Modelling, 38, 47-73.

H.E. Daly (1980): The Economic Thoughts of Frederick Soddy. History ofPolitical Economy, 4, 469-488.

P.S. Dasgupta and G.M. Heal (1979): Economic Theory and ExhaustibleResources. University Press, Cambridge.

E.F. Denison (] 979): Explanations o[ Declining Productivlty Growth.Survey of Current Business, 8.

R.M.E. Diamant (1986): Thermal and Acoustic lnsulation. Butterworths,London.

M. Faber, H. Niemes and G. Stephan (1987): Entropy, Environment andResout~es. Springer-VerIag. Berlin.

J.D. Fast (1954): Entropy in Science and Technology, Part 1: The EntropyConcept. Philips Technisch Tijdschrift, 7, 202-214 (in Dutch).

J.D. Fast (1954): Entropy in Science and Technology, Part 2: Exarnplesand Applications. Phi]ips Technisch Tijdschrift, 9, 278-288 (in Dutch).

J.D. Fast ( 1954): Entropy in Science and Technology, Part -~: Examplesand Applications (continued): Philips Technisch Tijdschrift, 11, 336-348 (inDutch).

ECN.I--g4-053 113


J.D. Fast (1956): Entropy in Science and Technology, Part 4: Entropy andInformation. Philips Technisch Tijdschrift, 56, 170-182 (in Dutch).

S. Faucheux (1993): The Role of Energy in Production Functlons.International Journal of Global Energy lssues, 1, 41-55.

K.J.V. Fowler (1985): Design of Thermal Loada for Building Structures.Bath Press, Avon~

J.C. Franken (1965): Thermodynamica. E.J. Brill, Leiden (in Dutch),

N. Georgescu-Roegen (1972): Process Analysis and the NeoclassicalTheory ofProduction. American Jouma[ of" Agricultural economics, 54,279-294.

IN. Georgescu-Roegen (1982): En ergetic Dogma, Energetic Economics andViable Technologles. In J.R. Moroney (Ed.), Advan(ìes in the Economics ofEnergy and Resources. Jai Press, Greenwich, 1-39.

M. Giampietro and D. Pimentel (1991): Energy Efficiency: Assessing thelnteraction between Humans and their Enüironment. EcologicalEconomics, 4, 117- ] 44.

W. van Gooi (~ 982): Two-Factor and Multi-Factor Trade-Offs in EnergyConser~ation. In Economic Theory o1: Naturel Resources, Physica-VeHag,Wurzburg-Wien, 515- 531.

W. van Gooi (1985): Entropy and Exergy. In Gooi, W. van and J.J.C.Bruggink (Ed.), Energy and Time in the Economic and Physical Sciences,North-Holland, Amsterdam, I 1-15.

C.A.S. Hall, C.J. Cleveland and R. Kaufmann (1992): Energg and ResourceQuality. University Press, Colorado.

B. Hannon (1975): Energy Consewation and the Consumer. Science, 189,95-102.

J.A. Hausman (1979): lndividual Discount Rates and the Purchase andUtilization ofEnergy-Using Durables. The ~3e~~ Jouma~ of Economics,Spring 1979, 33-54.

J.A. Hausman and P.L. Joskow (1982): E~aluting the Costs anti Benefits ofAppliance Efficiency Standards. AEA Papers and Proceedings, 2, 220-225.

D.I, Hertzmark (1978): Economics, Thermodynamica and the Allocation ofEnergy. University of Wisconsin, Madison.

W.W. Hogan and A.S. Manne (1977): Energy-Economy lnteractions: TheFable of the Elephant and the Rabbit. In C. J. Hitch (Ed.), ModellingEnergy-Economy lnteractions: Five Approaches, Resources for the Future,Washington D.C.

1 14 ECN4--94-053


W. Hoogendijk ( 1991): The Economic Revolution. Green Print/Jan vanArkel, Utrecht.

R. Hueting (1974): New Scarcity and Economic Growth. Agon Elsevier,Amsterdam (in Dutch).

R. Hueting (1987): An Economic Scenario that gives Top Priority to Savingthe Environment. Ecological Modelling, 38, 123-140.

R. Hueting (1990): The Brundtland Report: A Matter of Conflicting Goa/s.Ecological Economics, 2, 109-1

D.W. Jorgenson (]986): The Great Transition: Energy and EconomicChange. The Energy Journal, 3, 1-13.

M. Kamenetzky and R.H. Maybury (Ig8g): Technical and FinanclalRationality Concern for Human Needs and Ecosysterns: A ConceptualFramework. Science and Public Policy, 2, 68-72.

J.D. Khazzoom, M. Shelby and R. Wolcott (1990): The Conflict betweenEnergy Conservation anti Environmental t~olicy in the US TransportationSector, Energy Policy, June 1990, 456~158.

T. Kram (1993): National Energy Options for Reducing COz Emissions,Volume h The International Connection. Netherlands Energy ResearchFoundation, ECbI-C- ] 993-101.

J. Martinez Alier and J.M. Naredo (1982): A Marxist Precursor of EnergyEconomics: Podolinsky. The Journal of Peasant Studies, 9,207-224.

M. Munasinghe and G. Schramm (1983): Energy Economics, DemandManagement and Conservation Policy. Van Nostrand Reinhold, New York.

R. Pindyck (1979): The Structure of World Energy Demand. The MIT Press,Cambridge.

J.L.R. Proops (1985): Thermodynamics and Economics: From Analogy toPhysical Functioning. In W. van Gooi and J.J.C. Bruggink (Eds.), Energyand Time in the Economie and Physical Sciences, North-Holland,Amsterdam, 155-174.

B. Richmond (1980): A Dgnamic Framework for lnvestigating Policies tolncrease the Energy Efficiency of the lndustrial Secton Proceedings of thelnternatíonal Conference on Cybernetics and Society, 15-18.

E. Roll (1973): A History ofEconomic Thought. Faber, London.

G. Routh ( 1975): The Origin of Economic ldeas. Macmillan, London.

E.F. $chumacher (1973): Buddhist Economics. In Daly, H.E. (Ed.), Towarda Steady-State Economy, Freeman, Sari Fransi$co, 231-239.

115


J.L. Sweeney (1984): The Response of Energy Demand to Higher Prices:What have we Learned? American Economic Review, 2, 31-37.

M. Webb and M. Ricketts (1980): The Ec’onomics of Energy. Macmìllan,London.

T. Weyman-Jones (1986): The Economics of Energy Policy. GowerPublishing Company, Aldershot.

L.J. Zimmerman (1968): History ofEconomic ldeas. Albani, Den Haag (inDutch).

] 16 ECN-I--94-053

the energy-economy-environment … 1995 ecn-i--94-053 the energy-economy-environment interaction and...

Documents