The Everyday Life Context of Increasing Energy Demands: Time Use Survey Data in a Decomposition...

Vol. 9, Issues 1-2 - Winter-Spring 2005 A special issue of the Journal of Industrial Ecology, guest edited by Edgar Hertwich, program leader of the Industrial Ecology Programme and a professor in the Department of Energy and Process Engineering at the Norwegian University of Science and Technology in Trondheim, Norway. Support for this special issue was provided by the Garfield Foundation and the United Nations Environment Programme, Division of Technology, Industry, and Economics. If you'd like to stay informed with the latest news and articles from The Journal of Industrial Ecology via e-mail alerts, or if you would like information about our subscription rates, please visit us online at: http://mitpress.mit.edu/jie/consumption

http://www.tev.ntnu.no/edgar.hertwich

http://www.indecol.ntnu.no/

http://www.uneptie.org/

http://www.uneptie.org/

http://mitpress.mit.edu/jie/consumption

http://mitpress.mit.edu/jie/consumption

RESEARCH AND ANALYSIS

The Everyday Life Context ofIncreasing Energy DemandsTime Use Survey Data in a DecompositionAnalysis

Mikko Jalas

Summary

Industrial ecologists have modeled with precision the materialfoundations of industrial systems, but given less attention tothe demand for products and the drivers of structural changesin these systems. This article suggests that time use data com-plement data on monetary expenditure and can be used toelucidate the everyday life context in which the changes in theeconomy take place. It builds upon the claim that goods arenot direct sources of utility, but enter specific household activ-ities as inputs. A second argument for the proposed approachis that it can be used to introduce and foster human agency inanalyses of production systems. The article uses Finnish timeuse survey data, consumption expenditure data, and data onthe sectoral energy intensities of financial output in the Finnisheconomy. First, a measure of the energy intensity of activitiesis derived by relating consumer time use and the requireddirect and indirect energy requirements. Second, the resultsinclude a decomposition of changes in the energy require-ments of private consumption in Finland during the 1990s. Itis shown that although the same activities on average requireincreasing energy inputs per unit of time, Finns have simulta-neously changed the structure of their everyday life towardless energy-intensive activities.

Keywords

consumptionenergyhouseholdindustrial ecologyinput-output analysis (IOA)temporal

e-supplement available on the JIEWeb site

Address correspondence to:Mikko JalasHelsinki School of EconomicsLapuankatu 6, FIN-00101 Helsinki<[email protected]><www.hkkk.fi/organisaatiot/faculty/jalas.htm>

© 2005 by the Massachusetts Institute ofTechnology and Yale University

Volume 9, Number 1–2

http://mitpress.mit.edu/jie Journal of Industrial Ecology 129

R E S E A R C H A N D A N A LYS I S

Introduction

The IPAT identity, since its introduction inthe early 1970s, has been a powerful conceptualtool in environmental discussion. By relatingpopulation (P), affluence (A), and technology(T) to environmental impact (I), the equationI = P·A·T has framed debates on the causes andthe remedies of environmental problems. In aworld of growing population wishing to raise itsstandard of living, technological developmenttoward higher efficiency of resource use is anirresistible imperative. The technological factorT has thus received great interest in studiesof industrial metabolism (Chertow 2000).Commonly, the analyses further distinguishstructural changes in the economy and efficiencyimprovements within sectors (for example,Farla and Blok 2000; Hoffren et al. 2000).The levers of and the conditions for structuralchanges have received less interest, and char-acterization of demand remains elusive in thefield of industrial ecology. Thus, it has beenalso claimed that industrial ecology is lackinga theory of human agency (Jackson and Clift1998; Andrews 2000) and risks being reducedto a mere study of efficiency (Jackson and Clift1998).

In this article I argue and demonstrate thattime use survey data can be used to contextu-alize environmentally relevant changes in thestructure of demand. Although such an approachis novel in the field of industrial ecology, priorexamples from other fields exist. For example,Gershuny has claimed that “. . . if we are to un-derstand the processes of structural change in ‘theeconomy’ [referring to the ‘formal economy’], weneed to consider evidence about behaviour out-side it: we need to know more about the everydaylife” (1987, 57). In other words, he claims that afocus on time use might bring forward a morecomprehensive understanding of demand thanmonetary data alone. Whereas Gershuny wasinterested in tracing the employment effects ofchanging patterns of time use, the same questionis repeated in this article in terms of the energyrequirements. Essentially, I ask how changingpatterns of time use affect the direct and indi-rect energy requirements of private households.

Are lifestyles becoming more energy-intensiveand what particular changes in everyday life im-ply major increases?

Anthropologists (e.g., Sorokin and Merton1990; Coser and Coser 1963), historians(Thompson 1974; Cross 1993), and economists(Becker 1965; Linder 1970) have a long tradi-tion of focus on time and temporality as a socialphenomenon. For the current purpose, the workof Linder is especially relevant, in that he com-pellingly predicts that labor productivity growthwill increase the value of time and that materialgoods will be increasingly used and consumedto save time. This body of literature and theavailable statistics thus represent an underusedand potent resource in the discussion of the en-vironmental impacts of consumption. Pursuingsuch aims, Schipper and colleagues (1989) pointout in their early article that patterns of timeuse can be used as one explanatory factor forchanges in the aggregate use of energy. Godbey(1996) follows the line of Linder (1970) and fur-ther explicates the environmentally relevant out-comes of time famine. He claims that time famineand the resulting search for time-saving technol-ogy prompt the use of resource-intensive prod-ucts and imply higher levels of municipal solidwaste. Røpke (1999) and Binswanger (2001) ad-dress the issue of time squeeze with a differentargument, claiming that time-saving technolo-gies are themselves essential constituents of in-creasing levels of consumption. Furthermore, theconsumer society as a whole has been seen todepend on a deliberate or forced choice of work-and-spend (Schor 1991; Cross 1993; Aronowitzand DeFazio 1994; Sanne 2000) and on thechoice of material wealth instead of wealth-in-time (Reisch 2001). However, these observationson the founding elements of the consumption-oriented society have rarely been translatedinto studies of societal energy and materialsflows.

This article mainly pursues a methodologi-cal point as it seeks to expand energy analy-ses and commonly used decomposition methodswith regard to the everyday life of households.I will first repeat some of my earlier argumentsfor considering the temporality of consumption(Jalas 2002) and then proceed to demonstrate

130 Journal of Industrial Ecology


the use of the time use approach in an analysisof Finnish patterns of energy consumption duringthe 1990s. The analysis combines time use sur-vey data, consumer expenditure data, and data onthe sectoral energy intensities of financial outputof the Finnish economy, measured in megajoulesper Finnish mark [MJ/FIM].1 The outcomes ofthe analysis are twofold. First, they include a de-scription of the changes that occurred during the1990s in the energy intensity of consumption ac-tivities, measured in megajoules per hour [MJ/hr],which refers to the required inputs per unit oftime. The second part of the results is a decom-position analysis of the changes in the direct andindirect energy demand of private final consump-tion in Finland from 1990 to 1998, in whichthe decomposition factors include demographicvariables, patterns of time use, and the above-mentioned energy intensity figures of individualactivities.

Some guiding remarks may be made on thedecomposition factors of this study. Althoughthe intensity factor resembles the technical “effi-ciency” factors of earlier studies, it must be notedthat the intensity factor describes the technologyof everyday life in a very broad sense. For ex-ample, the intensity of car driving can of coursebe impacted by fuel efficiency, but it can alsobe impacted, more radically, by the number ofpassengers present in the car and the averagespeed of the car. Therefore the intensities of ev-eryday life activities hardly represent technicalefficiency alone. It should also be noted that thecurrent analysis does not allow, for practical rea-sons, any changes in production efficiency in theformal economy. However, there are no obsta-cles to inserting this variable into the proposeddecomposition framework.

It appears that both the patterns of time useand the energy intensities of the activities havechanged over the course of the 1990s. The fol-lowing section presents some theorizing on thedynamics of time use and substitutability betweentime and goods, but it is out of the scope of this ar-ticle to engage in a deep discussion of the reasonsfor the descriptive outcomes of the analyses per-formed. The concluding section briefly addressesthe fit between the results and the previously pro-posed relationships between time and goods.

A Time Use Approach toConsumption

Time use, or rather the rational allocation oftime, has been theorized in the field of householdeconomics (Becker 1965). This economic ortho-doxy represents humans as capable and willingmaximizers of their own utility. All consumptionactivities are thus thought to be driven by cog-nition of a goal and of the available means forstriving toward it. Accordingly, time is a scarceresource, which is sold on the labor market andcarefully invested in various free-time activitiesso that the marginal benefit of a unit of time is thesame in all activities. Because preferences are re-garded as stable, it is the opportunity costs and thehousehold technology that determine changes intime allocation decisions in this frame.

The psychological and sociological studies ofconsumption give rise to alternative theories ontime use. It has been claimed that the meaningsand goals of consumption have become more dif-ficult to interpret and more malleable (Firat andDholakia 1998). Goods are tools only in a lim-ited sense, as they simultaneously constitute, to-gether with language, our symbolic social reality(Dietmar 1992). In this reality, consumers mayincreasingly prefer to playfully engage in certainactivities, be it for reasons of self-actualization,interpersonal communication, or social cohesion.Substantial changes in consumption are parallel.On the one hand, the relative share of expendi-ture on such elementary items as nutrition hasdeclined. On the other hand, there is a growingsector of advertisement and marketing, which isharnessed to influence demand and mold prefer-ences (Røpke 1999). Hence, although it may beconvincingly argued that the drudgery of elemen-tary needs is only transformed into new “basic”needs, it also seems fair to claim that these newneeds are less obvious and stable. In other words,there are fewer granted necessities, or “functionalunits,” as in life-cycle assessment (LCA) stud-ies, against which to project efficiency consider-ations. There is also an important, long-standingsociological tradition pointing out that rationaldecisions of time allocation are also conditionedby various regulating institutions, as well as bycultural and social norms (Sorokin and Merton

Jalas, Time Use Survey Data in a Decomposition Analysis of Energy Demand 131


1990; Coser and Coser 1963; Cross 1993). Framedby these psychological and sociological expla-nations, time use as a phenomenon is thus nota mere reflection of available technology, butrather gains a status that is independent of thetechnologies that are being used.2

I have previously argued that time use dataare more comprehensively helpful to the analysisof consumption and rebound effects than expen-diture data alone (Jalas 2002). The argument isbased on the fact that time, unlike economic re-sources, is absolutely finite, and thus increases ina certain type of time use must be matched withdecreases in others. In other words, the time useapproach forces a step from analyzing consump-tion as expenditure toward a more tightly inter-woven system of consumption activities, whichtogether constitute a lifestyle. In such a system,the expenditures are interdependent not only viaa budget constraint, but also due to the varioususes of the acquired products and services.

Yet there is more to consumption than budgetsand restraints, be they monetary or temporal. Hu-mans pursue different, co-existing aims in theirconsumption activities and comply with differentnorms; apart from fulfilling functional and instru-mental roles, goods constitute symbolic social re-ality, enable communication, and serve the con-struction of identities (Douglas and Isherwood,1979; Belk, 1998; Dietmar, 1992). If consump-tion and demand are to be addressed as a broadsocial phenomenon, these widely accepted claimsshould be incorporated; they prompt the questionof how to introduce a less instrumental view ofconsumption and accommodate fragmented ra-tionality within the analysis of the economic sys-tems and the associated material flows.

In this article I claim that industrial ecologycould benefit from viewing consumer preferencesas temporal. In other words, consumption shouldbe regarded as a set of temporal activities in whichconsumers utilize or engage with the various prod-ucts of industrial systems and through which re-source flows pass, virtually or in the sense of in-duced, indirect flows. Accordingly, resource flowsenable the various ways in which consumers desireor come to spend their time and should be analyzedin respect to time use.

The proposed analytical frame is rooted inboth the economic and the sociological strands of

theorizing described above. First, it builds on thework of household economists and in particularon the claim that market products are not directsources of utility as such. Rather, goods enter thevarious activities of households as inputs, andthus the “production functions” of householdactivities establish a direct link between theactivities and the demand for market goods. Al-though the proposed approach thus partly stemsfrom a tradition that treats households essentiallyas rational mini-firms, I argue that the psycho-logical and the sociological theories of time useare necessary for and capable of introducinghuman agency into the analyzing frame.

According to Andrews (2000), human agencyresides in the obstacles and limitations of humanreasoning. In a nutshell, the psychological andthe sociological theories of time use claim thatthe rationality of time allocation is, first, frag-mented between the immediate affective out-comes and the instrumental reasoning of a longerterm, and, second, conditioned by the social con-text of human action. The suggested frame ofanalysis, which includes an independent time usefactor, leans heavily on the noneconomic expla-nations and the associated claims of limited ra-tionality of time allocation decisions. Hence, itworks toward a more comprehensive view of thecontext of structural changes in demand as wellas toward answering some of the concerns forconsidering the human agency in such analyses.

Time Use Data in aDecomposition Analysis

In lieu of arguing that the changes in timeuse can impact (energy) demand independent oftechnological changes, I have previously assumedthat household technology, or the (energy)inputs per unit of time in various activities, re-mains essentially the same (Jalas 2002). This as-sumption is empirically examined in the currentarticle. The decomposition analysis thus includesa factor indicating changes in patterns of time useand a factor indicating energy intensity of activi-ties, which is essentially the ratio of energy inputsto time use.

The subject of the analysis is the primary en-ergy that households use directly and indirectlythrough their purchase of energy carriers as well



as various products and services. The proposedframe of analysis draws on available time use sur-vey data in order to decompose monetary andrelated energy demand into two further factors:(1) structural changes in how individuals usetheir time and (2) intensity changes, which referto the energy inputs per unit of time in each activ-ity category. In other words, the analysis seeks toanswer the question of whether the phenomenonof rising energy requirements is due to consumerschanging toward more energy-intensive activi-ties or to the increasing energy requirements ofactivities in general.

The analysis requires that everyday life be se-quenced into a limited number of time use cate-gories. In this matter the analysis relies on the ag-gregate level of an activity classification that hasbeen proposed in the report on European effortsto harmonize time use studies (Eurostat 1999a)and applied in the most recent Finnish time usesurvey.

The second issue to be addressed is theallocation of consumption expenditures tospecific activities. Previous attempts at such astatistical maneuver have been variously called“matching” (Chadeau and Roy 1986), “transla-tion” (Gershuny 1987), and “recategorization”(Brodersen 1990). In addition to these schemes,the linking of time use and expenditure dataat the macro level has also been proposed anddemonstrated by Juster and colleagues (1981)and also by Eurostat (1999b) in an effort toextend national accounting toward householdproduction. The scheme of Brodersen (1990) hasserved as a suitable basis for the current work.The time use categories in the analysis presentedhere are different from those of Brodersen,but his principle of allocating expenditures tothe activities that are directly connected withthe use of specific goods and services has beenapplied in this work as well.

These maneuvers having been performed, itis possible to calculate the energy requirementsper unit of time [MJ/hr] in the householdactivities. This is achieved by (1) multiplyingeach expenditure item [FIM/day] by the sectoralintensity of the financial output [MJ/FIM], (2)summing all derived energy requirements ofa specific activity, and (3) dividing this totalenergy requirement by the time use in the

activity [hr/day] (see the appendix for a moreformal representation of the procedure and thee-supplement on the journal Web site for thedetails of the categorization schema).

The Classification Scheme andthe Data of the Current Work

The data used in the current study stem fromrepresentative surveys performed by StatisticsFinland and describe two discrete points in time.The first set of data includes a household expendi-ture survey from 1990 and a time use survey from1987–1988, and the latter set, a household sur-vey from 1998 and a time use survey from 1999–2000. All of the consumer expenditure data areadjusted by the consumer price index to accountfor inflation.

In order to link household expenditures to thecorresponding use of energy in the economy, thestudy makes use of the national input-output ta-bles from 1995 and the sectoral energy intensitiesof the financial output [MJ/FIM] of the nationaleconomy. These intensities have been computedaccording to a four-digit COICOP-coding (Clas-sification of Individual Consumption Accordingto Purpose) of the national accounts.3

The time use and expenditure data have beencollected separately from different sets of house-holds. Hence, the data on time use and consump-tion expenditure must be combined against se-lected background variables. This is done at theaggregate level of life-stages of families. This cat-egorization was selected for use in the presentanalysis, as family life-stages have been shown tosignificantly affect both patterns of consumptionexpenditure (Statistics Finland 2000) and timeuse (Robinson and Godbey 1997). The 1987–1988 time use survey does not include data of allindividuals of the same household and, thus, ithas been necessary to construct meta-data thatdescribe the time use of households at differentlife stages. The time use data represent the pop-ulation over 10 years of age. In the following, allchildren below 10 are also approximated by thedata representing older children. Table 1 presentsthe compilation of the household-type-specificdata.

The time use categories used in the analysisare presented in table 2. The analysis is built on



Tabl

e1

The

pairi

ngof

indi

vidu

altim

eus

eda

taw

ithho

useh

old-

type

-spe

cific

expe

nditu

reda

ta

Hou

seho

ldty

pe

Expe

ndit

ure

data

1ad

ult

1ad

ult+

2ad

ults

2ad

ults

+1

child

3ad

ults

Oth

ers

1–2

child

ren

(+2

child

ren)

(+1

child

)[+

3ch

ildre

n]

Tim

eus

eda

taA

sing

lepe

rson

Asi

ngle

-cus

tody

wom

an+

ape

rson

unde

r45

livin

gw

ith

his/

herp

aren

ts∗t

heav

erag

enu

mbe

rof

child

ren

inth

ese

hous

ehol

ds

Am

arri

edm

anan

da

mar

ried

wom

anw

itho

utch

ildre

n

Am

arri

edm

anan

da

mar

ried

wom

anw

ith

ach

ild/c

hild

ren

+a

pers

onun

der4

5liv

ing

wit

hhi

s/he

rpa

rent

s(∗2

)[∗3

]

Am

arri

edm

anan

da

mar

ried

wom

anw

ith

the

youn

gest

child

bein

g7–

17+

ape

rson

unde

r45

livin

gw

ith

his/

her

pare

nts(

∗2)

Am

arri

edm

anan

da

mar

ried

wom

anw

ith

ach

ild/c

hild

ren

+a

pers

onun

der4

5liv

ing

wit

hhi

s/he

rpa

rent

s∗th

eav

erag

enu

mbe

rof

child

ren

inth

ese

hous

ehol

ds

Not

e:T

hepa

rent

hese

sand

the

squa

rebr

acke

tsre

fert

odi

ffere

ntm

odifi

cati

onso

fthe

sam

eba

sic

hous

ehol

dty

pe(i

.e.,

incr

easi

ngnu

mbe

rsof

child

ren)

.Ast

eris

ksin

dica

tea

mul

tipl

icat

ion

ofth

ere

leva

ntfa

ctor

s.



Table 2 Time use categories analyzed in this study and the related energy requirements

Time use category Consumption items and direct energy requirements

Work-related trips (02)School and study trips (13)

The share of transportation-related goods, consumables, andservices according to the share of these trips in overallmobility by distance traveled. Shares according to theMinistry of Transportation (1999).

Household work (03) andMaintenance work (04)

Household equipment; water; electricity; other relatedconsumables; equipment repair; maintenanceconsumables; tools; gardening materials; pets and pet food.

Services and civic matters (25/82) Medical services; health and beauty services; financialservices; other services.

Shopping and family business (07)excluding Services and civic matters(25/82)Trips related to housework,shopping, and childcare (08)

Trips related to daily shopping and family business(calculated in the same fashion as the work-related trips).

Eating (10) All foodstuff and services such as restaurants and canteens.Personal hygiene and dressing up (11) Goods and consumables for personal hygiene; beauty-related

goods.Outdoor activities and sports (16) Equipment; services such as entrance and course fees and

rents; equipment repair.Culture and amusement events (17) Entrance fees to spectator sports and amusement and culture

events.Reading (18) All books and magazines except for education-related books,

encyclopedias, and dictionaries.TV viewing (20) TV and VCR equipment; VCR recordings; program services;

equipment repair; electricity.Telephone calls (64/82) Telecommunication equipment; servicesHobbies (23) Audio equipment; recordings; photography-related goods

and services; musical instruments; computer equipment;games; toys; gambling; electricity; no telecommunicationcosts.

Leisure-time travel (25) All trips other than those relating to work, shopping andfamily business, housework, and childcare (calculated insame fashion as work-related trips); tourism services; boats.

Sleeping (09) No energy requirements allocated.Household infrastructure, no time use

specifiedAll other items, most significantly, heating, housing,

lighting, furniture, and clothing.

Note: Numbers in parentheses refer in general to a 26-activity classification scheme and those with a /82 to an 82-activityclassification scheme (see the work of Niemi and Paakkonen 2001 for further information on the classification schemes).The e-supplement presents the allocation of expenditure items in detail.

the same categorization that has been used in theinitial time use surveys and includes 26 activitycategories. Both the categorization and the allo-cation of expenditures are guided by the alloca-tion principles used by Brodersen (1990): activi-ties that require direct inputs of market goods aredistinguished and then each of them is matchedwith a set of activity-specific energy inputs.

The allocation is relatively straightforward forsome activities. For example, television (TV) and

videotape (VCR) equipment can be assigned towatching TV. But many time use categories areless-specifically defined by the objects that arebeing used. For example, it remains to be studiedwhat physical inputs are utilized during the timethat is reported as spent in socializing. On theother hand, some goods have an infrastructurerole in the everyday life of consumers. Housingand related heating and lighting are significant interms of natural resources, but cannot reasonably



be allocated to specific activities. In the follow-ing, all such goods that cannot be allocated tospecific activities will be denoted as householdinfrastructure; they include other items such asclothing in addition to housing-related expendi-tures, but the latter is clearly the most significantin terms of both money and energy.

Some categories have been excluded from theanalysis and some have been combined due to thelack of suitable criteria or data for allocating en-ergy requirements to specific activities. For exam-ple, work- and education-related trips have beencombined and the activities relating to socializinghave been left out of the analysis. Furthermore,time-wise education is not a significant categorybecause the energy inputs of this activity mainlycomprise significant public consumption expen-ditures, which are not included in the currentanalysis.

There is no single “right” categorization ofactivities, much less a means-ends configurationbetween these activities. Individuals who reportengaging in the same activity may assign differ-ent meanings to their activity (Elchardus 1991).In the same vein, what constitutes infrastructureis an arbitrary question. Some may regard, forexample, meals or personal hygiene as necessary“infrastructure” for conducting other meaningfulactivities. However, the categorization and allo-cation principles that have been applied in thisanalysis lead to distinction of all such activitiesthat require specific physical inputs, be they “pro-ductive” activities or “consumption.” The catego-rization thus distinguishes, for example, personalhygiene as an activity, which requires hot waterand consumables (e.g., soap). It also allocates allcooking utensils to the activity of preparing foodinstead of the activity of having meals; similarly,all renovation-related goods and consumables areallocated to the activity of maintenance work in-stead of the household infrastructure category.Altogether, the activity categories in table 2 rep-resent a partly arbitrary attempt to decomposeeveryday life into sequences, toward which hu-mans orient their attention.

Because of the excluded time use categories,the daily time use does not sum up 24 hours.The activities specified in table 2 occupied, onaverage, 18.7 hr in a day in the survey of 1987–1988 and 19.3 hr respectively in the survey of

1999–2000. If one considers the average work-ing time, for males above 10 years of age of 3 hr12 minutes (min) and for females above 10 yearsof age of 2 hr 12 min in 1987–1988, the cover-age encompasses more than 20 hr of the daily24 hr. The household expenditures that havebeen allocated to the specified time use accountedfor 60% of all household expenditures in 1990and the energy use that is related to these activi-ties covered 61% of the direct and indirect energyuse of Finnish households in 1990.

Results of the Analysis

Energy Intensity of Household Activities

This analysis aims to illustrate the energyrequirements of specific patterns of time useand the impacts of the changes in these pat-terns over a decade on the energy requirements.In the following, I will first consider the is-sue in the light of average per capita figuresand then presents some results, which are ar-ranged by household type. Table 3 presents theresults of recategorizing expenditure data ac-cording to the time use categories presented intable 2. The presented data include the sum ofall related expenditures, the consequent energyrequirements, the time use per person, and theenergy intensity of the activities for both pointsof observation, 1987–1990 and 1998–2000, andfor each activity.

Each of the analyzed activities is relatedto multiple expenditure categories. For exam-ple, leisure-time travel includes expenditures ontourist services and on capital goods such as boatsand a share of vehicles, as well as a share of fuelexpenditures. Thus no one-to-one link over timeexists between expenditures and energy, eventhough the analysis includes only one set of datafor 1995 derived from the input-output tables de-scribing the sectoral intensities [MJ/FIM]. How-ever, it can be observed that expenditure lev-els and energy requirements are closely linked intable 3.

The energy and time use data of selectedhousehold activities are presented in figure 1along the axis of the average time use per per-son and the average energy requirement re-lated to the activity. The arrows describe the



Tabl

e3

The

expe

nditu

res,

ener

gyre

quire

men

ts,t

ime

use

and

ener

gyin

tens

ityof

hous

ehol

dac

tiviti

esin

Finl

and

in19

87–1

990

and

in19

98–2

000

1987

–199

019

98–2

000

Ene

rgy

Ene

rgy

Exp

endi

ture

sE

nerg

yT

ime

use

inte

nsity

Exp

endi

ture

sE

nerg

yT

ime

use

inte

nsity

Act

ivity

[FIM

/yr/

pers

on]

[MJ/

yr/p

erso

n][h

r/da

y/pe

rson

][M

J/hr

][F

IM/y

r/pe

rson

][M

J/yr

/per

son]

[hr/

day/

pers

on]

[MJ/

hr]

Leis

ure-

tim

etr

avel

6,93

318

,096

0.60

835,

804

16,1

210.

5383

Wor

k-an

ded

ucat

ion-

rela

ted

trip

s2,

725

8,64

30.

3177

2,51

98,

152

0.31

73H

avin

gm

eals

11,3

6019

,050

1.26

4111

,600

19,7

621.

3141

Pers

onal

hygi

ene,

dres

sing

up1,

093

10,0

140.

8034

1,36

210

,158

0.77

36Se

rvic

esan

dci

vic

mat

ters

2,00

21,

268

0.12

301,

668

1,05

80.

0646

Shop

ping

,fam

ilybu

sine

ss1,

852

6,05

70.

5829

1,72

85,

750

0.66

24H

ouse

wor

k2,

735

11,8

692.

0016

3,15

913

,595

1.96

19Ph

one

calls

680

489

0.10

131,

581

1,14

20.

1227

Cul

ture

and

amus

emen

teve

nts

417

324

0.10

938

830

20.

108

Hob

bies

1,32

31,

203

0.47

71,

544

1,42

90.

616

Rea

ding

1,02

691

00.

873

982

870

0.71

3Sp

orts

and

recr

eati

on50

651

30.

592

571

583

0.69

2T

Vvi

ewin

g80

01,

135

1.74

280

21,

225

2.25

1Sl

eepi

ng0

08.

550

00

8.65

0Pa

idw

ork

00

2.93

00

02.

460

Una

lloca

ted

expe

ndit

ures

26,7

4550

,599

26,8

2847

,206

Uns

peci

fied

tim

e-us

e2.

982.

79T

otal

60,1

9713

0,17

124

60,5

3412

7,35

424

Not

e:Se

eth

ee-

supp

lem

ent

for

mor

ede

tails

.FIM

=Fi

nnis

hm

arks

;yr=

year

;MJ=

meg

ajou

les;

hr=

hour

;one

meg

ajou

le(M

J)=

106

joul

es(J

,SI)

≈23

9ki

loca

lori

es(k

cal)

≈94

8B

riti

shT

herm

alU

nits

(BT

U).



Figure 1 The average time use (hours/day/person) and average energy requirements (megajoules/day/person) of selected household activities in 1987–1990 and in 1998–2000. The arrows describe the directionof the change.

direction of change during the 1990s in Finland.The energy dimension reveals the familiar prior-ities of environmentally relevant consumption:housing, nutrition, and mobility require a signifi-cant share of total material inputs (Spangenbergand Lorek 2003). Housework, which includespreserving and preparing food, is also significant.The time use dimension reveals another aspect ofconsumption. In addition to working and sleep-ing, consumers fill their days by watching TV,doing housework, and having meals. Personal hy-giene, reading, sports, and shopping are the nextsignificant activities during the course of an av-erage day.

Despite the fact that the comparisons betweenactivities are sensitive to the allocation criteria,some observations are due. High-intensity activ-ities include, fairly obviously, commuting, leisuretravel, and eating. However, low-intensity ac-tivities are also of interest; TV viewing has thelowest intensity of all analyzed activities. Con-cerning figure 1, it can be noted that, in general,

time use on the low-intensity activities has in-creased the most. Most notably, and in accordwith the trends in other countries, the timespent on watching TV has increased dramati-cally. Activities that people spend less time oninclude reading, personal hygiene and dressingup, housework, services and civic matters, andleisure-related trips. For leisure trips, services andcivic matters, and reading, the average daily en-ergy requirements per person have declined aswell; although for services and civic matters, thechange is not proportional to changes in timeuse. On the other hand, housework, and shop-ping and family business-related activities are in-teresting exceptions; for the first, the energy re-quirements have increased despite the slightlydiminished time use, and for the latter, energyrequirements of its related mobility seem to havedeclined, whereas total time use has increased.In both of these cases the result is a substantialchange in the energy intensity of these activi-ties.



Tabl

e4

The

impa

ctof

hous

ehol

dty

peon

the

1990

tota

lexp

endi

ture

san

den

ergy

requ

irem

ents

ofth

eho

useh

olds

and

onth

esp

ecifi

cfig

ures

for

hous

ewor

kan

dfo

rvi

ewin

gTV

Hou

seho

ldty

pe

Tw

oT

wo

Oth

ers

One

Tw

oad

ults

+T

hree

adul

ts+

(Ave

rage

One

adul

t,1–

2T

wo

adul

ts+

aT

hree

two

adul

ts+

ath

ree

size:

5.25

adul

tch

ildre

nad

ults

child

adul

tsch

ildre

nch

ildch

ildre

npe

rson

s)

Tot

alen

ergy

requ

irem

ents

148,

017

222,

955

309,

346

362,

435

459,

277

430,

926

527,

773

482,

410

584,

543

[MJ/

yr/h

ouse

hold

]T

otal

ener

gyre

quir

emen

ts14

8,01

794

,875

154,

673

120,

812

153,

092

107,

731

131,

943

96,4

8211

1,34

1[M

J/yr

/per

son]

Hou

sew

ork

Expe

ndit

ures

2,66

65,

381

6,70

19,

138

7,70

69,

911

9,91

911

,336

11,3

45[F

IM/y

r/ho

useh

old]

Ener

gy[M

J/yr

/hou

seho

ld]

12,7

4326

,248

26,9

7836

,878

35,1

4343

,016

43,1

8151

,303

53,1

30T

ime

use

2.3

4.0

5.1

5.6

5.6

6.6

6.6

7.6

8.0

[hr/

day/

hous

ehol

d]In

tens

ity

[MJ/

hr]

14.9

18.0

14.5

17.9

17.1

17.8

17.9

18.6

18.2

Vie

win

gT

VEx

pend

itur

es1,

352

1,86

61,

844

2,22

02,

257

2,26

92,

460

2,71

62,

388

[FIM

/yr/

hous

ehol

d]En

ergy

[MJ/

yr/h

ouse

hold

]1,

712

2,66

02,

642

3,27

33,

331

3,36

13,

531

4,11

73,

890

Tim

eus

e1.

83.

83.

94.

84.

86.

66.

68.

49.

2[h

r/da

y/ho

useh

old]

Inte

nsit

y[M

J/hr

]2.

61.

91.

91.

91.

91.

41.

51.

31.

2

Not

e:M

J=m

egaj

oule

s;yr

=ye

ar;F

IM=

Finn

ish

mar

ks;h

r=ho

urs;

one

meg

ajou

le(M

J)=

106

joul

es(J

,Sl)

≈23

9ki

loca

lori

es(k

cal)

≈94

8B

riti

shT

herm

alU

nits

(BT

U).



Intensity, that is, energy use per unit of time,also varies across the different household typesunder consideration. Table 4 presents the inten-sities in 1990 per household type in two activi-ties. First, in housework, the intensities tend torise with the average size of the household. This ismainly due to the fact that time use in houseworkdoes not increase in proportion to the size of thehousehold or to the expenditures and the energyrequirements. On the other hand, in viewing TVthere is a reverse scale-related impact. The invest-ment in the equipment and the service fees areshared by an increased number of viewers in thelarger households. In a similar vein, the averagenumber of simultaneous viewers may be greaterin larger households. The intensity of viewingTV thus diminishes with the growing size of thehousehold.

Decomposition Analysis

The previous sections of this article have dis-cussed some factors that influence the energy re-quirements of households. Changing patterns oftime use are only one of these factors. In addi-tion, the intensity of each type of activity—theresources needed per unit of time—may changeas was shown in figure 1. Furthermore, it hasbeen shown that demographic changes in thehousehold types affect the energy requirements(Nurmela 1996).

The following analysis aims to isolate the rela-tive significance of the time-related factors with adecomposition analysis. The gross change in theenergy requirements of private final consumptionin Finland from the late 1980s to the late 1990shas been analyzed in terms of the contributionsof the following factors:

� Population growth (P)� Changes in relative share of different house-

hold types (H)� Changes in time use within the analyzed

time use categories (t)� Changes in the energy intensity of these

time use categories (I)� Changes in household infrastructure (f)

Figure 2 presents the results of a base-year de-composition in which the figures represent thecontribution of a factor assuming that all the

other factors remained constant or changed atthe very end of the period.4 As this is not thecase, the real contribution of the factors is some-thing different and their sum does not equal theobserved gross increase in materials requirements.The residual factor in figure 2 indicates the dif-ference between the gross changes and the sumof the contributions of the factors separately.

Demographic changes—population growthand decreasing average size of households—areimportant factors behind the trends in energy re-quirements. However, the changes in time useand in the energy intensity of activities are alsoof relevant magnitude and are the particularfocus of this article. Whereas the time use fac-tor describes what people do in terms of time use,the intensity factor describes how people do whatthey do. The changes in the intensity can thusbe regarded as technological changes in the com-position and volume of the inputs of a specificactivity.

The result achieved—according to which thetime use changes have contributed toward lowerenergy demands—matches figure 1; the activi-ties with low energy intensity have gained in theaverage time budget. Important contributions interms of energy use stem, however, both fromthe increased low-intensity activities such as TVviewing and from reduced mobility-related time.

The changes in the intensities of the activi-ties have contributed toward an increase in en-ergy requirements. The activity-specific changeswere presented in table 3, but it should be notedthat in the decomposition analysis, the inten-sity impacts depend naturally also on the initialtime use in the activity. The most important in-tensifying impact thus stems from housework. Inaddition, telephone conversations and the use ofservices have become more energy-intensive. Onthe other hand, the combined activity of trans-portation for shopping and the act of shoppinghas contributed to a lowering of its intensity fac-tor.

The significant negative impact of the house-hold infrastructure factor consists of many co-herent changes. In an average household, theoverall expenditure on these items has fallenslightly. More importantly, there is a structuralshift within the expenditure. Rents and the costsof owned homes have increased significantly,



Figure 2 The results of a base-year decomposition of the change in the gross energy requirements (TJ) ofFinnish private final consumption from 1987–1990 to 1998–2000. The bars indicate the contributions of thefactors assuming that all other factors remained constant or changed at the very end of the period. Oneterajoule (TJ) = 1012 joules (J, Sl) ≈ 9.48 × 108 British Thermal Units (BTU).

whereas those of energy have fallen.5 Also, theclothing expenditures have declined during theperiod from 1990 to 1998. On the other hand,as the impacts of changes in household sizehave been isolated into a separate factor, thesechanges, which in general raise the residentialenergy demand, do not show in the householdinfrastructure factor.

Concerning the changes in time use, it shouldalso be noted that the analyzed activities do notaccount for the whole 24 hours of one day. Hence,it is possible that the duration of one of the activ-ities changes without corresponding changes inother time use. In fact, the total time allocatedto the analyzed activities has grown from 18.1 to18.7 hr on average. This 0.6 hr has been reducedfrom the activities that are not specified in thetime use classification scheme and for which therelated energy requirements are categorized underthe household infrastructure factor. Thus, someof the negative contribution of the infrastructurefactor should rather appear as a greater negativetime use factor.

Conclusions

This article has proposed and demonstrateda method of using time use survey data to ex-tend the scope of the decomposition analyses,which have been used to study changes in theeconomy. This additional layer focuses on thepatterns of time use and the intensity of timeuse as explanatory factors for the changes in con-sumption expenditure and the related energy re-quirements. The time use approach was arguedfor from two distinctive points. First, I claimedthat a time use focus is better able to internalizeand contextualize economic growth and struc-tural changes as sources of rising energy require-ments. This is apparent in the applied decompo-sition factors, which do not include factors foreconomic growth and for the structure of produc-tion activities. Rather, the analysis forces one tosituate growth in the changes of everyday life andelucidates the conditions of structural changes inthe economy. Second, I argued that this frame ofanalysis implies the possibility of human agency.



The proposed frame of analysis assumes thatconsumption expenditures and consequent en-ergy requirements are dependent on patterns oftime use. In addition, it assumes that time usedecisions are social phenomena beyond merechanges in technology and marginal productiv-ity. These assumptions were argued for on thebasis of two theories on time use. Firstly, theo-ries on household production suggest that mar-ket goods are not direct utilities as such, but thatthey are rather inputs to specific household ac-tivities. These economic theories of time use alsosuggest that the time use and the intensity factorsare interdependent, as the technologies used inthe activity determine both the intensity and themarginal productivity of allocating time to theactivity. However, leaning more on sociologicalexplanations of time use, this article has assumedthat it is relevant to study the two factors as ifthey were mutually independent.

The results show that both the time use andintensity factors have contributed to changes inenergy requirements. During the 1990s, Finns en-gaged in less energy-intensive activities, but atthe same time the activities on average requiredincreasing energy inputs per unit of time. Someactivities, such as leisure-time travel, compliedwith the assumption that intensities remainedfairly stable. This result, however, is not clear, asthere also were activities, such as housework, inwhich the intensities changed significantly. Thesame ambiguity is also present in the results ofthe decomposition analysis. The time use factorhas not been overwhelmingly strong; it has notbeen insignificant, but has contributed less thanthe intensity factor.

The rising average energy intensities of activ-ities and the specific findings concerning house-work are in line with economic theories of timeallocation and the claims of time-squeeze or time-deepening as a factor behind increasing lev-els of consumption (Linder 1970, Ropke 1999,Binswanger 2001, Robinson and Godbey 1997).However, the change toward low-intensity activ-ities is new and somewhat surprising consideringthe simultaneous rise of the energy requirementsof private consumption.

The arbitrary nature of the allocation of ex-penditures and the related energy requirements tospecific activities should be kept in mind. These

allocation decisions affect the energy-intensityfigures obtained for each activity. However, asthe allocation rules have remained the same overthe analyzed period, they affect less the decom-position analysis. The result that energy intensi-ties have risen while patterns of time use havechanged toward low-intensity activities appearsrobust.

The results raise many questions for further re-search. An anthropological focus on the subjec-tive meanings of material possessions might beable to answer whether the time use categoriesthat were used in this work correspond to themeanings consumers assign to their possessionsand activities. One might also question the allo-cation of expenses to the activities. The answerto this question would greatly benefit from a sin-gle set of data on both time use and expenditure,which, however, is scarcely available. With suchdata it would be possible to study the correlationbetween time use and expenditures on the level ofindividual households. The correlation betweentime use changes and the environmentally sig-nificant household infrastructure is especially inneed of elaboration.

The role of public consumption should alsobe considered. Apart from the true public goodssuch as defense, a large share of the public sectorserves the direct needs of individuals. Educationand health expenditures could be matched withitems of specific time use. Similarly the road in-frastructure, which is used by private consumers,constitutes a consumption item that could be al-located to private consumption.

Finally, the underlying causes of changes intime use and in energy intensity of activities callfor further research. It can be asked, for example,how voluntary was the change toward less energy-intensive activities in Finland during the 1990s.During this period, the level of unemploymentrose significantly and despite the fact that theaverage GDP per capita also rose, it may be thatless energy-intensive lifestyles were forced uponthe less fortunate.

The changes in the intensity factor also callfor further investigation. What are, for exam-ple, the conditions of the rising energy demandsand lower time use demands of housework? Thepenetration rates of household appliances haveincreased during the period, but this may not



explain all. Households are also increasingly sub-stituting commercial services for their own labor.Thus, even if commercial household servicesare often regarded as environmentally benefi-cial, ecoefficient, and labor-intensive sectors ofthe economy (Heiskanen and Jalas 2003), theyalso may intensify private consumption and con-tribute to a net increase in the gross demand fornatural resources.

Notes

1. One megajoule (MJ) = 106 joules (J, SI) ≈ 239kilocalories (kcal) ≈ 948 British Thermal Units(BTU).

2. For the rational agents of home economics, time useis just a reflection of the available technology. Newtechnology that increases labor productivity withina certain activity changes the allocation of time.For example, the central tenet of Linder (1970)is that consumer goods will be used to save timeand that those activities in which the productiv-ity can be enhanced will gain on the time budget.The intensity factor, which, in the work presentedin this article, measures energy inputs per unit oftime would thus cause and explain changes in timeuse.

Analyzing time use as a separate factor is a radicaldeparture from economic orthodoxy. This assump-tion implies that time is not considered simply asa resource for achieving instrumental goals or gain-ing utility. Patterns of time use are thus not mereresults of individual reasoning and optimization ofpersonal utility and, in the same vein, people arenot always able to explicate the reasons for their ac-tions or the “functions” they strive to fulfill. Rather,the patterns of everyday life result from conformingto norms, playing and acting life, and approachingliminal experiences, which all mix together witha limited capability for rationalizing present actionagainst future prospects. Another phenomenon ofeveryday life is that the instrumental outcomes un-fold in such different dimensions that the reasoningof actions becomes complex. If there is no “pro-duction function” for the activities of free time,the causal connection between technology, or in-tensity, and time use dissolves. Hence, economistshave also considered joint production, affective ex-periences, and “process benefits” (see Pollack andWachter 1975; Juster et al. 1981; Dow and Juster1985; Winston 1982; Gershuny and Haplin 1996).

In reality, people do reason their action. Con-sumers are, for example, price sensitive. But this

is not the main story. It hard to conceive of thereasons for the increasing time spent on watchingTV (e.g., Robinson and Godbey, 1997). Likewise,Csikszentmihalyi (2000) has found out that peopleare bored and do not know how to use their time.The former story of rational agents is the one we aremost keen on, but the latter also reflects reality andcalls for different approaches to increasing levels ofconsumption.

3. The sectoral energy intensities [MJ/FIM] used inthe current study have been calculated by IlmoMaenpaa, who has also used the same input-output data within his work on materials intensity(Maenpaa and Juutilainen 2001). Due to space lim-its these data are presented in the e-supplement ofthis article.

4. Decomposition analysis of empirical discrete obser-vations encounters the so-called integral path prob-lem: the path of change of each parameter betweenthe observations is not known and the analysis is leftwith a residual term that cannot be allocated to anyparameter. Various methods have been proposedto allocate the residual term, but lacking knowl-edge of the process of change, these methods arearbitrary (de Bruyn 2000). The base-year method,which is used in the current work, does not allocatethe residual factor.

5. The degree-days in the calendar years 1989–1991were less than those of 1997–1999, which indicatesthat the fall in energy expenditure from 1990 to1998 was not due to temperature variations.

References

Andrews, C. 2000. Building a micro-foundation for in-dustrial ecology. Journal of Industrial Ecology 4(3):35–52.

Aronowitz, S. and W. DeFazio. 1994. The jobless future.Minneapolis: University of Minnesota Press.

Becker, G. 1965. A theory of the allocation of time.The Economic Journal 65(Sept.): 493–517.

Belk, R. 1988. Possessions and the extended self. Journalof Consumer Research 15: 139–168.

Binswanger, M. 2001. Technological progress and sus-tainable development: What about the reboundeffect? Ecological Economics 36(1): 119–132.

Brodersen, S. 1990. Reanalysis of consumer surveys.Classification and method. In Time and consump-tion, edited by G. Viby Mogensen. Copenhagen:Danmarks Statistiks.

Chadeau, A. and C. Roy. 1986. Relating households’final consumption to household activities: Sub-stitutability or complementarity between market



and non-market production. Review of Income andWealth 32(4): 387–407.

Chertow, M. R. 2000. The IPAT equation and itsvariants. Journal of Industrial Ecology 4(4): 13–29.

Coser, L. and R. Coser. 1963. Time perspectiveand social structure. Reprinted in The sociologyof time, edited by J. Hassard (1990). London:MacMillan.

Cross, G. 1993. Time and money. The making of theconsumer culture. London: Routledge.

Csikszentmihalyi, M. 2000. The costs and benefits ofconsumption. Journal of Consumer Research 27(2):267–272.

De Bruyn, S. 2000. Economic growth and the environment.Dordrecht: Kluwer Academic Publishers.

Dietmar, H. 1992. The social psychology of material pos-sessions. To have is to be. Hemel Hempstead, UK:Harvester Wheatsleaf.

Douglas, M. and Isherwood, B. 1979. The world of goods.Harmondsworth, UK: Penguin.

Dow, G. K. and F. T. Juster. 1985. Goods, time, andwell-being: The joint dependence problem. InTime, goods, and well-being, edited by F. T. Justerand F. P. Stafford. Ann Arbor, MI: The Univer-sity of Michigan.

Elchardus, M. 1991. Rationality and specialization ofmeaning. A sociological approach to the alloca-tion of time. In The consumption of time and thetiming of consumption. Proceedings of the Interna-tional Colloquium Amsterdam, 6–8 November 1990,edited by G. Antonides, W. Arts, and W. F. vanRaaij. Amsterdam: Royal Netherlands Academyof Arts and Sciences.

Eurostat. 1999a. Survey on time use. Activ-ity coding list. Eurostat DOC E2/TUS/3.6/99.Helsinki: Statistics Finland. Also availableat <www.iser.essex.ac.uk/conferences/iatur/1999/pdf/abstract31/word7.pdf>. Accessed 14 February2005.

Eurostat. 1999b. Proposal for a satellite account ofhousehold production. Final report of the ProjectSC96LO9 Time Use Survey: Development of aEuropean Satellite system of household produc-tion. Luxembourg: Eurostat.

Farla, J. C. M. and K. Blok. 2000. Energy efficiency andstructural change in the Netherlands, 1980–1995.Journal of Industrial Ecology 4(1): 93–117.

Firat, A. F. and N. Dholakia. 1998. Consuming people.From political economy to theaters of consumption.London: Routledge.

Gershuny, J. 1987. Time use and the dynamics of theservice sector. The Service Industry Journal 7(4):56–72.

Gershuny, J. and B. Haplin. 1996. Time use, qualityof life, and process benefits. In In pursuit of thequality of life, edited by A. Offer. Oxford, UK:Oxford University Press.

Godbey, G. 1996. No time to waste: Time use and thegeneration of residential solid waste. Yale Work-ing Papers on Solid Waste Policy #4. New Haven,CT: Yale School of Forestry and EnvironmentalStudies.

Heiskanen, E. and M. Jalas. 2003. Can services leadto radical eco-efficiency improvements? A reviewof the debate and evidence. Corporate Social Re-sponsibility and Environmental Management 10(4):186–198.

Hoffren, J., J. Luukkanen, and J. Kaivo-oja. 2000. De-composition analysis of Finnish material flows:1960–1996. Journal of Industrial Ecology 4(4): 105–126.

Jackson, T. and R. Clift. 1998. Where is the profitin industrial ecology? Journal of Industrial Ecology2(1): 3–5.

Jalas, M. 2002. A time use perspective on the materi-als intensity of consumption. Ecological Economics41(1): 109–123.

Juster, F. T., P. N. Courant, and G. K. Dow. 1981.The theory and measurement of well-being: Asuggested framework for accounting and analysis.In Social accounting systems: Essays on the state ofthe art, edited by F. T. Juster and K. C. Land. NewYork: Academic Press.

Linder, S. 1970. The harried leisure class. New York andLondon: Columbia University Press.

Maenpaa, I. and A. Juutilainen. 2001. Material flows inFinland. Resource use in a small open economy.Journal of Industrial Ecology 5(3): 33–48.

Niemi, I. and H. Paakkonen. 2001. Ajankayton muutok-set 1990-luvulla. [Changes in time use in the 1990s.].Statistics Finland, Culture and the Media, 2001:6.Helsinki: Statistics Finland.

Nurmela, J. 1996. Kotitaloudet ja energia vuonna 2015.[Households and energy in the year 2015.]. Statis-tics Finland, Research Report 216. Helsinki:Edita.

Pollack, R. A. and M. L. Wachter. 1975. The rele-vance of the household production function andits implications for the allocation of time. Journalof Political Economy 83(2): 255–277.

Reisch, L. A. 2001. Time and wealth. The role oftime and temporalities for sustainable patternsof consumption. Time & Society 10(2/3): 367–385.

Robinson, J. and G. Godbey. 1997. Time for life. Thesurprising ways Americans use their time. UniversityPark, PA: Pennsylvania State University Press.



Røpke, I. 1999. The dynamics of the willingness toconsume. Ecological Economics 28(3): 399–420.

Sanne, C. 2000. Dealing with environmental savingsin a dynamic economy—how to stop chasing yourtail in the pursuit of sustainability. Energy Policy28(6–7): 487–497.

Schipper, L., S. Bartlett, D. Hawk, and E. Vine. 1989.Linking life-style and energy use: A matter oftime. Annual Review of Energy 14(Nov.): 273–320.

Schor, J. 1991. The Overworked American. New York:Basic Books.

Sorokin, P. and R. Merton. 1990. Social-time: Amethodological and functional analysis. 1937Reprinted in The sociology of time, edited by JohnHassard. London: MacMillan.

Spangenberger, J. H. and S. Lorek. 2003. Envi-ronmentally sustainable household consumption:from aggregate environmental pressures to prior-ity fields of action. Ecological Economics 43(2–3):127–140.

Statistics Finland. 2000. Kotitalouksien kulutusmenojenmuutokset 1990-luvulla. [Changes in household con-sumption expenditure in the 1990s.] Tulot ja kulutus2000:28. Helsinki: Yliopistopaino.

Thompson, E. P. 1974. Time, work-discipline, and in-dustrial capitalism. 1974. Reprinted in Essays insocial history, edited by M. W. Flinn and T. C.Smout. Oxford, UK: Clarendon Press.

Winston, G. C. 1982. The timing of economic activities.Cambridge, UK: Cambridge University Press.

About the Author

Mikko Jalas is a research fellow in organiza-tions and management at the Helsinki School ofEconomics in Helsinki, Finland.

Appendix

The decomposition of the energy require-ments of private final consumption can be pre-sented in the form of the equation

Etot = P ×o∑

k=1

Hk

×(

m∑i =1

n∑j =1

tk,i × ck,i, j I j /tk,i + Ek f

)

in which

Etot represents the total energy requirements ofprivate final consumption

P represents the populationHk represents the share of the population liv-

ing in household type k = 1, . . . , otk,i represents time use on activity i =

1, . . . , m in household type kck,i, j represents the costs j = 1, . . . , n related to

the activity i in household type kI j represents the energy intensity of expendi-

ture jEk f represents the energy requirements of

household infrastructure and other suchcosts that have not been allocated to spec-ified time use in household type k.

The change in the total energy requirements canbe decomposed into the following factors:

�E = �EP + �EH + �Et + �EI + �E f + R

The decomposition equations of each factor are

�EP = �P × Etot

�EH = P ×o∑

k=1

�Hk

×(

m∑i =1

n∑j =1

tk,i × ck,i, j I j /tk,i + Ek f

)

�Et = P ×o∑

k=1

Hk

×(

m∑i =1

n∑j =1

�tk,i × ck,i, j I j /tk,i

)

�EI = P ×o∑

k=1

Hk

×(

m∑i =1

n∑j =1

tk,i × �(ck,i, j I j /tk,i )

)

�E f = P ×o∑

k=1

Hk × �Ek f

R = �E − (�EP + �EH + �Et + �EI

+�E f )

in which all constant figures refer to the base year.Change is interpreted in relation to the base year.For example, �P = (P98 − P90)/P90.


The Everyday Life Context of Increasing Energy Demands: Time Use Survey Data in a Decomposition...

Documents

Transcript of The Everyday Life Context of Increasing Energy Demands: Time Use Survey Data in a Decomposition...