Journal of Economic Psychology 20 (1999) 53±81

Spillover processes in the development of a sustainable

consumption pattern
1
John Thùgersen
Department of Marketing, Aarhus School of Business, Haslegaardsvej 10, DK-8210, Aarhus V, Denmark
Received 23 October 1996; received in revised form 3 May 1998; accepted 30 November 1998

Abstract

Hypotheses about possible mechanisms for spillover processes between pro-environmental

behaviours are developed and tested by means of structural equation modelling. Data were
collected by means of telephone interviews with a representative sample of Danish adults.
Personal norms concerning recycling and packaging waste prevention are found to be rooted
in the same more general, internalised values. Further, a predicted positive spillover e€ect from
recycling to packaging waste prevention is con®rmed. However, whereas a positive spillover
e€ect from recycling to personal norms concerning packaging waste prevention is predicted,
the reverse is found. Neither is it con®rmed that performing an environmentally friendly
behaviour makes attitudinal and more distal antecedents of related behaviours more predictive
of the next step in the assumed causal chain. Ó 1999 Elsevier Science B.V. All rights re-
served.

PsycINFO classi®cation: 3920; 4070

JEL classi®cation: D10

Keywords: Consumption patterns; Recycling; Waste prevention; Values; Norms; Structural

equation modeling

1
Tel.: +45 89 48 66 88; fax: +45 86 15 39 88; e-mail: john.thogersen@mar.hha.dk

0167-4870/99/$ ± see front matter Ó 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 7 - 4 8 7 0 ( 9 8 ) 0 0 0 4 3 - 9
54 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

1. Introduction

During the past few decades, the understanding of causes and of how to
avoid negative e€ects of consumer behaviour on the natural environment has
become an important, albeit still somewhat exotic research area within eco-
nomic psychology. With few exceptions these studies have focused on speci®c
behaviours, such as participation in a recycling programme (e.g., Pieters,
1991; Taylor & Todd, 1995; Thùgersen, 1994), buying organic food (e.g.,
Grunert & Juhl, 1995; Thùgersen & Andersen, 1996), avoiding excessive
packaging (e.g., Bech-Larsen, 1996; Thùgersen, 1996a), regulating home
heating (e.g., Van Raaij & Verhallen, 1983; Yates & Aronson, 1983), or
choice of mode of transportation (e.g., GaÈrling & Sandberg, in press; Ver-
planken, Aarts, Knippenberg & Knippenberg, 1994). Studies like these have
improved our understanding of the determinants of environmentally sensitive
consumer behaviour and of possible measures for changing environmentally
undesirable behaviours. However, the ``piecemeal'' approaches dominating
consumer behaviour research in this area fall short of the challenges currently
facing environmental policymakers (Gray, 1985; O È lander & Thùgersen,
1995). For example, one of the prominent conclusions of the Earth Summit
in Rio in 1992 was that ``altering consumption patterns is one of humanity's
greatest challenges in the quest for environmentally sound and sustainable
development'' (Sitarz, 1994, p. 39). Since the Rio summit, strategies for the
attainment of a sustainable consumption pattern has been on the agenda of a
large number of international, national, and NGO conferences and meetings
(e.g. Norwegian Ministry of Environment, 1994; Oslo Roundtable Con-
ference on Sustainable Production and Consumption, 1995; Stù, 1995; The
Directorate Industry-, Construction-, Product-, and Consumer-a€airs et al.,
1995). Increasingly, consumer-directed environmental policy is guided by the
ambition to change overall consumption patterns, not just speci®c, limited
behaviours.
But does it make a di€erence whether environmental policies target the
overall consumption pattern or speci®c, and probably urgent, problems?
After all, a consumption pattern is composed of speci®c behaviours. Some
authors have warned that the previous narrow focus on single behaviours
and problems is inecient and too slow, measured by the challenges ahead
of us (e.g., Gray, 1985; O È lander & Thùgersen, 1995), and others have
claimed that a too narrow focus often leads to the creation or worsening of
environmental problems outside the focus (e.g., Jùrgensen, 1989). Also, the
ethical defensibility of focusing on achieving speci®c behavioural change
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 55

goals when, in reality, there may be several other, equally valid ways to
de®ne and solve the problem has been questioned (O È lander & Thùgersen,
1995).
Maybe the most obvious alternative to a narrow focus on speci®c be-
haviour changes is to use educational and other means to target fundamental
beliefs, values and norms with implications for a broad range of behaviours
(Berger, 1991; Gray, 1985; Stern, 1992). However, whereas education un-
doubtedly is an indispensable element of a long-run strategy for a sustainable
society, previous educational experiments have shown meagre results with
regard to changing environmental behaviour (Gardner & Stern, 1996; Gray,
1985). In addition, most empirical research has found no or a weak corre-
lation between general environmental attitudes or values and any speci®c
behaviour (Alwitt & Pitts, 1996; Moisander & Uusitalo, 1994; Stern &
Oskamp, 1987; Weigel, 1983).

As emphasised by Olander and Thùgersen (1995) the problem is probably
not whether to choose a targeted behavioural change tactics or a broad ed-
ucational strategy, but how to combine the two approaches to best advan-

tage. Olander and Thùgersen (1995) suggest policy should be focussed on
activation and behaviour change, but instead of trying to install a speci®c and
well-de®ned new behaviour ± which may produce reactance in some and
which others may ®nd is prohibitively inconvenient ± the goal should be to
produce some relevant behaviour change and to stimulate activity around
serious environmental problems in general (see also Devine & Hirt, 1989).
Education is an important means to produce quali®ed activity, but so are,
e.g., speci®c information about activities and opportunities for acting locally
and the provision of eventual physical prerequisites for action (for example
systems for separate collection of recyclable waste or public transportation
systems).
The idea of broadening the scope of behavioural change campaigns in the
environmental ®eld undoubtedly will be met with some resistance in both the
administrative and research communities. It may require more complicated
research designs which may be more dicult to cut into standard, journal
length pieces. Environmental administrators may resist broadening the scope
because it is likely to lead to the formulation of change goals that cross
administrative borders, are more demanding of money, time and personal
skills, and are more dicult to monitor and control. They may also fear that
widening the scope of desirable behaviours may result in people getting
confused and ®nding it easier to avoid doing anything at all which could
actually lead to less behavioural change. The fear that widening the
56 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

perspective will confuse most people is nurtured by empirical research re-

porting that ``environmentally bene®cial choices... are not re¯ective of a
general conservation stance, but are instead made on an activity-to-activity
basis'' (Pickett, Kangun & Grove, 1993, p. 240). Even within a certain en-
vironmental sector actions have often been found not to be closely related
(Stern & Oskamp, 1987).
There are, however, also strong arguments supporting the idea of broad-
ening the scope of behavioural change campaigns. Maybe the strongest ar-
gument, suggested by researchers from a broad range of behavioural sciences,
is that a change in attitude and/or behaviour concerning a speci®c activity,
produced by a targeted e€ort or otherwise, may ``spill over'' into related areas
and, hence, become more general (Frey, 1993; Jensen, 1992; Sugden, 1989;
Weigel, 1983; Williamson, 1975). If such ``spillover'' processes are common,
clearly the focus of behavioural change policy in the environmental ®eld
should be broad enough to manage and monitor not only the direct e€ects of
the policy on the target behaviour, but also its (positive and negative) spill-
over e€ects on other behaviours. However, until now only anecdotal evidence
supports the existence of spillover processes. About a decade ago, one re-
viewer of the empirical literature concluded that ``there is very little evidence
of transfer of heightened pro-ecological behaviour from the target behaviour
to other behaviors'' (Gray, 1985, p. 188).
The empirical research referred to by Gray (1985) was small in volume and
generally inadequate for detecting ``transfer of heightened pro-ecological
behaviour''. Especially, there is a lack of theorising about how the transfer
(or spillover) is supposed to happen. Since the performance of practically any
environmentally bene®cial behaviour is bound to be in¯uenced by a host of
other, both motivational and situational, factors besides an eventual spillover
e€ect (e.g., Guagnano, Stern & Dietz, 1995; Weigel, 1983), the risk of a Type
I error (rejecting the spillover hypothesis when in fact it is true) is high in the
absence of clear hypotheses about how spillover e€ects expresses themselves.
Hence, more and better empirical research is needed in order to test the
existence and e€ectiveness of transfer or spillover processes between envi-
ronmentally bene®cial behaviours.
This paper develops and tests some hypotheses about possible mechanisms
for spillover processes between pro-environmental behaviours. It is a basic
assumption that the performance of an environmentally friendly activity (or
any other involving activity) leads to heightened psychic activity (salience,
learning, rationalisation) in the person which may have implications for how
other activities are conceived and/or performed.
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 57

2. Hypotheses

2.1. Hypothesis 1

A large literature supports Fazio's hypothesis that attitudes based on direct

experience are both more salient and more predictive of behaviour than at-
titudes that are not (e.g., Doll & Ajzen, 1992; Fazio & Zanna, 1981; Regan &
Fazio, 1977; for a comprehensive review, see Eagly & Chaiken, 1993). Most of
this literature has investigated the hypothesis that experience with a speci®c
behaviour moderates the relationship between that behaviour and the attitude
towards it. An exception is Berger and Kanetkar (1995) who successfully test
the hypothesis that experience with a speci®c behaviour (recycling at the
workplace) can increase the strength of the relationship between other, but
related attitudes and behaviours. Cognitive psychology explains this type of
attitudinal spillover by the notion of spreading activation (e.g., Anderson,
1983; Collins & Loftus, 1975). When experience has made a speci®c attitude
more accessible in memory (Fazio, 1986), other mental constructs that the
person associates strongly with this attitude are made more accessible too,
through an unconscious spreading activation process. The strength of the
spreading activation depends on the semantic or conceptual proximity of the
two mental constructs. (For an empirical illustration and test within the en-
vironmental ®eld, see Cialdini, Reno & Kallgren, 1990, Study 5.) Hence, we
may hypothesise that, due to spreading activation processes,

Hypothesis 1 (H1). The performance of an environmentally friendly

behaviour (e.g., the participation in a recycling programme for household
waste) increases the salience of attitudes towards other consumer activities
targeting the same environmental problem (e.g., attitudes towards avoiding
excessive packaging or reusing packaging for other purposes in the
household) and, hence, strengthen the correlation between these speci®c
attitudes and behaviour.

2.2. Hypothesis 2

Hypothesis 1 makes no reference to eventual changes in the person's beliefs

or value structures. However, it seems fairly obvious that learning processes
and subsequent changes in speci®c beliefs regarding a behaviour can be ini-
tiated by the performance of another behaviour, thus providing a second
route that spillover processes could take.
58 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

Popular cognitively based attitude and norm theories like, for example,
Fishbein and Ajzen's (Fishbein & Ajzen, 1975; Ajzen & Fishbein, 1980)
Theory of Reasoned Action and Schwartz's (Schwartz, 1977) Norm Acti-
vation Theory assume that the proximal attitudinal and/or normative ante-
cedents of a behaviour are rooted in beliefs about consequences of that
behaviour. It is also acknowledged within this tradition that ``a change in
behavior implies a new set of relationships between the individual and the
attitude object


[which] may then lead to the learning of new beliefs about
the attitude object, and thus to attitude change'' (Fishbein, 1967, p. 482).
Researchers studying environmentally sensitive behaviour have often in-
cluded this type of feedback in their models (Ester & Mandemaker, 1994;
Stern, 1992; Thùgersen, 1994; Van Raaij & Verhallen, 1983), and it is well
documented in empirical research as well (e.g., O È lander & Thùgersen, 1995;
Thùgersen, 1997, 1998b). For example, it has been found that beliefs about
the consequences of recycling change substantially with experience, cf. Jo-

hansson (1993) described in Olander and Thùgersen (1995). Most of the
learning that goes on while performing an environmentally friendly activity,
such as recycling, is undoubtedly relevant only for that particular activity.
However, it is not unlikely that some of it has implications for ± and may lead
to revisions of attitudes towards ± other behaviours as well; in the case of
recycling for example learning related to the goals of the recycling pro-
gramme or about the ``waste consequences'' of various consumer goods or
procedures. Hence, we may hypothesise as follows.

Hypothesis 2 (H2). The performance of an environmentally friendly

behaviour (for example, recycling) may lead to learning about environmental
consequences of other consumer behaviours which may then produce
attitude changes regarding these behaviours and, subsequently, behaviour
changes.

2.3. Hypothesis 3

It is widely acknowledged in psychology and consumer behaviour research

that the understanding (if not the prediction) of behaviour can be improved
by studying more distal mental antecedents (general values and attitudes) in
addition to the proximal determinants included in the most popular models
(speci®c intentions, attitudes and/or norms) (e.g., Eagly & Chaiken, 1993;
Homer & Kahle, 1988; Rokeach, 1968, 1973). Some researchers, among them
some that have focused particularly on explaining environmentally friendly
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 59

behaviour, also include a hypothesis about feedback from the performance of

a speci®c behaviour to the behaviour's more distal and abstract antecedents
in their model (e.g., Stern, 1992). In these cases, it is assumed that the vol-
untary performance of a behaviour triggers mental self-justi®cation processes
and that this heightened mental activity makes abstract values, attitudes and
norms that may serve as justi®cation for the behaviour more salient (Homer
& Kahle, 1988; Stern, 1992). If the individual becomes aware of inconsis-
tencies among important values, changes in the value system may follow too
(Rokeach, 1973, 1979). For example, (provided it is perceived to be volun-
tary) participation in a recycling programme may make a person's general
environmental values and attitudes more salient. The increased salience of
these mental constructs may, next, sensitise the person to environmentally
harmful consequences of other behaviours and subsequently lead to changes
in these behaviours (e.g., Stern, Dietz, Kalof & Guagnano, 1995; Thùgersen
& Grunert-Beckmann, 1997). In this indirect way, the initial behaviour
change (beginning to recycle) may spill over into other behaviours. Hence, we
may hypothesise as follows.

Hypothesis 3 (H3). The (voluntary) performance of an environmentally

friendly behaviour increases the salience of the person's general environ-
mental values and attitudes which, in a second step, makes the person change
other speci®c behaviours with harmful environmental consequences.

2.4. Hypothesis 4

The possible spillover processes described until now all involve some
reasoning. The individual is assumed to establish conscious and motivating
association links between speci®c or more abstract evaluations and some
speci®c behaviour(s); the evaluation having been altered or made more sa-
lient by performing an environmentally friendly behaviour, such as partici-
pating in a recycling programme. However, it is also possible to envisage
more spontaneous spillover processes where related behaviours, gradually
and unconsciously, become aligned according to a principle or goal. Such
processes could, for instance, evolve because individuals have a preference
for category-based evaluation and, at the same time, a propensity to perceive
their activities in as broad categories as possible.
Several popular models in consumer behaviour research assume that, when
encountering an attitude object (external stimuli), an individual's ®rst reaction
60 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

is to attempt to categorise it (e.g., Fazio, 1986; Peter & Olson, 1996). If the
categorisation is successful, a€ect associated with the category label may be
activated spontaneously and guide behaviour (Fiske & Pavelchak, 1986;
Sujan, 1985). Fiske and Pavelchak (1986) suggest that only if the encountered
object hold salient characteristics that are inconsistent with the category, or,
if no appropriate category can be retrieved from memory, will the person
engage in more time-consuming and demanding ``piecemeal-based'' evalua-
tion of the attitude object. Category-based evaluation has been shown to be
both preferred to and more ecient (in terms of response latency) than
piecemeal-based evaluation both regarding other people (e.g., Fiske & Pa-
velchak, 1986) and consumer goods (e.g., Sujan, 1985), and there is no ob-
vious reason why it should be di€erent in other areas (Fiske & Pavelchak,
1986).
Add to this Vallacher and Wegner's (Vallacher & Wegner, 1985; Wegner &
Vallacher, 1986) suggestion that we are inclined to conceive of our activities
in as encompassing categories as possible, because it allows a more ecient
execution of most acts. Only when we cannot perform a broadly conceptu-
alised act, do we concern ourselves with thinking of details of the act (i.e.,
activate more narrow action categories). This is typically the case when a
person begins performing a somewhat dicult new activity. In such cases, he
or she tends to focus at the operations that have to be borne out. However,
with experience the speci®c operations become routinised, thus making a
change in focus away from the speci®c operations and towards the goals of
the activity possible (and desirable).
Hence, due to routinisation, individuals tend to conceive of repeated acts
in broader categories over time. A participant in a recycling programme, for
example, may in the beginning conceive of this activity as separating the
waste in designated containers, later as recycling, and even later as waste
avoidance. Imagine a recycler who perceives her participation in a source
separation programme as an instance of waste avoidance, and whose attitude
towards waste avoidance (a broad behavioural category) is based on expe-
rience with participation in the source separation programme only. Imagine
also that when buying, say, detergents this person switches between a limited
number of brands that overall are evaluated equally positive. If, suddenly,
one of these brands comes in a new ``re®ll'' packaging with a label and a text
explaining that the new packaging helps avoiding waste, the ``waste avoid-
ance'' category may be activated in the recycler's mind and the a€ect asso-
ciated with it determine her brand choice.
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 61

A€ective responses are more likely to be category-based if the stimuli is

complex, if the individual is under time pressure or not knowledgeable about
the category, or if there is low risk (e.g., of retaliation from a target person)
(Fiske & Pavelchak, 1986; Sujan, 1985). In the environmental ®eld it is clearly
often the case that the issue is complex and the average individual not very
knowledgeable about the issue (e.g., Ellen, 1994; Morris, Hastak & Mazis,
1995; Stern, 1992). As the re®ll packaging example illustrates, time pressure
and low risk are also common characteristics of environmentally sensitive
behaviours.
Based on the research on category-based evaluation and action identi®-
cation we may hypothesise as follows.

Hypothesis 4 (H4). Performing an environmentally friendly activity (for

example, recycling) increases the likelihood that other activities, that can be
categorised in a more general mental concept that also includes this activity,
are performed as well.

3. Method and data

3.1. The case

The speci®c case investigated in the following is the eventual transfer of

pro-environmental behaviour between waste handling (recycling) and shop-
ping (waste prevention by ``packaging conscious'' buying). This case was
chosen because of its high practical relevance. In the last ®ve to seven years,
most industrialised countries have witnessed substantial and widespread
behaviour changes with regard to the disposal of waste in response to a fairly
intense promotion of recycling by central and local authorities. At the same
time it is widely acknowledged that recycling is neither the only nor the best
solution to the waste problem (e.g., Eurostat et al., 1995; Gardner & Stern,
1996) and the idea of preventing waste from arising in the ®rst place is getting
increasing prominence in many countries (e.g., Commission of the European
Communities, 1996; Miljù-og Energiministeriet, 1995; Rathje & Murphy,
1992; Wuttke, 1993). Hence, if a person's participation in a recycling pro-
gramme increases the likelihood that he or she will also change other be-
haviours in order to reduce the ``wastefulness'' of consumption (Thùgersen,
1996b), this would be a substantial side-bene®t. As demonstrated by research
cited in the introduction, it is not self-evident that positive spillover of this
62 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

kind will arise. In fact, it is sometimes claimed that participation in a recy-

cling programme may result in negative spillover because it provides con-
sumers with a convenient excuse for continuing an otherwise environmentally
harmful lifestyle (e.g., Wenke, 1993).
The hypotheses developed above were applied to this case and tested on
cross-sectional survey data by means of structural equation modelling and
LISREL8 for Windows (J oreskog & Sorbom, 1993). Structural equation
modelling was chosen because of its strong features regarding the modelling
of complex relationships. However, the reliance on correlational cross-sec-
tion data of course severely limits the possibilities for making causal infer-
ences.

3.2. The data

In summer 1995, a telephone interview survey was conducted with a

sample of Danish residents aged 18 years and older (N ˆ 1002). Households
were drawn randomly from a telephone register covering the whole country,
and respondents within households were selected using the ``next birthday''
method. The appropriate individual completed a survey in 53.8% of the
households reached. Data were collected concerning a number of attitudinal
constructs and behaviour with regard to recycling and avoiding packaging
waste (and a number of other issues not discussed in this paper). Separate
analyses not reported here (but see Thùgersen, 1998a) show that the per-
sonal norm is the best predictor of behaviour among the attitudinal con-
structs and that the in¯uence of the degree of overall favourability and the
subjective social norm (as conventionally measured) is reduced to insignif-
icance (recycling) or close to insigni®cance (avoiding packaging waste) when
the personal norm is included in the structural equation. Hence, for the
present purpose the attitudinal antecedents of behaviour are represented by
the personal norm. In addition, one calculation includes some items from
Schwartz's Value Indicator (Schwartz, 1992, 1994; Schwartz & Bilsky, 1987,
1990).
Schwartz (1977, p. 227) de®nes a personal norm as ``the self-expecta-
tions for speci®c action in particular situations that are constructed by the
individual''. In continuation he states that ``activated personal norms are
experienced as feelings of moral obligation''. Here, the concept is opera-
tionalised as the feeling of personal obligation to (a) source separate glass
and bottles for recycling and appropriate kitchen waste for composting
(two items), and (b) to avoid groceries with excessive packaging, and to
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 63

buy beverages in re®llable bottles (two items), measured on a 4-point

unipolar scale with the end points ``no obligation'' and ``very strong ob-
ligation''.
The behavioural criteria are two relatively narrow behavioural categories
that both involve a limited set of recurrent actions with a well-de®ned target
and context (see Ajzen & Fishbein, 1980). Recycling of household waste is
operationalised as the frequency of source separating (a) glass and bottles
and (b) appropriate kitchen waste, and avoiding packaging waste is opera-
tionalised as the frequency of (a) avoiding groceries with excessive packaging
and (b) buying beverages in re®llable bottles, in all cases measured as the
number of times out of the last ®ve opportunities. Since it was judged that
these behaviour questions are formative rather than re¯ective (Bagozzi,
1994), a recycling index and a packaging waste prevention index were formed
by summing the responses on the individual items.

3.3. Data preparation

The quality of the data was assessed and the data was prepared for the
analysis by means of PRELIS (SPSS, 1993). The pre-analysis showed that
several of the measures contained substantial skewness and kurtosis that
could not be corrected with a simple transformation of the data. Hence,
Generally Weighted Least Squares (which is not dependent on assumptions
about the distribution of the data) was used to estimate the models.
Listwise deletion was used in the case of missing values. In order to reduce
the detrimental e€ect of this method on the sample size, it was assumed that
no answer to a behaviour frequency question is equal to a behaviour fre-
quency of 0. Still, due to missing values the sample size is reduced to 794
respondents.

4. Hypothesis tests

4.1. Hypothesis 1

Speci®ed to the present case, Hypothesis 1 states that the participation in a

recycling programme for household waste makes attitudes about avoiding
waste in connection with other consumer activities more salient (for example,
attitudes about avoiding excessive packaging or reusing packaging for other
purposes in the household) and, hence, strengthen the correlation between
64 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

these speci®c attitudes and behaviour. Independent measures of the salience

of the various concepts are not included in the data set so the in¯uence of
participation in a recycling programme on salience can only be detected in an
indirect way, through analysing the assumed in¯uence of recycling on the
correlation between constructs.
In order to analyse whether the attitude±behaviour (or more speci®cally
the personal norm-behaviour) relationship concerning packaging waste
prevention depends on how much the person recycles, 2 the sample is di-
vided at the mean of the recycling index. The simplest way of testing the
hypothesis is to compare the product±moment correlation of the personal
norm and behaviour indexes between the groups. If LISREL8 is used to
perform the test, the focus is on whether the b parameter for the path from
personal norm to behaviour vary between the two groups. This can be done
by comparing the ®t of two models to the data: one model where the b
parameter for the relationship between the personal norm and behaviour is
®xed to be the same in both groups and one where the parameter is set free
to vary between the groups (see, e.g., Bagozzi, 1994; JoÈreskog & SoÈrbom,
1993). If the strength of the correlation between the personal norm and
behaviour regarding packaging waste prevention depends on how much the
person recycles (an interaction e€ect), setting the parameter free should
produce a signi®cant improvement in ®t, measured by v2d . In the present
case, neither of the two methods produce any indication that personal
norms and behaviour regarding waste avoidance depends on how much the
person recycles. The correlation coecient is not signi®cantly di€erent be-
tween the two groups (at the 5% level) and the model ®t is not improved by
setting the b parameter free to vary. Hence, the hypothesis is not supported
by the data. 3

4.2. Hypothesis 2

Speci®ed to the present case, Hypothesis 2 states that participation in a

recycling programme changes attitudes concerning other waste related

2
Notice that we only measured the number of times the person recycled two speci®c types of waste out
of the last ®ve opportunities which, of course, is a rough indicator of the person's direct experience with
recycling. It makes little sense to compare recyclers and non-recyclers, since 86% of the respondents
claimed that they recycle all of their glass waste and an additional 9% claimed that they recycle some of it.
3
In the interest of saving space, the calculations are not shown. However, they can be obtained from the
author.
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 65

behaviours because of the learning it generates about the waste-conse-

quences of these behaviours. This hypothesis can be tested by comparing
the ®t of two nested structural equation models to the correlation matrix
in Table 1 by means of the v2d test. Model 1 treats personal norms as
independent and the two behaviours as dependent variables, letting each of
the measured personal norms predict their respective behaviours. Model 2
is identical to Model 1 except that it adds a causal path from recycling to
the norm of packaging waste prevention, the hypothesis being that this
path will be signi®cant, positive, and increase the overall ®t of the model
(see Fig. 1).
Fig. 1 shows the results of adding a causal path from recycling to the
personal norm of packaging waste prevention. The invariance test and the
parameter value show that the path is highly signi®cant. Hence, a spillover
e€ect from recycling does indeed seem to exist in this case. However, it is not
the kind of spillover e€ect expected. The sign of the parameter linking re-
cycling to the personal norm of packaging waste prevention is negative, in-
dicating that recycling reduces the feeling of obligation to do other things to
prevent waste. Hence, Hypothesis 2 is not supported by the data. Instead, the
data indicates that some consumers perceive recycling as sucient to solve
the waste problem, making waste avoidance super¯uous, or, as has been
suggested by some researchers (Schahn, 1993a; Wenke, 1993), use recycling
as a convenient excuse for continuing a traditional, unsustainable lifestyle in
other areas. This e€ect of recycling on the propensity to prevent waste may
still be regarded as a spillover e€ect, but a negative one.

Table 1
Polychoric correlation matrix for the combined data regarding recycling and packaging waste prevention
(N ˆ 794)

x1 x2 x3 x4 x5 x6

Personal norms
x1 : Recycle glass 1.00
x2 : Recycle compostable waste 0.33 1.00
x3 : Avoid overpackaged goods 0.29 0.36 1.00
x4 : Choose beverages in recyclable packaging 0.30 0.37 0.57 1.00
Behaviour
x5 : Recycling 0.17 0.37 0.11 0.08 1.00
x6 : Packaging waste prevention 0.09 0.19 0.29 0.34 0.15 1.00
66 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

Fig. 1. Spillover from recycling to the personal norm with respect to packaging waste prevention.

4.3. Hypothesis 3

Hypothesis 3 consists of two sub-hypotheses: (1) that recycling increases

the salience of more general environmentally relevant attitudes and values,
and (2) that, when made more salient, these general attitudes and values
increase the propensity to perform other behaviours with similar environ-
mental consequences. This later in¯uence is usually supposed to be mediated
through speci®c attitudes and norms (e.g., Alwitt & Pitts, 1996; Homer &
Kahle, 1988; Stern & Oskamp, 1987; Thùgersen & Grunert-Beckmann,
1997). Hence, Hypothesis 3 implies that speci®c attitudes and norms con-
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 67

cerning recycling and packaging waste prevention have common roots in

more general attitudes and values which then function as more distal deter-
minants of behaviour (Eagly & Chaiken, 1993).
The high correlation between the two personal norms shown in Fig. 1
lends support to the second sub-hypothesis. The correlation may indeed be
caused by them having shared antecedents. The survey contains a sample of
13 items from Shalom Schwartz's Value Indicator (Schwartz, 1992, 1994;
Schwartz & Bilsky, 1987, 1990; see Thùgersen & Grunert-Beckmann, 1997);
one of the most thoroughly tested values measurement instruments. These
data allow us to investigate whether environmentally relevant values are
included in the (possible) set of shared determinants.
The full version of Schwartz's Value Indicator contains 52 items, typically
(and also in this case) measured on a nine-point scale with the end points
``not important at all'' and ``of decisive importance as a guiding principle in
my life.'' According to Schwartz (1992), individuals use values to evaluate
objects, events, other people, and themselves, and to select and justify ac-
tions. However, it seems unlikely that anyone would ®nd it necessary to
consider all the values represented by the 52 items in Schwartz's Value In-
dicator in order to evaluate an everyday activity, such as recycling or
avoiding wasteful packaging. In fact, it seems much more likely that only one
or a couple of values would be involved in making such an evaluation. The
problem is, of course, that it is dicult to know on beforehand which values
will determine the attitude (or personal norm) towards a given activity.
Hence, stepwise regression ± an exploratory form of data analysis ± was
chosen to investigate whether personal norms concerning the two activities
are rooted in the same values. Since it seems reasonable to assume that in-
ability to answer a value question indicates that the particular value is un-
important to the individual, missing values for these items have been recoded
as ``not important at all''. Personal norm indexes for each activity ± con-
structed by summing responses to the individual personal norm questions ±
were used as criterions. Table 2 shows the value items that contribute sig-
ni®cantly to the explanation of personal norms, their marginal contribution
in terms of DR2 and their coecients in the ®nal equation.
Table 2 con®rms that personal norms regarding the analysed behaviours
are rooted, and that they share common roots, in the person's general values.
The expectation that only a few of a person's complete set of values are in-
volved in the evaluation of these activities is also con®rmed (with the reser-
vation that only a subset of the items in Schwartz's Value Indicator is
included in the analyses). The ®rst of the included value items to enter the
 68

Table 2
The in¯uence of values on personal norms regarding recycling and waste avoidance

Personal norm regarding: Recycling (N ˆ 794) Waste avoidance (N ˆ 794) Waste avoidance, Waste avoidance,
Values ``light'' recyclers ``heavy'' recyclers
N ˆ 395 N ˆ 399

Coecient Adj. R2 DR2 Coecient Adj. R2 DR2 Coecient Adj. R2 Coecient Adj. R2
after in- after in-
clusion clusion

Intercept 4.15 2.05 1.96 0.072 2.35 0.049

Unity with nature 0.18 0.046 0.046 0.21 0.037 0.037 0.20 0.19
Curious 0.06 0.056 0.010 0.13 0.057 0.020 0.15 0.10
Choosing own goals 0.09 0.061 0.005
In¯uential 0.13 0.067 0.010 0.11 0.015

Note: The ®rst two calculations were made by means of stepwise regression. The coecients are those that appear in the ®nal equation. In the last
two calculations, multiple regression was used. Besides the ones mentioned in the ®rst column, the following value items were included in the
stepwise regression analysis: Responsible, Helpful, Enjoying life, A world of beauty, A world at peace, Independent, Preserving my public image,
Respect for tradition, Social order.
J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 69

equation in both cases, and by far the most important of the set, is ``unity
with nature''. Other studies have also found this item to be the one from
Schwartz's Value Indicator most closely associated with environmental
concern (e.g., Stern, Dietz & Kalof, 1993; Stern et al., 1995) which increases
the face validity of this ®nding.
``Being curious'' is the second item to enter the equation in both cases.
There is no previous research indicating why this value may in¯uence per-
sonal norms concerning environmental issues. A possible explanation could
be that an active interest in new features in one's surroundings, as indicated
by this value, facilitates the development of personal norms regarding
emerging issues (as many environmental issues are still perceived to be; cf.
Stern et al., 1995; Thùgersen & Grunert-Beckmann, 1997).
Di€erent values enter the two equations third and last. The personal norm
regarding recycling is marginally in¯uenced by the value ``choosing my own
goals'', and the personal norm regarding waste avoidance by the value ``being
in¯uential''. A common trait of these two values is that they both indicate an
active, self-controlled approach to life. Perhaps individuals with such an
approach are more likely than others to form an attitude towards any rela-
tively new issue, environmental or not.
Values (at least those included here) explain only a modest share of the
variation in personal norms. This should come as no surprise in view of the
wide gap in abstraction levels between criterion and predictors (e.g., Ajzen &
Fishbein, 1977). Further, norm research suggests that the formation of a
personal norm depends on the person's perceptions regarding a number of
situation speci®c issues, besides his or her values (e.g., Schwartz, 1977).
As mentioned earlier, the survey contained no measures of the salience of
the included mental constructs. However, the indirect procedure that was
used to test the in¯uence of recycling on attitude salience when investigating
Hypothesis 1 also applies here. If recycling has made environmentally rel-
evant values more salient, the correlation between these values and personal
norms regarding waste avoidance should be expected to be stronger for
``heavy'' than for non- or ``light'' recyclers. However, this expectation rests
on the assumption that the individual's perception of the relevance of waste
prevention for reaching the valued goal is not (negatively) a€ected by his or
her recycling. Compared to non- or light recyclers, heavy recyclers who
perceive recycling as a sucient solution to the waste problem, as some
seem to do according to the results of the test of Hypothesis 2, may per-
ceive their environmentally relevant values as less relevant for decisions
about waste avoidance. As a consequence, the correlation between values
70 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

Fig. 2. Spillover from recycling to packaging waste prevention and from recycling to both the personal
norm with respect to packaging waste prevention and packaging waste prevention.

and personal norms may be weaker for heavy than for non- or light re-
cyclers.
The right half of Table 2 shows the results of running multiple regression
analyses of the personal norm regarding packaging waste avoidance on the
three value items identi®ed by the stepwise regression separately for heavy
and for non- or light recyclers. The two groups were formed by splitting the
sample at the mean of the recycling index. A comparison of the results shows
that values in fact explain less of the variation in personal norms regarding
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 71

waste avoidance among heavy than among non- or light recyclers, indicating
that recycling makes a number of people perceive their values as less relevant
for waste avoidance (perhaps because they believe that recycling solves the
waste problem). However, the di€erence in the coecient for ``unity with
nature'' ± the only speci®cally ``environmental'' value among the three ± is
not statistically signi®cant. Hence, either neither the salience nor the rele-
vance of this value is in¯uenced by recycling or the two (counteracting) forces
cancel each other. However, most importantly for the present purpose, the
evidence provides no support for the hypothesis that recycling increases the
salience of environmental values.

4.4. Hypothesis 4

Hypothesis 4 implies that recyclers conceive of recycling activities in an

increasingly inclusive category with increasing experience and that this in-
creases the likelihood that other (waste preventing or environmentally
friendly) activities covered by the same mental category are performed as
well. The hypothesis was meant to explain a fairly ``mindless'' spillover
from recycling to related behaviours. It implies that a path from recycling
to waste prevention can explain some of the unexplained variance in the
latter variable when mental antecedents have been accounted for. Hence,
the procedure used to test Hypotheses 2 can be used to test this hypothesis
too. The test is performed by comparing the ®t of the models in Fig. 1 to
the ®t of a similar model that includes a causal path from recycling to
packaging waste prevention (Fig. 2). The invariance test is signi®cant
(p < 0.01) and since the sign of the path is positive (as predicted) it can be
concluded that the type of spillover implied by Hypothesis 4 is supported by
the data.
In Fig. 2, the signi®cant paths identi®ed in connection with the test of
Hypotheses 2 and 4 are combined. Both the causal path from recycling to
the personal norm referring to packaging waste prevention and the one
from recycling directly to packaging waste prevention still produce a sig-
ni®cant improvement in model ®t when the other path has been controlled
for.
Judged by the size of the v2 the ®nal model ®ts poorly. However, this is
primarily due to the v2 -testor being dependent on the sample size. The CFI is
very satisfactory and so are other testors produced by LISREL8 (for example
a RMSEA of 0.049). Hence, the ®nal model gives an acceptable represen-
tation of the data.
72 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

The combined model explains approximately the same share of the vari-
ation in the two behaviours as the individual models. 4 Hence, the primary
advantage of combining the models is not that it increases the explanation,
but that it increases our understanding of the behaviours in question. The
combined model reveals important spillover e€ects between the two behav-
iours.

5. Discussion

This paper began with a plea to behavioural research and policy in the
environmental ®eld to leave its narrow focus on speci®c behaviours and start
dealing with the more dicult problem of changing the whole lifestyle in
a‚uent societies. Compared to this grand plea, the paper itself takes a small
step, focusing on two speci®c behaviours rather than one, and even behav-
iours within the same environmental ``sector''. However, the most important
contribution of this paper is not new knowledge about these speci®c be-
haviours. Its novel contribution is demonstrating that environmentally
friendly behaviours are not independent. General environmental values that
people hold foster feelings of behavioural obligation and commitment in
many settings, and when people start to act in an environmentally friendly
way in one area, this behaviour tends to spill over into other areas. Both of
these observations are of tremendous importance for developing strategies to
produce general lifestyle changes.

5.1. General environmental values

This study took o€ from an analysis showing that the attitude towards
recycling or preventing packaging waste mainly depends on the strength of
the person's feeling of obligation to perform these behaviours. By demon-
strating that the feeling of a personal obligation to perform speci®c, envi-
ronmentally friendly behaviours is rooted in more general, internalised
values, the analysis suggests that education is an important tool for pro-
ducing behaviour change in the environmental ®eld (Gray, 1985; Rokeach,
1979).

4
In the ®nal model, R2 's regarding recycling and waste prevention are 0.24 and 0.19. The calculations
regarding the individual models are reported in Thùgersen (1998a).
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 73

More than a quarter of a century ago, Heberlein (1972) suggested that

environmental attitudes have changed from an economic to a moral orien-
tation mainly because technological change has provided new non-polluting
options and, hence, made producers and consumers responsible for choosing
the polluting ones, and because science has uncovered a large number of
previously unknown or ignored consequences of human behaviour for the
environment. Compared to the early 1970s the availability of environmen-
tally friendly choice options and the scienti®c knowledge about environ-
mentally harmful consequences of human behaviour have increased and
disseminated tremendously in the industrialised countries. However, the
rapid growth in knowledge within this area and the perennial addition of new
generations maintain a continuing need of environmental education, both
within and outside the formal education system. The literature on normative
decision-making suggests that education and information campaigns should
provide both information about environmental consequences of human be-
haviour, speci®c how-to-do-it information, and information showing the
positive e€ects of changing behaviour (e.g., feedback) (Schwartz & Howard,
1981).
However, according to this literature, knowledge in this broad sense is
only one of the prerequisites of forming environmentally friendly personal
norms. Another prerequisite is that reasonable alternatives to environ-
mentally harmful behaviours are available, which is much too often not the
case. For example, a poll published in Business Week, 19 February 1996,
found that 63% of adult Americans feel that they could not live without an
automobile. This perception mirrors the current American society's total
dependence on automotive transportation and that no real alternative to a
private automobile exists in most areas (Dholakia, Dholakia & Firat,
1983). Hence, strengthening personal norms concerning environmentally
friendly behaviour and increasing the range of speci®c behaviours that
these norms are applied to depends on increasing the number and avail-
ability of environmentally friendly choice options as much as on increasing
knowledge.

5.2. Positive spillover e€ects

The ®nding that behaving in an environmentally friendly way in one type

of situation (separating the household's waste for recycling) tend to spill over
and lead to more environmentally friendly choices in a di€erent situation
(shopping), seemingly without mediating cognitive processes, has a number
74 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

of implications. One implication (if this is a general phenomenon) is that it

becomes less important which activity targeting an environmental problem a
hitherto passive person adopts, as long as he or she starts doing something
that contributes to solve the problem. Hence, the existence of spillover e€ects

support Olander and Thùgersen's (O È lander & Thùgersen, 1995) suggestion
that activity goals should substitute ``state goals'' (the goal of producing a
speci®c ``state'' in the target audience) in behaviour change campaigns. An-
other implication is that when a behaviour change campaign is based on
narrow ``sector-thinking'', there is a risk that important consequences of the
campaign are overlooked. Consequences that are not monitored are likely to
be ignored when choosing between behaviour change approaches. If di€erent
behaviour change approaches produce di€erent spillover e€ects, as argued by
Frey (1993), sub-optimal policy decisions are almost certain to follow from
ignoring these e€ects.

5.3. Negative spillover e€ects

The analysis also detects a negative spillover e€ect from recycling to re-
lated behaviours. Performing an environmentally friendly behaviour, such as
recycling, seem to have a negative impact (spillover e€ect) on the feeling of
obligation to do other things.
Two possible reasons for this e€ect were suggested above. The gentler
explanation is that some people (mistakenly) believe that recycling solves
the waste problem. Hence, when one recycles there is no need to do more
about this problem. This is not the explanation most observers prefer,
however. Most observers of this phenomenon have been concerned with the
possibility that some people adapt to demands for minor behaviour changes
(like participation in a convenient recycling programme) as an excuse for
not making more radical, and personally more costly, changes in their
lifestyles (Schahn, 1993a,b; Van Raaij, 1995; Wenke, 1993). If this is the
true explanation, it seems to be a tougher job to reverse it than if the only
problem is mistaken beliefs. But perhaps this is a mistake too. Van Raaij
(1995) has suggested that a person who uses this excuse has in fact accepted
that he or she ought to change behaviour. ``Partial yielding'', as he terms it,
may in fact make it easier to persuade a person to make more radical be-
haviour changes.
Two ways of handling excuses for not changing behaviour in spite of the
availability of more environmentally friendly options have been suggested in
the literature: counterarguing the excuses (e.g., Schahn, 1993a; Van Raaij,
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 75

1995; for an example, see Schahn, 1993b) and changing the balance between
personal cost and bene®ts from choosing the environmentally friendly op-
tion (e.g., Diekmann & PreisendoÈrfer, 1992). Counterarguing is appropriate
in cases where the resistance against adapting to an obviously more envi-
ronmentally friendly behaviour is caused only by the convenience of main-
taining an old habit. Counterarguing basically means confronting people
with the same types of information that are needed in order to correct
mistakes about the need for waste prevention even though one recycles.
Hence, in this respect it makes little di€erence whether the negative spillover
e€ect is produced by one or the other of the two suggested explanations.
However, it is important to remember that campaigns that promote be-
haviour change in the environmental ®eld typically ask the individual to take
responsibility for solving problems that he or she did not cause (Salmon,
1989). To the degree possible, the balance between personal cost and bene®ts
from choosing environmentally friendly options should be adjusted to re¯ect
this fact.

5.4. Not supported hypotheses

Three of the suggested spillover mechanisms were not supported by the

present analyses. They did not con®rm that performing an environmentally
friendly behaviour (such as recycling) makes attitudinal and more distal
antecedents (values) of related behaviours (waste prevention) more predictive
of the next step in the assumed causal chain (due to increased salience). The
analyses also failed to con®rm that performing an environmentally friendly
behaviour (such as recycling) may lead to learning about environmental
consequences of other behaviours and, hence, change the attitude towards
these behaviours in a more pro-environmental direction.
As regards the latter, the study does not prove that no such learning takes
place in the analyses case, but it gives a fairly unequivocal indication that any
attitudinal in¯uence of such learning is more than balanced by the just
mentioned negative spillover e€ect. The balance between these two forces
obviously may vary from case to case. There may be other cases where the
learning spin-o€ from performing an environmentally friendly behaviour
may more than outweigh eventual negative spillovers. Future research should
investigate whether there are identi®able conditions that are more facilitating
for positive spillover than others.
The non-con®rmation of the other two hypotheses indicates that if there
are spillover mechanisms that work via increased salience of various mental
76 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

constructs they are rather weak. However, again this ®nding of course is
in¯uenced by the peculiarities of the present case. Hence, further empirical
studies are needed before any conclusions can be drawn with con®dence. It
should also be kept in mind that the present data set is not ideal for detecting
salience e€ects. Future research should include direct measurement of sa-
lience in the research design. Also, a more balanced sample of performers and
non-performers of the behaviour assumed to generate spillover should be
striven for.

5.5. Limitations

The most important limitation of this study is that it is based on a single

case only. Connections, like common normative antecedents and spillover,
are bound to vary between dyads of behaviours. Hence, replications on many
other dyads are needed before we will know to what degree the ®ndings are
generally valid and to what degree they depend on conditions that are special
for the analysed behaviours. Further, it should be remembered that the
analysis is based on correlational data which gives limited possibilities for
concluding about the causal properties of the detected relationships. The
certainty with which conclusions can be drawn depends on the precision in
measurement which in the present case su€ers form the low number of items
used to estimate the concepts of interest. Finally, some of the analyses were
weakened by low variation in the available data. However, even given these
limitations this research has contributed to the literature by pointing out
some mechanisms for spillover processes between pro-environmental be-
haviours, or ± in a more general formulation ± processes whereby a new
behavioural pattern may evolve, and by demonstrating a method for ana-
lysing such spillover processes.

Acknowledgements

The ®rst draft of this paper was written during my visit at the Institute for
International Studies, Stanford University, in the academic year 1995±1996. I
am grateful to Richard P. Bagozzi and two anonymous reviewers for com-
ments on an earlier draft. The study was supported by a grant from the
Danish Government's Strategic Environmental Research Programme. It is
part of a collaborative research project which looks into the prerequisites for
reducing household waste production. Besides this author, the project group
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 77

consists of Suzanne C. Beckmann, Anne Marie Dahler-Larsen, Erik Klop-


penborg Madsen, and Folke Olander.

References

Ajzen, I., & Fishbein, M. (1977). Attitude±behavior relations: a theoretical analysis and review of
empirical research. Psychological Bulletin, 84, 888±918.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cli€s,
NJ: Prentice-Hall.
Alwitt, L. F., & Pitts, R. E. (1996). Predicting purchase intentions for an environmentally sensitive
product. Journal of Consumer Psychology, 5, 49±64.
Anderson, J. R. (1983). A spreading activation theory of memory. Journal of Verbal Learning and Verbal
Behavior, 22, 261±275.
Bagozzi, R. P. (1994). Structural equation models in marketing research: basic principles. In R. P. Bagozzi,
Principles of marketing research. Cambridge: Blackwell.
Bech-Larsen, T. (1996). Danish consumers' attitudes to the functional and environmental characteristics
of food packaging. Working paper (32). The Aarhus School of Business, Centre for Market
Surveillance, Research and Strategy for the Food Sector.
Berger, I. E., & Kanetkar, V. (1995). Increasing environmental sensitivity via workplace experiences.
Journal of Public Policy & Marketing, 14, 205±215.

Berger, K. (1991). Okologische Verantwortung. Ein Ziel f ur die Umwelt- und Verbraucherbildung
Erwachsener. Bad Heilbrunn/Obb.: Julius Klinkhardt.
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: recycling the
concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58,
1015±1026.
Collins, A. M., & Loftus, E. F. (1975). A spreading activation theory of semantic memory. Psychological
Review, 82, 407±428.
Commission of the European Communities (1996). Communication from the commission on the review of
the community strategy for waste management. Brussels: Commission of the European Communities.
Devine, P. G., & Hirt, E. R. (1989). Message strategies for information campaigns: a social psychological
analysis. In: C. T. Salmon, Information campaigns: Balancing social values and social change (pp. 229±
259). Newbury Park, CA: Sage.
Dholakia, R. R., Dholakia, N., & Firat, A. F. (1983). From social psychology to political economy: a
model of energy use behavior. Journal of Economic Psychology, 3, 231±247.
Diekmann, A., & Preisend orfer, P. (1992). Pers onliches Umweltverhalten ± Diskrepanzen zwischen
Anspruch und Wirklichkeit. K olner Zeitschrift f
ur Soziologie und Sozialpsychologie, 44, 226±251.
Doll, J., & Ajzen, I. (1992). Accessibility and stability of prediction on the theory of planned behaviour.
Journal of Personality and Social Psychology, 63, 754±765.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth: Harcourt Brace Jovanovich
College Publishers.
Ellen, P. S. (1994). Do we know what we need to know? Objective and subjective knowledge e€ects on pro-
ecological behaviors. Journal of Business Research, 30, 43±52.
Ester, P., & Mandemaker, T. (1994). Socialization of environmental policy objectives: tools for
environmental marketing. In: Dutch Committee for Long-Term Environmental Policy, The environ-
ment: Towards a sustainable future (pp. 49±80). Dordrecht: Kluwer Academic Publishers.
Eurostat et al. (1995). Europe's environment: Statistical compendium for the Dobris assessment (8A).
Luxembourg: Oce for Ocial Publications of the European Communities.
78 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

Fazio, R. H. (1986). How do attitudes guide behavior? In: R. M. Sorrentino, & E. T. Higgins, The
handbook of motivation and cognition: Foundations of social behavior (pp. 204±243). New York:
Guilford Press.
Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude-behavior consistency. In: L.
Berkowitz, Advances in experimental social psychology (pp. 161±202). New York: Academic Press.
Fishbein, M. (1967). Attitude and the prediction of behavior. In: M. Fishbein, Readings in attitude theory
and measurement (pp. 477±492). New York: Wiley.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. Reading, MA: Addison±Wesley.
Fiske, S. T., & Pavelchak, M. A. (1986). Category-based versus piecemeal-based a€ective responses. In: R.
M. Sorrentino, & E. T. Higgins, The handbook of motivation and cognition: Foundations of social
behavior (pp. 167±203). New York: Guilford Press.
Frey, B. S. (1993). Motivation as a limit to pricing. Journal of Economic Psychology, 14, 635±664.
Gardner, G. T., & Stern, P. C. (1996). Environmental problems and human behavior. Boston, MA: Allyn &
Bacon.
Garling, T., & Sandberg, L. (in press). A commons-dilemma approach to households' intentions to change
their travel behavior. In: M. E. H. Lee-Goselin, & P. R. Stopher, Understanding travel behavior in an
era of change. New York: Pergamon Press.
Gray, D. B. (1985). Ecological beliefs and behaviors: Assessment and change. Westport, CT: Greenwood
Press.
Grunert, S. C., & Juhl, H. J. (1995). Values, environmental attitudes, and buying of organic foods. Journal
of Economic Psychology, 16, 39±62.
Guagnano, G. A., Stern, P. C., & Dietz, T. (1995). In¯uences on attitude-behavior relationships. A natural
experiment with curbside recycling. Environment and Behavior, 27, 699±718.
Heberlein, T. A. (1972). The land ethic realized: some social psychological explanations for changing
environmental attitudes. Journal of Social Issues, 28, 79±87.
Homer, P. M., & Kahle, L. R. (1988). A structural equation test of the value-attitude-behaviour hierarchy.
Journal of Personality and Social Psychology, 54, 638±646.
Jensen, R. (1992). Reputational spillovers, innovation, licensing and entry. International Journal of
Industrial Organization, 10, 193±212.
Johansson, M. G. (1993). Et adfrdsteoretisk grundlag for regulering af husholdningernes a€aldsbortska€-
else (A behaviour theoretical framework for the regulation of houshold waste disposal). Unpublished
Ph.D. Dissertation. Aarhus: Institut for Markedsùkonomi, Handelshùjskolen i Aarhus.
J
oreskog, K. G., & S orbom, D. (1993). LISREL8: Structural equation modelling with the SIMPLIS
command language. Chicago: Scienti®c Software International.
Jùrgensen, C. E. (1989). Om hundrede  ar er intet glemt. Industri og naturgrundlag (In hundred years nothing
is forgotten. Industry and nature). T astrup: Forlaget Niche.
Miljù-og Energiministeriet (1995). Natur-og miljpolitisk redegrelse 1995 (Nature and environmental
policy statement 1995). Kùbehavn: Miljù-og Energiministeriet.
Moisander, J., & Uusitalo, L. (1994). Attitude±behavior inconsistency: limitations of the reasoned action
approach in predicting behavior from proenvironmental attitudes. In G. Antonides, & W. F. v. Raaij,
Proceedings of the IAREP/SABE Conference, Rotterdam, 10±13 July, pp. 560±579.
Morris, L. A., Hastak, M., & Mazis, M. B. (1995). Consumer comprehension of environmental advertising
and labeling claims. The Journal of Consumer A€airs, 29, 328±350.
Norwegian Ministry of Environment (1994). Symposium: Sustainable consumption. 19±20 January 1994,
Oslo, Norway. Oslo: Norwegian Ministry of Environment.

Olander, F., & Thùgersen, J. (1995). Understanding of consumer behaviour as a prerequisite for
environmental protection. The Journal for Consumer Policy, 18, 317±357.
Oslo Roundtable Conference on Sustainable Production and Consumption (1995). Elements for an
International Work Programme on Sustainable Production and Consumption. Report for conference.
Oslo: Norwegian Ministry of Environment.
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 79

Peter, J. P., & Olson, J. C. (1996). Consumer behavior and marketing strategy (4th ed.). Homewood, IL:
Irwin.
Pickett, G. M., Kangun, N., & Grove, S. J. (1993). Is there a general conserving consumer? a public policy
concern. Journal of Public Policy & Marketing, 12, 234±243.
Pieters, R. G. M. (1991). Changing garbage disposal patterns of consumers: Motivation, ability, and
performance. Journal of Public Policy & Marketing, 10, 59±76.
Rathje, W., & Murphy, C. (1992). Rubbish! The archaeology of garbage. New York: HarperCollins.
Regan, D. T., & Fazio, R. (1977). On the consistency between attitudes and behavior: look to the method
of attitude formation. Journal of Experimental Social Psychology, 13, 28±45.
Rokeach, M. (1968). Beliefs, attitudes, and values. San Francisco, CA: Jossey Bass.
Rokeach, M. (1973). The nature of human values. New York: Free Press.
Rokeach, M. (1979). Value theory and communication research: Review and commentary. In: D. Nimmo,
Communication yearbook 3 (pp. 7±28). New Brunswick: Transaction Books.
Salmon, C. T. (1989). Campaigns for social `improvements': an overview of values, rationales, and
impacts. In C. T. Salmon, Information campaigns: Balancing social values and social change (pp. 19±
53). Newbury Park, CA: Sage.
Schahn, J. (1993a). Die Rolle von Entschuldigungen und Rechtfertigungen f ur umweltsch adigendes
Verhalten. In J. Schahn, & T. Giesinger, Psychologie f ur den Umweltschutz (pp. 51±61). Weinheim:
BELTZ, Psychologie Verlags Union.
Schahn, J. (1993b). Psychologische Maûnahmen zur F orderung von M ulltrennung und -vermeidung: Ein
Anwendungsbeispiel f ur das Modell von Fietkau und Kessel. In J. Schahn, & T. Giesinger, Psychologie
f
ur den Umweltschutz (pp. 123±131). Weinheim: BELTZ, Psychologie Verlags Union.
Schwartz, S. H. (1977). Normative in¯uence on altruism. In L. Berkowitz, Advances in experimental social
psychology, vol. 10 (pp. 221±279). New York: Academic Press.
Schwartz, S. H. (1992). Universals in the content and structure of values: theoretical advances and
empirical tests in 20 countries. In M. P. Zanna, Advances in experimental social psychology, vol. 25 (pp.
1±65). San Diego, CA: Academic Press.
Schwartz, S. H. (1994). Are there universal aspects in the structure and content of human values?. Journal
of Social Issues, 50 (4), 19±45.
Schwartz, S. H., & Bilsky, W. (1987). Toward a universal psychological structure of human values. Journal
of Personality and Social Psychology, 53, 550±562.
Schwartz, S., & Bilsky, W. (1990). Toward a theory of universal content and structure of values: extensions
and cross-cultural replications. Journal of Personality and Social Psychology, 58, 878±891.
Schwartz, S. H., & Howard, J. A. (1981). A normative decision-making model of altruism. In J. P.
Rushton, & R. M. Sorrentino, Altruism and helping behavior: Social, personality, and developmental
perspectives (pp. 189±211). Hillsdale, NJ: Lawrence Erlbaum.
Sitarz, D. (Ed.) (1994). Agenda 21: The earth summit strategy to save our planet. Boulder: EarthPress.
Worldwatch environmental alert series.
SPSS (1993). SPSS/PC+. LISREL7 and PRELIS. Chicago: SPSS Inc.
Stern, P. C. (1992). Psychological dimensions of global environmental change. Annual Review of
Psychology, 43, 269±302.
Stern, P. C., Dietz, T., & Kalof, L. (1993). Value orientations, gender, and environmental concern.
Environment and Behavior, 25, 322±348.
Stern, P. C., Dietz, T., Kalof, L., & Guagnano, G. (1995). Values, beliefs, and proenvironmental action:
attitude formation toward emergent attitude objects. Journal of Applied Social Psychology, 25, 1611±
1636.
Stern, P. C., & Oskamp, S. (1987). Managing scarce environmental resources. In D. Stokols, & I. Altman,
Handbook of environmental psychology (pp. 1043±1088). New York: Wiley.
Stù, E. (1995). Sustainable Consumption ± Report from the International Conference on Sustainable
Consumption. Working report (2). Lysaker: National Institute for Consumer Research (SIFO).
80 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81

Sugden, R. (1989). Spontaneous order. Journal of Economic Perspectives, 3, 85±98.

Sujan, M. (1985). Consumer knowledge: e€ects on evaluation strategies mediating consumer judgments.
Journal of Consumer Research, 12 (June), 31±46.
Taylor, S., & Todd, P. (1995). Understanding household garbage reduction behavior: a test of an
integrated model. Journal of Public Policy and Marketing, 14, 192.
The Directorate Industry-, Construction-, Product-, and Consumer-a€airs et al. (1995). Facilities for a
sustainable household. Workshop report. The Hague, The Netherlands: Ministry of Housing, Spatial
Planning and the Environment, Directorate Industry-, Construction-, Product-, and Consumer-a€airs,
Directorate of International Environmental A€airs.
Thùgersen, J. (1994). A model of recycling behaviour. With evidence from Danish source separation
programmes. International Journal of Research in Marketing, 11, 145±163.
Thùgersen, J. (1996a). The demand for environmentally friendly packaging in Germany. MAPP Working
Paper (30). The Aarhus School of Business, Centre for Market Surveillance, Research and Strategy for
the Food Sector.
Thùgersen, J. (1996b). Wasteful food consumption. Trends in food and packaging waste. The
Scandinavian Journal of Management, 12, 291±304.
Thùgersen, J. (1997). Forbrugeradfñrd og miljùbelastning (Consumer behaviour and environmental
impacts). Samfundskonomen (1997/2), 5±11.
Thùgersen, J. (1998a). The norms±attitude±behavior relationship. Theory and applications in the
environmental domain. Working Paper 98-3, Aarhus School of Business, Department of Marketing,
The Faculty of Business Administration.
Thùgersen, J. (1998b). Understanding behaviours with mixed motives. An application of a modi®ed
Theory of Reasoned Action on consumer purchase of organic food products. In B. G. Englis & A.
Olofsson, European Advances in Consumer Research, vol. 3 (pp. 286±287). Provo: Association for
Consumer Research.
Thùgersen, J., & Andersen, A. K. (1996). Environmentally friendly consumer behavior: the interplay of
moral attitudes, private costs, and facilitating conditions. In R. P. Hill, & C. R. Taylor, Marketing and
Public Policy Conference Proceedings, vol. 6 (pp. 80±96). Washington, DC: American Marketing
Association.
Thùgersen, J., & Grunert-Beckmann, S. C. (1997). Values and attitude formation towards emerging
attitude objects: from recycling to general, waste minimizing behavior. In D. MacInnis, & M. Brucks,
Advances in consumer research, vol. XXIV (pp. 182±189). Provo: Association for Consumer Research.
Vallacher, R. R., & Wegner, D. M. (1985). A theory of action identi®cation. Hillsdale, NJ: Lawrence
Erlbaum.
Van Raaij, F. (1995). A hierarchy of excuses for not behaving in a pro-environmental way. In: E. Stù,
Sustainable consumption ± report from the International Conference on Sustainable Consumption (pp.
209±244). Lysaker: National Institute for Consumer Research (SIFO).
Van Raaij, W. F., & Verhallen, T. M. M. (1983). A behavioural model of residential energy use. Journal of
Economic Psychology, 3, 39±63.
Verplanken, B., Aarts, H., Knippenberg, A. v., & Knippenberg, C. v. (1994). Attitude versus general habit:
antecedents of travel mode choice. Journal of Applied Social Psychology, 24, 285±300.
Wegner, D. M., & Vallacher, R. R. (1986). Action identi®cation. In R. M. Sorrentino, & E. T. Higgins,
The handbook of motivation and cognition: Foundations of social behavior (pp. 550±582). New York:
Guilford Press.
Weigel, R. H. (1983). Environmental attitudes and the prediction of behavior. In N. R. Feimar & E. S.
Geller, Environmental psychology. Directions and perspectives (pp. 257±287). New York: Praeger.
Wenke, M. (1993). Konsumstruktur, Umweltbewutsein und Umweltpolitik. Eine markro okonomische
Analyse des Zusammenhanges in ausgew ahlten Konsumbereichen. Berlin: Duncker & Humblot.
Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust implications. New York: Free
Press.
 J. Thùgersen / Journal of Economic Psychology 20 (1999) 53±81 81

Wuttke, J. (1993). Regulatory measures on avoidance, reuse and material recycling in Germany. TIMES ±
Members' Information from the ISWA General Secretariat, 2, 11±15.
Yates, S., & Aronson, E. (1983). A social psychological perspective on energy conservation in residential
buildings. American Psychologist, 38, 435±444.