Content uploaded by Patrik Sörqvist
Author content
All content in this area was uploaded by Patrik Sörqvist on Oct 20, 2023
Content may be subject to copyright.
Resources, Conservation & Recycling Advances 19 (2023) 200186
Available online 14 October 2023
2667-3789/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Managing waste behavior by manipulating the normative appeal of trash
bins: Lessons from an urban eld experiment
Noah Linder
a
,
d
,
*
, Patrik S¨
orqvist
a
,
e
, Therese Lindahl
b
,
f
, Robert Ljung
c
a
Department of Building Engineering, Energy Systems and Sustainability Science, University of G¨
avle, G¨
avle, Sweden
b
Beijer Institute of Ecological Economics, Royal Swedish Academy of Sciences, Stockholm, Sweden
c
Swedish Agency for Work Environment Expertise, Sweden
d
Global Economic Dynamics and the Biosphere, Royal Swedish Academy of Science, Stockholm, Sweden
e
Department of Health, Learning and Technology, Luleå University of Technology, Luleå, Sweden
f
Stockholm Resilience Centre, Stockholm University, Stockholm, Sweden
ARTICLE INFO
Keywords:
Pro-environmental behavior
Littering
Saliency
Norms
Design
Physical environment
ABSTRACT
Littering is a problem in many human societies. In this study, 9 individual street bins were manipulated on a
central street in the city of G¨
avle, Sweden. The aim was to explore if changing the appearance of the bins, thereby
manipulating the different types of social norms they signal, can increase the amount of trash they collect and
mitigate littering. A eld experiment tested the effectiveness of two alternatives to the conventional grey street
bin; one bin foliated with pictures drawn by school children containing a normative anti-littering message
(explicit norm), and one bright orange salient bin (implicit norm). Observed behavioral data was collected, and
both the weight and volume of trash in the bins were measured each day for a period of one month. The results
showed a tendency for the salient orange bin to increase trash collection compared to other bins; an effect most
tangible towards the end of the weeks. The biggest effect was, however, that the explicitly normative bin reduced
trash collection overall. These results provide lessons on how the appearance of bins can inuence trash
collection, potentially resulting in both desirable and undesirable outcomes.
1. Introduction
A powerful way to steer behavior toward desirable outcomes is
through the design and manipulation of the physical environment
(Scott, 2005). Recent research highlights the potential of changing the
physical environment (viz. the immediate decision situation) to
encourage sustainable actions, such as promoting recycling (Linder
et al., 2021; Rosenthal and Linder, 2019), limiting food waste (Kall-
bekken and Sælen, 2013), or promoting biking over car use (Kaaronen
and Strelkovskii, 2020a). Such approaches could complement the more
conventionally used strategies of communication, education, and
normative information that are often implemented to foster sustainable
behavior changes (S¨
orqvist, 2016). Still, exploring how changes in the
physical environment can be used to promote pro-environmental
behavior remains largely understudied (Clayton et al., 2016; G¨
arling,
2014; Linder, 2022; S¨
orqvist, 2016; Steg and Vlek, 2009).
The current study aims to address this research gap by focusing on
how changes in the urban environment can be leveraged to prevent
littering. Littered environments are often considered both unsightly and
antisocial and communities across the world spend a great deal of re-
sources on prevention measures. More importantly, accumulated and
misplaced trash can be harmful at larger scales. For example, fossil-fuel-
derived plastics take centuries to degrade, meaning that today, the
oceans, soil, and air contain billions of tons of increasingly fragmented
(and toxic) pieces of plastic (MacLeod et al., 2021). To prevent littering,
city planners tend to place trash bins throughout the city landscape at
strategic points (with alternating concentration). General guidelines for
city planners are to make street bins blend into the environment. While
this approach may be logical in terms of architectural and aesthetic
considerations, it may not fully maximize the bins’ intended purpose: to
promote usage and reduce littering. This study asks the following
question: How does the appearance of street bins affect the amount of
trash they collect, and can they be manipulated to become more effec-
tive at collecting trash?
* Corresponding author at: Global Economic Dynamics and the Biosphere, The Royal Swedish Academy of Science, Lilla Frescativ¨
agen 4A, SE-104 05 Stockholm,
Sweden.
E-mail address: noah.linder@kva.se (N. Linder).
Contents lists available at ScienceDirect
Resources, Conservation & Recycling Advances
journal homepage: www.sciencedirect.com/journal/
Resources-Conservation-and-Recycling-Advances
https://doi.org/10.1016/j.rcradv.2023.200186
Resources, Conservation & Recycling Advances 19 (2023) 200186
2
2. Theoretical background
2.1. Social norms
A promising way to motivate behaviors, or behavior change, is to
make people aware of a social norm in a situation that aligns with the
desired behavior (Nyborg et al., 2016). Social norms can be seen as the
predominant behavioral pattern within a group, supported by a shared
understanding of acceptable actions (Nyborg et al., 2016). Social norms
are often built on the foundation of shared values which can be regarded
as general guiding principles for actions and attitudes (Axelrod, 1986).
These values also assist in evaluations and judgments, signaling what is
important, or what is deemed good and bad for individuals and collec-
tives (Rokeach, 1968). Social norms can be an important motivator for
behavior, and have been highlighted as an especially promising
approach to help address sustainability challenges in their capacity to
instigate pro-environmental behavioral change (Bergquist et al., 2019;
Chapin et al., 2022; Farrow et al., 2017; Nyborg et al., 2016). For
example, normative messages can be persuasive and have been exten-
sively used to promote different sustainable behaviors in different do-
mains, such as energy, water savings, and recycling (Allcott, 2011;
Ferraro et al., 2011; Linder et al., 2018; Miranda et al., 2020). Note
however, that the focus theory of normative conduct (Cialdini et al.,
1990) suggests that human behavior is both systematically and signi-
cantly affected by norms — but only when these norms are made salient
to the individual.
2.2. Explicit and implicit norms
Social norms can be utilized to foster sustainable behavior, but it
matters in which way they are communicated. For example, a recent
meta-analysis showed that social norms tend to have a greater impact on
behavior when they are implicitly emphasized through cues in the
environment, in comparison with when norms are stated explicitly, e.g.,
through communication messages (Bergquist et al., 2019). The authors
argue that this effect likely arises because norms conveyed by the
environment often lead to automatic behavioral responses, that occur
with minimal conscious deliberation (Bergquist et al., 2019). Conse-
quently, these implicit norms are less prone to evoke anti-conformity
responses, such as psychological reactance, i.e., the resistance or de-
ance that can occur when an individual perceives that their freedom or
choices are being threatened (Brehm, 1966). Such results highlight the
potential of changing behavior through the manipulation of the physical
environment so that it conveys implicitly stated norms. One way of
creating such normative environments could be by making the desired
behavior more salient so that it is clear what behavior is expect-
ed/promoted in a certain situation. If successful, those environments
would signal the implicit norm that most people behave accordingly.
Furthermore, salient objects in the environment might increase the
chance of choosing one option over another because it is better at
grabbing people’s attention (Milosavljevic et al., 2012). Saliency is a
property by which an object, person, or event stands out from the sur-
rounding context and can be a result of the physical intensity and
characteristics of the object (e.g., color; Itti and Koch, 2001) but also a
result of emotional or cognitive associations made by the individual who
perceives the object (Wood and Cowan, 1995). The tendency for a sa-
liency bias in choices (i.e., the tendency to select the more salient op-
tions) has been utilized in many areas to inuence decisions (e.g.,
(Castelo et al., 2015; Chetty et al., 2009), wherein perhaps the most
prominent is within marketing design (Milosavljevic et al., 2012).
Hence, one idea of conveying implicit norms is by making the desired
behavior more salient in the decision situation. There is some research in
support of this suggestion; for example, it has been shown that installing
charging stations for electric vehicles at greater density in a neighbor-
hood increases intentions to buy electric cars amongst residents in the
area, and that this association is mainly due to a perceived increase in
social norms in support of electric vehicles amongst the neighbors
(White et al., 2022). Another example is the biking revolution that
occurred in Copenhagen over the last decades; this shift in behavior,
from a car-centric culture to one embracing cycling, is mainly attributed
to physical changes in the urban landscape (Kaaronen and Strelkovskii,
2020b). These modications involve the allocation of greater road space
for dedicated bike lanes, increased safety measures for bikers, and the
use of colored roads to demarcate bike lanes: all changes that collec-
tively contribute to creating an environment that is signaling an implicit
pro-cycling social norm.
2.3. Aims and hypotheses
The overarching aim of the current study was to test if it is possible to
make trash bins collect more trash by manipulating their physical
appearance. An additional aim of the study was to contribute to the
understanding of how social norms inuence decisions, by exploring
how trash bins designed to signal different social norms (explicit or
implicit norms) inuence waste behaviors. At the time of data collec-
tion, the conventional street bins at the city center of G¨
avle (where the
data collection was conducted) were grey and arguably made to blend in
with the surrounding environment (Fig. 1, panel A). And the research
area had an average litter rate of about 0.75 misplaced trash (excluding
cigarettes and chewing gum) per 10 m
2
s, according to a report from the
previous year (Håll Sverige Rent, 2016). Based on the theory presented
above, two alternative types of bins to the conventional street bins were
designed. One type of bin, henceforth called the explicitly normative
bin, was covered in drawings that stated an explicit pro-environmental
norm (Fig. 1, panel B). Another type of bin, henceforth called the
implicitly normative bin, was foliated with a bright orange color (Fig. 1,
panel C). To explore the effectiveness of these bin manipulations, a eld
experiment (Harrison and List, 2004), was conducted, in the city center
of G¨
avle, Sweden.
Grounded in the insights gleaned from previous literature presented
above, the study seeks to test the following hypotheses:
H1: The physical appearance of street bins affects the amount of trash
they collect.
H2: Both the implicitly normative bins and the explicitly normative
bins will collect more trash than the conventional control bin.
3. Method
3.1. Design and procedure
The design of the explicitly normative bin was done in collaboration
with a local school. Children painted pictures with an anti-littering
message “e.g., do like me, it is important to put your trash in the bin”
signed by the kids’ name and the name of the school (see example in
Fig. 1, panel B). Different versions of the kids’ drawings and messages
were used under the same theme. The idea was to create a trash bin that
explicitly stated the local social norm of caring for the environment and
the importance of reducing litter, as well as activating a pro-
environmental norm.
The bin with an implicit normative appeal was foliated to be bright
orange, aiming to make it more salient and stand out from the sur-
rounding context. The intent was to draw attention to the bin, clearly
indicating the desired behavior within the current context without
relying on explicit normative text, symbols, or signs. See Fig. 2 for the
differences in saliency between the orange bins and the conventional
bins (panel A compared to panel B).
3.2. Manipulation check
To test if the manipulation of the implicitly normative bins was
successful, a post-hoc manipulation check was conducted using a
N. Linder et al.
Resources, Conservation & Recycling Advances 19 (2023) 200186
3
modied version of the implicit association test (Greenwald et al.,
1998). A convenient sample of 31 non-informed colleagues observed a
series of 5 different pictures on a computer, with and without trash bins.
The rst picture was a generic picture of central park, with a path and no
bin (this picture was always shown rst and acted as a warmup trial that
was later discarded from the analysis), the following pictures were a
series of photoshopped versions of the same picture of the street, con-
taining the different versions of the bins. And one last picture showing
the street without any bins (See Fig. 3). The order of the pictures was
alternated between participants. In response to each picture, partici-
pants were asked to identify if there were any trash bins in the picture, as
fast and as accurately as possible. The reaction time was compared
across the different versions of the bins.
The results showed that participants were fastest at identifying the
orange implicitly normative bin, followed by the explicitly normative
bin, and slowest to identify the conventional control bins (See Fig. 4).
A paired samples t-test was conducted to further explore the differ-
ence in reaction time between conditions. When the reaction time for the
control bin (M =1.49 s, SD =1.03) and for the orange bin (M =1.16 s,
SD =0.78) were compared, the mean difference was 328.1 ms (SD =
745.3). This difference was statistically signicant, t(30) =2.45, p =
.020, 95 % CI [54.75, 601.51]. That is, the participants were signi-
cantly faster to identify a trash bin in the street when the trash bin
wasorange compared to the conventional color. There was no signicant
differences between the other pairs. Categorization accuracy was also
analyzed. The participants had no problem correctly identifying
whether there was a bin in the street or not. Only 6 times (out of a total
of 93 times a participant saw one of the 3 versions of the bins) did
participants mistakenly say that there were “no trash bins in the picture”
when the picture in fact contained a bin. Three times when the picture
included a control bin, one time for the explicit bin, and two times for
the implicit bin. These results suggest that people have no problem
Fig. 1. The three different manipulations of street bins: a conventional control bin (A), the explicit normative bin (B), and the implicit normative bin (C).
Fig. 2. The pictures show the same bin area with two different manipulations of the bins: the implicitly normative bin (panel A) and the conventional control bin
(Panel B).
Fig. 3. Shows the different pictures in the implicit association test.
N. Linder et al.
Resources, Conservation & Recycling Advances 19 (2023) 200186
4
identifying the trash bins as indeed being trash bins, even when bins
don’t have the standard grey visual appearance. Moreover, the faster
response times suggest that orange color facilitates detection of the trash
bins, arguably by being better at capturing attention.
3.3. Field experiment
In total 9 bin-positions on a walking street in G¨
avle, Sweden, were
located and included in the study. The street bins that occupied these
positions alternated between being manipulated (according to the two
treatments), and not (i.e., being in the control group) during the
experiment period. The experiment took place in the year 2017 from the
14th of August to the 10th of September. It was conducted on the street
Nygatan which was divided into 3 areas that contained a cluster of 3 bin-
positions in each. These areas were separated by roughly 300 m along
the street to avoid spill-over effects. In these “bin areas,” three trash bins
in close proximity had the same manipulation for one week each (see
Fig. 2, for an example of an area with three bins), after which they
alternated between the different bin manipulations, following the
experimental design presented in Table 1.
The experiment ran for 4 weeks, the rst week served as a baseline
measure when no bin was manipulated. Data on both the weight and
estimated volume of trash collected was measured by the end of each
day for all 9 bins as the experiments’ dependent variables. To measure
weight, a research assistant removed the bag from the bin, attached it to
a scale, and recorded the weight (in kilograms). For volume, the assis-
tant used a measuring stick placed on the highest trash in the bin and
estimated the total volume of trash in the bin (from 0 to 100 %).
3.4. Data
After observing data from the 9 bins, every day for 4 weeks,
measuring both volume and weight, the data set consisted of 504 (9 ×7
×4 ×2) individual weight and volume measurements from the bins.
There were 6 missing values, which were replaced by the mean weight/
volume collected by the same bin for the week in question, but there was
never more than one missing value for a single bin/week. In total, the
trash bins collectively collected 270 kg of litter which amounted to
approximately 75 full bins.
4. Results
4.1. Baseline measure
The baseline measure served two main purposes: to explore pre-
existing differences between the areas and to look at potential time
trends of the collection over the week. An initial observation was that
people used the trash bins more towards the weekend in comparison
with the rst days of the week (Fig. 5), a trend that lasted throughout the
experiment (see Fig. 6). Furthermore, the baseline measure indicated
that areas 2 and 3 were strikingly similar in how much trash their bins
collected; in total, area 2 collected 30.8 kg of trash, and area 3 collected
31.3 kg during the baseline week, and they collected approximately 7
full trash bins each in volume. Area 1, however, collected less trash on
average, with a total of 24.4 kg and about 5 full bins in volume.
Note that these pre-existing differences were controlled for by the
experimental design (where bin order and area were counterbalanced,
Table 1), as well as by including the time order of the manipulations as a
variable in the nal analysis.
4.2. Experimental period
To test our hypotheses, we analyzed how the physical appearance of
street bins affects the amount of trash they collect. There was no
discernible difference between the trash bins during the rst weekdays,
in terms of how much trash they collected. However, during the latter
part of the week, when the trash bins were used more often, there was a
systematic difference between the different bins. In terms of descriptive
statistics, the implicitly normative trash bins were used more than both
the explicit ones and the conventional grey trash bins. This difference
was observed both for the weight measure (Fig. 6, Panel A) and the
volume measure (Fig. 6, Panel B). The amount of trash that was tossed in
the implicitly normative (orange) trash bins was generally higher during
the end of the week while the explicitly normative trash bins, in turn,
were used systematically less than the competing bins, especially during
the end of the week, suggesting that bin use behavior was better induced
by the attention-grabbing orange trash bins while it was steered away
from the explicitly normative ones.
These conclusions were supported by a within-between mixed 3
(Trash bins: “control”, “implicitly normative”, and “explicitly norma-
tive”) ×7(days of the week) ×3(condition order) multivariate analysis
of variance with trash weight and trash volume as dependent variables.
The multivariate analysis revealed a signicant effect of trash bin,
Wilks’ Lambda =0.35, F(4, 22) =3.78, p =0.017,
η
p
2
=0.41, and a
signicant effect of day, Wilks’ Lambda =0.37, F(12, 70) =3.70, p <
0.001,
η
p
2
=0.39. The main effect of condition order was not signicant,
F ≈0.5, for either dependent variable. The interaction between trash bin
and days of the week was signicant for the weight measure, F(12, 72)
=1.94, p =0.043,
η
p
2
=0.24, but not for the volume measure, F(12, 72)
=1.16, p =0.329,
η
p
2
=0.16. Thus, we conclude that Hypothesis 1 is
supported by the data.
The tendency to use the bins more during the weekend, and the in-
teractions between bin and day of the week, motivated a closer look at
the differences between the bins at the weekend (day 5–7). The data for
weekday 5–7 were collapsed into an average for both dependent vari-
ables, respectively. Thereafter, pairwise comparisons with t-tests were
conducted to test the differences between the bins, at the weekend.
These analyses revealed a difference between the implicitly normative
bins and the explicitly normative bins (M
diff
=1.12, SE =0.38), t(8) =
2.98, p =0.018, 95 % CI [0.25–1.99], such that the implicitly normative
bin collected trash with more weight. Similarly, the implicitly normative
bins were, on average, more lled than the explicitly normative bins
(M
diff
=18.70, SE =7.88), t(8) =2.37, p =0.045, 95 % CI [0.53–36.87].
Differences in the same direction were found for the comparison be-
tween the control bins and the explicitly normative bins. The control
bins collected more weight (M
diff
=0.64, SE =0.24), t(8) =2.68, p =
Fig. 4. Shows the time it took (on average) for participants to indicate whether
there was a trash bin in the picture across different manipulations.
Table 1
The table shows the order of manipulation in each area of the street.
Week 1 (baseline) Week 2 Week 3 Week 4
Bin area 1 Control Implicit Control Explicit
Bin area 2 Control Explicit Implicit Control
Bin area 3 Control Control Explicit Implicit
N. Linder et al.
Resources, Conservation & Recycling Advances 19 (2023) 200186
5
0.028, 95 % CI [0.09–1.20], and were more lled (M
diff
=18.37, SE =
5.32), t(8) =3.25, p =0.012, 95 % CI [5.32–31.42], than the explicitly
normative bins. There were, in turn, no differences between the
implicitly normative bins and the control bins, either for weight (M
diff
=
0.48, SE =0.41), t(8) =1.17, p =0.27, 95 % CI [-1.41-0.45], or for
volume (M
diff
=0.33, SE =9.25), t(8) =0.04, p =0.972, 95 % CI
[-21.66-20.66]. While the mean difference was in the expected direc-
tion, the difference could not reach statistical signicance, probably due
to low statistical power. Thus, we conclude that Hypothesis 2 is not
supported by the data. While part of the evidence, and in particular the
descriptive statistics, points toward the conclusion that the implicitly
normative trash bin collects more trash than the control trash bin, this
was not the case for the explicitly normative trash bin.
5. Discussion
What is clear from these results is that the appearance of the bin
signicantly inuenced how well it encouraged bin use, supporting
Hypothesis 1. The biggest effect seen, however, was that the explicitly
normative bins foliated with kids’ drawings and an environmental
message reduced trash collection overall. This unexpected outcome
disconrms Hypothesis 2 and highlights the need for testing in-
terventions, in eld settings, before large-scale implementation (Linder,
2022; McKenzie-Mohr, 2013; Steg and Vlek, 2009). Without proper
prior testing, well-intended interventions may end up worsening a sit-
uation instead of solving the problem they sought to solve.
One possible explanation for why the bins using an explicit norma-
tive appeal reduced rather than increased bin use is because they
invoked a reactance effect toward the normative messages. Some people
without an “environmental identity” might experience an adverse re-
action to the explicitly normative message on the bin, something a more
neutral bin, such as the control and the implicitly normative (salient)
bin, would avoid. On the other hand, it is also possible that some pe-
destrians appreciated the bins and didn’t want to “soil them” with their
litter. Lastly, it is possible that the appearance of the explicitly norma-
tive bins was changed too much, so that pedestrians no longer under-
stood the trash bins were, in fact, trash bins, but perhaps instead
associated them with an advertising campaign or something along that
line. However, the results from the manipulation check reported in the
methods section of the paper contradict this last potential explanation,
since the manipulation check indicated that participants seem to have
no problem identifying the explicitly normative bins as trash bins. The
implicitly normative bins, on the other hand, showed some trends of
increasing trash collection and were signicantly more effective than
the explicitly normative bins at doing so. In view of the results from the
manipulation check, one possible explanation of this effect is that the
orange (implicitly normative) bins attract attention. These results give
support to studies on how saliency inuences action (e.g., Itti and Koch,
2001; Milosavljevic et al., 2012) and indicate that saliency can be uti-
lized in attempts to promote pro-environmental behavior. The orange
bins were best at capturing the pedestrians’ attention and perhaps also
at reminding people to disperse their trash.
More research is needed to untangle the mechanisms underpinning
the poor performance of the explicitly normative bin and the potential
success of the implicitly normative orange bin, to adjudicate these po-
tential explanations. From the current study, it is impossible to draw
strong conclusions about the use of the two different norms. The results
do, however, give some promising indications that signaling implicit
Fig. 5. The gure shows the mean amount of trash, measured in kilograms (Panel A) and in volume/percentage lled (Panel B), that was thrown in the conventional
grey trash bins across the seven days of the week before the experimental period in the different bin areas on the street.
Fig. 6. The gure shows the mean amount of trash, measured in kilograms (Panel A) and in volume/percentage lled (Panel B), that was thrown in various trash bins
across the seven days of the week. During the experimental period of three weeks, a third of the trash bins were implicitly normative (orange), a third were control
(grey/regular) and a third were explicitly normative (white with normative drawings pasted on them), with positions counterbalanced across the three measure-
ment weeks.
N. Linder et al.
Resources, Conservation & Recycling Advances 19 (2023) 200186
6
social norms through changes in the physical landscape could be a tool
that can be both effective and help avoid potential reactance effects. We
encourage more research exploring similar interventions.
The differences between bins were greatest at the end of the week.
This most likely reects the fact that more people moved around in the
town center during the weekends. This also aligns with the baseline
measure, demonstrating a similar trend of more bin use towards the
weekend. It could be the case that differences between bins—in terms of
their capacity to attract bin use — are only noticeable during periods of
high usage. The reason the bins did not differ substantially during the
weekdays could hence partly stem from a oor effect, as some bins
collected very little trash during the weekdays. Of course, there is also a
possibility that different days of the week are associated with a different
type of behavior. For example, trash collection during the weekend
could be inuenced by more food consumption or increased rates of
intoxicated people moving around within the city center.
5.1. Practical implications
The result of this study mainly underscores the signicance of con-
ducting eld testing before implementing interventions on a large scale.
It showcases how even well-intentioned interventions can inadvertently
worsen a situation if not rigorously evaluated in real-world settings. The
study makes it clear that waste managers and city planners should pay
close attention to the external design of trash bins if their goal is to
maximize bin effectiveness, and especially consider designs that make
street bins stand out from the surrounding context rather than blending
in — contrary to the standard praxis of today. The study also indicates
that street bins with designs containing explicitly normative messages
should perhaps be avoided, or their effects should at least be extensively
tested before large-scale implementation, as they might be more prone
to backre. More research exploring when and why explicitly stated
norms work (and don’t work) to manage waste behavior is advised.
5.2. Theoretical implications
The study contributes to the understanding of social norm theory by
showcasing the differential impact of explicit and implicit normative
appeals in inuencing behavior. It supports previous research suggest-
ing that implicit social norms prompted by the environment can be more
effective than explicit messages in certain situations (Bergquist et al.,
2019). When the bins were orange, this circumstance might have
created an urban environment that more clearly conveyed behavioral
expectancies, which could inuence behavior through implicit social
norms without risking an adverse reaction towards the
pro-environmental message (as well as better prompting bin use).
Furthermore, these ndings indicate that reactance theory may play a
role in how individuals respond to explicit normative appeals. For
example, it is possible that normative environmental messages on bins
could trigger a reactance effect among individuals who do not strongly
identify with environmental concerns, which would underscore the need
to consider individual differences and potential reactance effects when
designing interventions. More research is needed to conrm this po-
tential explanation of the backlash effect. Finally, the study aligns with
previous research on saliency and attention by demonstrating that the
visual saliency of bins, particularly the use of the color orange, is better
at capturing people’s attention and potentially inuencing their
behavior, supporting the idea that salient cues in the environment can be
used effectively to prompt specic actions.
5.3. Limitations
One obvious limitation concerns sample size. Although four weeks of
data collection is a rather long time for a eld experiment in a public
area, one measurement a day results in a relatively small number of data
points. It is, for example, possible that a longer data collection, or
inclusion of more bins, would paint a different picture, and differences
between bins would be more noticeable. The fact that the largest dif-
ferences between bins were found at times of high bin usage points in
this direction. Furthermore, there is a multitude of potential con-
founding variables that could have inuenced the different areas on the
street disproportionately. These limitations are common in eld- and
quasi-experimental designs (Campbell and Stanley, 2015), and are hard
to avoid. However, since the conditions were counterbalanced, the effect
of uncontrolled variables should be limited.
5.4. Conclusions
The results reported in this study highlight how seemingly small
changes in the urban environment can have disproportional effects on
citizens’ behaviors. Considering both the surprising backlash effect from
the explicitly normative bin, and the promising effect of the implicitly
normative bin, this study provides some valuable lessons on what not to
do and creates concrete suggestions for how to improve the design of
urban trash bins with the goal of mitigating littering.
CRediT authorship contribution statement
Noah Linder: Conceptualization, Methodology, Investigation, Vali-
dation, Writing – original draft, Visualization, Writing – review &
editing. Patrik S¨
orqvist: Formal analysis, Software, Writing – original
draft, Visualization, Writing – review & editing, Resources. Therese
Lindahl: Writing – original draft, Writing – review & editing, Concep-
tualization. Robert Ljung: Conceptualization, Writing – original draft,
Methodology, Supervision.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgments
We would like to acknowledge “G¨
astrike återvinnare" (the municipal
body responsible for collection, treatment, planning, and information
regarding household waste in G¨
avle) for making this study possible,
taking this project on board and providing us with the tools we needed to
realize this study. Linder’s work is enabled by the University of G¨
avle
and with support from FAIRTRANS, a program that aims to promote a
fair transformation to a fossil-free future, which is nanced by FORMAS
and MISTRA.
References
Allcott, H., 2011. Social norms and energy conservation. J. Public Econ. 95 (9–10),
1082–1095. https://doi.org/10.1016/j.jpubeco.2011.03.003.
Bergquist, M., Nilsson, A., Schultz, W.P., 2019. A meta-analysis of eld-experiments
using social norms to promote pro-environmental behaviors. Glob. Environ. Change
59, 101941.
Brehm, J.W. (1966). A theory of psychological reactance.
Campbell, D.T., Stanley, J.C., 2015. Experimental and Quasi-Experimental Designs For
Research. Ravenio books.
Castelo, N., Hardy, E., House, J., Mazar, N., Tsai, C., Zhao, M., 2015. Moving citizens
online: using salience & message framing to motivate behavior change. Behav. Sci.
Policy 1 (2), 57–68.
Chapin, F.S., Weber, E.U., Bennett, E.M., Biggs, R., van den Bergh, J., Adger, W.N.,
Cr´
epin, A.-S., Polasky, S., Folke, C., Scheffer, M., 2022. Earth stewardship: shaping a
sustainable future through interacting policy and norm shifts. Ambio 1–14.
N. Linder et al.
Resources, Conservation & Recycling Advances 19 (2023) 200186
7
Chetty, R., Looney, A., Kroft, K., 2009. Salience and taxation: theory and evidence. Am.
Econ. Rev. 99 (4), 1145–1177.
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. J. Pers. Soc.
Psychol. 58 (6), 1015–1026. https://doi.org/10.1037/0022-3514.58.6.1015.
Clayton, S., Devine-Wright, P., Swim, J., Bonnes, M., Steg, L., Whitmarsh, L., Carrico, A.,
2016. Expanding the role for psychology in addressing environmental challenges.
Am. Psychol. 71 (3), 199.
Farrow, K., Grolleau, G., Ibanez, L., 2017. Social norms and pro-environmental behavior:
a review of the evidence. Ecol. Econ. 140, 1–13.
Ferraro, P.J., Miranda, J.J., Price, M.K., 2011. The persistence of treatment effects with
norm-based policy instruments: evidence from a randomized environmental policy
experiment. Am. Econ. Rev. 101 (3), 318–322.
G¨
arling, T., 2014. Past and present environmental psychology. Eur. Psychol. 19 (2),
127–131. https://doi.org/10.1027/1016-9040/a000184.
Greenwald, A.G., McGhee, D.E., Schwartz, J.L.K., 1998. Measuring individual differences
in implicit cognition: the implicit association test. J. Pers. Soc. Psychol. 74 (6), 1464.
Håll Sverige Rent, HSR (2016). Skr¨
apfacit 2016 G¨
avle. Statistiska centralbyrån. https://ol
d.gavle.se/PageFiles/4125/Skr%C3%A4pFacit_G%C3%A4vle_2016.pdf.
Harrison, G.W., List, J.A., 2004. Field experiments. J. Econ. Lit. 42, 1009–1055. https://
doi.org/10.1257/0022051043004577.
Itti, L., Koch, C., 2001. Computational modelling of visual attention. Nature Reviews
Neuroscience 2 (3), 194–203.
Kaaronen, R.O., Strelkovskii, N., 2020a. Cultural evolution of sustainable behaviors: pro-
environmental tipping points in an agent-based model. One Earth 2 (1), 85–97.
Kaaronen, R.O., Strelkovskii, N., 2020b. Cultural evolution of sustainable behaviors: pro-
environmental tipping points in an agent-based model. One Earth 2 (1), 85–97.
Kallbekken, S., Sælen, H., 2013. Nudging” hotel guests to reduce food waste as a win-win
environmental measure. Econ. Lett. 119 (3), 325–327. https://doi.org/10.1016/j.
econlet.2013.03.019.
Linder, N., 2022. Make It Easier: A psychological Perspective On Sustainable Behavior
Change. G¨
avle University Press.
Linder, N., Lindahl, T., & Borgstr¨
om, S. (2018). Using behavioural insights to promote food
waste recycling in urban households — evidence from a longitudinal eld experiment. 9
(March), 1–13. 10.3389/fpsyg.2018.00352.
Linder, N., Rosenthal, S., S¨
orqvist, P., Barthel, S., 2021. Internal and external factors’
inuence on recycling: insights from a laboratory experiment with observed
behavior. Front. Psychol. 12.
MacLeod, M., Arp, H.P.H., Tekman, M.B., Jahnke, A., 2021. The global threat from
plastic pollution. Science 373 (6550), 61–65.
McKenzie-Mohr, D., 2013. Fostering Sustainable behavior: An introduction to
Community-Based Social Marketing. New society publishers.
Milosavljevic, M., Navalpakkam, V., Koch, C., Rangel, A., 2012. Relative visual saliency
differences induce sizable bias in consumer choice. J. Consum. Psychol. 22 (1),
67–74.
Miranda, J.J., Datta, S., Zoratto, L., 2020. Saving water with a nudge (or two): evidence
from Costa Rica on the effectiveness and limits of low-cost behavioral interventions
on water use. World Bank Econ. Rev. 34 (2), 444–463.
Nyborg, K., Anderies, J.M., Dannenberg, A., Lindahl, T., Schill, C., Schlüter, M.,
Adger, W.N., Arrow, K.J., Barrett, S., Carpenter, S., 2016. Social norms as solutions.
Science 354 (6308), 42–43.
Rosenthal, S., & Linder, N. (2019). Effects of convenience and procedural prompts to
increase the use of recycling bins and reduce the contamination of recyclables:
evidence from a laboratory experiment in Singapore. Manuscript Submitted for
Publication.
Scott, M.M., 2005. A powerful theory and a paradox: ecological psychologists after
Barker. Environ. Behav. 37 (3), 295–329.
S¨
orqvist, P., 2016. Grand challenges in environmental psychology. Front. Psychol. 7.
Steg, L., Vlek, C., 2009. Encouraging pro-environmental behaviour: an integrative review
and research agenda. J. Environ. Psychol. 29 (3), 309–317. https://doi.org/
10.1016/j.jenvp.2008.10.004.
White, L.V., Carrel, A.L., Shi, W., Sintov, N.D., 2022. Why are charging stations
associated with electric vehicle adoption? Untangling effects in three United States
metropolitan areas. Energy Res. Soc. Sci. 89, 102663.
Wood, N., Cowan, N., 1995. The cocktail party phenomenon revisited: how frequent are
attention shifts to one’s name in an irrelevant auditory channel? J. Experim. Psychol.
Learn. Memory Cogn. 21 (1), 255.
N. Linder et al.
