Final Report The Effectiveness of Providing Labels and other Pre-Purchase Factual Information in encouraging more Environmentally Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment Project code: POS013 Research date: May 2018 to January 2019 Date: February 2019
78
Embed
The Effectiveness of Providing Labels and other Pre ... Sust Product... · The Effectiveness of Providing Labels and other Pre-Purchase Factual Information in encouraging more Environmentally
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Final Report
The Effectiveness of Providing Labels and other Pre-Purchase Factual
Information in encouraging more Environmentally Sustainable
Product Purchase Decisions: Expert Interviews and a Rapid Evidence
Assessment
Project code: POS013
Research date: May 2018 to January 2019 Date: February 2019
WRAP’s vision is a world in which
resources are used sustainably.
Our mission is to accelerate the move to
a sustainable resource-efficient economy
through re-inventing how we design,
produce and sell products; re-thinking
how we use and consume products; and
re-defining what is possible through re-
use and recycling.
Find out more at www.wrap.org.uk
Document reference: WRAP, 2019, The Effectiveness of Providing Pre-Purchase Factual Information in encouraging
more Environmentally Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment
Prepared by Dr. Colin Whittle, Fiona Brocklehurst, Catriona McAlister & Prof. Lorraine Whitmarsh
Written by: Dr. Colin Whittle, Cardiff University; Fiona Brocklehurst, Ballarat Consulting;
Catriona McAlister, Sea Green Tree; & Prof. Lorraine Whitmarsh, Cardiff University
Quality Control by: Christian Reynolds, WRAP
Front cover photography:
While we have taken reasonable steps to ensure this report is accurate, WRAP does not accept liability for any loss, damage, cost or expense incurred or arising from
reliance on this report. Readers are responsible for assessing the accuracy and conclusions of the content of this report. Quotations and case studies have been
drawn from the public domain, with permissions sought where practicable. This report does not represent endorsement of the examples used and has not been
endorsed by the organisations and individuals featured within it. This material is subject to copyright. You can copy it free of charge and may use excerpts from it
provided they are not used in a misleading context and you must identify the source of the material and acknowledge WRAP’s copyright. You must not use this report
or material from it to endorse or suggest WRAP has endorsed a commercial product or service. For more details please see WRAP ’s terms and conditions on our
◼ Prof. Lorraine Whitmarsh, Department of Psychology Cardiff University, Project
Director & Reviewer (Environmental Psychology)
The work was guided by a steering group formed of representatives of Defra and WRAP
whose members were:
◼ Jacks Guiness (initially, then Nathan Simmonds), Project manager, WRAP
◼ Keith James, WRAP
◼ Christian Reynolds, WRAP; Overall Assurer
◼ Debs Reynolds, Defra
◼ Leila McElvenney, Defra
◼ Ladislav Tvaruzek, Defra
◼ Maya De Souza, Defra
◼ Matt Stocks, Defra
The rest of this report describes the work as follow:
Section 2 – Methodology
Section 3 – Results
Section 4 – Implications
6 Collins, A.M., Coughlin, D., Miller, J., Kirk, S. (2015) The Production of Quick Scoping Reviews and Rapid Evidence Assessments: A
How to Guide
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 8
2.0 Methodology
2.1 Development of Rapid Evidence Assessment protocol
The REA protocol was developed in association with the Defra/WRAP steering group and
in line with the Defra/NERC guidance on evidence reviews7. It was designed to provide a
systematic and objective survey of the evidence, to be transparent and to minimise bias.
The initial protocol proposed by the project team were refined in the light of feedback
from the steering group, suggestions by the experts interviewed and trial database
literature searches.
Specifically, keywords for the literature search were developed based on the PICO8
elements of the primary research question. Synonyms, antonyms and conceptually
similar terms for these components were determined using the research team’s
knowledge and guidance from the experts who were interviewed as a component of the
first part of the project. Likewise, a list of existing environmental sustainability labels
(individually searched for) were generated through the same processes. The keywords
were developed in an iterative process of trialling and refining by the research team such
that the results returned from the search were optimised to provide a comprehensive
overview of the available evidence. The draft list was then reviewed by the steering group.
Inclusion and exclusion criteria for evidence were determined by the research team, in
conjunction with the Steering Group, based on the PICO components of the primary
research question and the initial trialling and refining of search terms. For example, it
was decided to include research published between the year 2000 and current date, with
no geographic restrictions as long as the publication was in English; research on market
impacts (as against direct consumer response) were excluded. The full details are in
Appendix 1.
The draft protocol, as formally agreed with the steering group is given in Appendix 1.
2.2 Overview of the REA protocol
The REA protocol consisted of several stages:
1. Evidence search
2. Search using agreed terms search academic databases Scopus and Web of
Science and the internet (using an online search engine; call for evidence from
the steering group, expert interviewees (see below) and wider stakeholders.
The searches took place in in September 2018.
3. Screening search results
Screen search results for relevance to research topic undertaken in two stages:
firstly, on the paper or report title; secondly by the content of the abstract or
executive summary
7 Ibid.
8 Population, Impact, Comparator and Outcome
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 9
4. Extracting the evidence
Extract evidence extracted from the full paper or report text using a standard
form which captured: the details of the source; the nature of the study (eg
quantitative observation – such as a survey or a quantitative experiment such as
a choice experiment, the population studied) and the results (in terms of the
REA’s primary and secondary questions).
Rate each piece of selected evidence for robustness and relevance against
pre-determined criteria in the protocol and combine the scores to give an overall
confidence categorisation.
(Some further evidence was rejected at this stage as being irrelevant, based on
the full text or if found to be a duplicate.)
5. Synthesising the evidence
Review the evidence to answer the research questions and report on the
adequacy of the evidence base: describe the volume and characteristics of the
evidence base; describe what the evidence indicates; indicate the implications.
(Some further evidence was excluded at this stage , based on a more detailed
reading of the full text or if found to be a duplicate.)
These stages are described in detail in the protocol in the Appendix 1 and the results are
given below.
The protocol was refined in the course of its application. For example, more exclusions
were added to the search terms and some suggested search keywords were found to be
too broadly used (for example “bio”) and had to be excluded. A description of the
adaptations of the protocol are given in Appendix 2.
The greatest challenge in this stage of the process was making the search for ‘grey’
literature systematic; some sources were identified via the steering group, and a wider
group of stakeholders (via interviews and a call for evidence – see below). However, it was
essential for the completeness and objectivity of the REA that a systematic search was
made via an online search engine. It was found that the search syntax used in the search
of academic databases was not applicable in online search engines - this had to be
adapted. Professional librarians and experienced stakeholders were consulted as to how
best to do this.
2.3 Expert interviews
The interviews were intended to overcome the risk of publication bias by identifying:
◼ results from studies which do not find effects or impacts, and therefore are less likely
to be published; and
◼ recently completed or ongoing research
They also offered the opportunity to get additional suggestions on the development of
the protocol.
Interviewees were identified from amongst the research team and steering group’s
contacts to represent experts in the field of consumer behaviour, labelling, resource
efficiency and sustainability and reflecting a balance of academics (including different
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 10
disciplines), industry, non-governmental and policy representatives. In total, ten
interviewees agreed to participate.
Interviews took place by telephone, lasting a maximum of one hour. They were conducted
using a set of questions (Appendix 3) derived from the research objectives. These included
questions about the interviewees’ own work or expertise in sustainable information
provision (e.g. labelling), suggestions for literature that should be included in the review
and interviewees’ views on and responses to the primary and secondary research
questions. Answers were recorded by interviewers who took detailed notes during the call
on a standardised template.
Interviewees’ suggestions for literature were included in the main REA (screening,
extraction and synthesis). Interviewees’ views relating to the research questions are
presented in Appendix 4.
2.4 Call for evidence
The experts interviewed reinforced the project team’s initial view that a call for evidence
would be useful to extend the pool of evidence gathered by the review, so this was
undertaken. A text was agreed with the steering group and sent to contacts suggested by
the experts and those known to the team and the steering group. The call included asking
newsletters, associations and networks to publicise the call, as well as using social media.
The additional sources which were suggested in response were included in the REA.
3.0 Results
The flow of evidence through each stage of the evidence identification and selection
process is shown in the Table 1. A total of 72 pieces of evidence were included in the
synthesis. In this section, the characteristics of the evidence base will be described first.
Then, a synthesis of the findings from the evidence base will be reported. Throughout the
synthesis, confidence statements will be allocated to the conclusions. The confidence
statements (see Table 1) are based on the amount of evidence and the assessment scores
the evidence achieved. As such, confidence can be either High, Medium, Low or
Contested. These statements are only provided for the primary research question
evidence.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 11
Source of
evidence
Search of
academic
databases
Online
search
Stakeholder proposals
(steering group, expert
interviewees, call for
evidence)
Total
Number at
each stage
Evidence
search
5315 NA9 NA >5000
Initial
screening
492 100 136 706
Screened
evidence
165 27 63 252
Evidence
extraction
68 9 30 107
Included in
synthesis
72
Table 1 Description of the evidence identification and selection
Class Description
High Evidence from several studies assessed as ≥6 and 1 or more studies
assessed as ≥7 10
Medium Evidence from one or more studies that have been assessed as ≥6
Low Evidence from a small number of studies or studies assessed all of which
assessed as <611
Contested Evidence that differs in its conclusions (present the assessment for each
study/evidence)
Table 2 Description of confidence statement classifications
3.1 Evidence base characteristics
Overall, the evidence base consisted of 72 pieces of evidence. Although there can be no
expectations or standards for the number of pieces of evidence in an REA, 72 might be
considered relatively large. However, this overall size may be indicative of the wide range
of environmental sustainability information that has been explored as well as the range
of products. Consequently, although a large number cumulatively, the numbers decrease
once the evidence is grouped by product or by information type. That is, though seemingly
large, the evidence base is heterogeneous in nature. To describe the evidence
characteristics as fully as possible, a number of different groupings and matrices have
been used.
9Millions of results were brought up by each set of search terms
10 This was changed from the draft protocol to reflect the fact that we allowed non-integer scores for robustness and relevance
and therefore scores of between 6 and 9 were possible. Therefore, a score of 7 was possible.
11 This was changed from the draft protocol so that low categorisations were distinct from medium
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 12
3.1.1 Environmental sustainability information
As shown in Figure 1, the most consistent type of environmental sustainability
information found within the evidence was related to the energy efficiency and energy
consumption of a product. Although, cumulatively, there were more studies that
investigated non-energy related information, these studies are more diverse in respect to
the information content conveyed, as opposed to the energy studies which were more
consistent in using an energy efficiency rating, kilowatt hours (kWh) and/or monetary
running costs to convey the energy consumption of the product. It should be noted that
for existing labels, we classified them by both the specific label and the information the
label gives, hence the greater number of information types than evidence.
Figure 1 Environmentally sustainable information used in the evidence base
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 13
Figure 2 shows the form and content of some of the labels that were investigated in the
evidence.
Figure 2 Images of some of the labels investigated in the evidence
In addition to consistency in the information content, the energy consumption evidence
also shows more consistency in their presentation form and product. For instance, as
shown in Table 3, the energy consumption was predominantly explored in relation to
appliances, and heating, ventilation and air-conditioning (HVAC). This might be expected
due to the international adoption of energy efficiency standards for appliances. Indeed,
the energy literature has a far larger proportion of existing labels compared to study-
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 14
specific labels (i.e. labels or information that was generated by the researchers for their
study’s purposes). This is likely to represent the relatively longstanding nature of some of
the energy labels.
The proportion of existing labels to study-specific labels is reversed for the non-energy
evidence, which has a range of environmental sustainability information being tested (see
Table 3).
Table 3 indicates the number of studies and the effect they found for the information type
by the product category. The indication of effect has been simplified for tabulation and so
may not fully represent each study and the synthesis provides more detailed discussion
on the different effects. However, the table does show the areas that have been
investigated and the average findings. Note that some studies investigated multiple
products and so the numbers in Table 3 do not equal the evidence base number.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally Sustainable Product Purchase Decisions: Expert Interviews
and a Rapid Evidence Assessment 15
Table 3 Matrix of information type by product with indication influences
EU energy label 1 2 8
Energy Star 1 2 1
Energy Rating 1 1
EnergGuide 1 2
Euro Topten 1 1 1
Energy Frontier 1 1
China Energy Label 1
Nordic Swan 1 1
Building energy performance (incl. BEAM, BER and EPC) 1 2
Green Seal 1
Wood certification (incl. Swiss Quality label and FSC) 5
Energy consumption (kWh) 4 2 12 1 1 2 1 1
Energy efficiency 4 2 14 1 1 4 1
Monetary running costs 1 3 5 2
Carbon footprint/emissions 1 2 1 1 1 1 1
Lifespan 1 1 1 1
Recycled/ reclaimed materials 1 1 1 2 1 1
Remanufactured/refurbished 1
Organic 1
Environmental impacts (incl. water and resource depletion)1 2 1 2 1 1 4 1 4 1 2 1
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 16
3.1.2 Products
The range of products and their frequency in the evidence base is shown in Figure 3.
Appliances are the best represented, followed by cleaning/home cleaning products. Some
products appeared only once in the evidence base. The frequency of studies investigating
appliances is related to the high level of studies investigating energy consumption. In
contrast, the high frequency of clearing/home products may be due to researchers
intending to select products which are regularly purchased by a large number of
consumers and are, therefore, familiar to most consumers.
Figure 3 Products investigated in the evidence base
3.1.3 Measures of influence (method)
As shown in Figure 4, a range of measures has been used to assess the influence of the
environmental sustainability information on participant/consumer behaviour. Actual
purchase behaviour is the least frequent measure, whilst purchase intentions, willingness
to pay (WTP) and choice of product (indicating product preference) were the highest.
Studies which focused on intentions to purchase or choice preference between two or
more products were frequently used as the outcome to assess the influence of consumer
information. In addition, many of these studies also measured other factors to test
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 17
hypothesised relationships, mediations and moderations between the factors and
intentions/choice.
Several studies measured self-reported past-purchases; that is, those which asked people
if they had bought a product recently, if that product had a label and if so, did it influence
them. This category also includes studies which asked people if they recalled using a
label/information when purchasing a product in the past. These studies typically
presented descriptive data in the form of percentage of people asked.
Willingness to pay (WTP) was measured in multiple ways. Frequently, participants were
asked how much extra (premium) they would be willing to pay for a product that was
associated with environmentally sustainable information. This was then indicated either
as a monetary amount or percentage increase in price, in response to either an open-
ended question or multiple-choice options. These results are distinct from WTP measured
in a choice experiment - in which the amount participants would be WTP for the product
is statistically inferred based on the product selection across multiple choices.
3.1.4 Intervention types
A large number of studies in the evidence base used an experimental design to explore
the effects of different information. Although there was not always a control group, the
experimental designs typically enabled the different effects of the information content, or
no information, on participants’ choices, preferences, intentions, or purchases.
A brief (and certainly not comprehensive) explanation of choice experiments will be given
to aid the understanding of the results. The choice experiment methodology involves
presenting participants with products that are given a number of attributes. These
attributes are then given discrete levels. For example, an appliance may be given an
attribute of “colour” and that attribute given three levels; silver, black or white. The
attributes the product have stays the same, but the level of the attribute is randomised.
As such, participants are shown two or more products with comparable attributes, but
the levels of those attributes will be different (e.g. a silver fridge vs a white fridge). The
participant then compares the products and selects the one they would prefer. This
design enables researchers to statistically infer the attribute that had the most influence
over the participants’ product choices (e.g. if colour was most important and if so, what
colour). Within this evidence base, it was typical to present a product with at least one
attribute which related to environmental sustainability information/labelling.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 18
Figure 4 Methods used in the evidence base
3.1.5 Geographic distribution of studies
There was a requirement that the publication was written in English but the search was
not restricted geographically and evidence for studies conducted in several different
countries or regions were found. The largest proportion of studies were in the EU with
several including studies in several Member States or associated countries, perhaps
representing the high proportion of EU Energy Label studies. A proportion of studies were
conducted in the US (particularly ENERGY STAR studies), whereas a smaller proportion
were conducted in China and South Korea.
3.1.6 Assessment scores12
Average assessment scores by product are indicated in Figure 5. As a large amount of the
evidence base is composed of peer-reviewed literature, the frequency of evidence
achieving the maximum of three for robustness was high. However, due to the
assessment criteria only awarding a three for relevance if the study was a field trial, of
which there are few, only a small number of studies achieved the maximum nine on the
assessment.
12 The criteria used to assign relevance and robustness scores are defined in Annexes B and C of Appendix 1, the REA protocol
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 19
Figure 5 Relevance and robustness scores by product
3.2 Evidence synthesis
For presentation and discussion, the evidence is grouped by product type under the broad
product categories of appliances (including heating and cooling), consumer electronics,
cleaning/home chemical products and cosmetic products, automotive, textiles and
clothing, buildings and construction, and paper and wood products. Where there is
sufficient evidence, the evidence is further grouped in to those that have the same
environmental sustainability information. This was done to facilitate the synthesis of the
evidence and to highlight converging and diverging findings. As highlighted in the
evidence base characteristics section, some products have been investigated in only one
or two studies13, whereas others have been investigated more frequently. Likewise, as
some studies investigated more than one product, some evidence is cited in more than
one section.
The findings have been split into those that relate to the primary research question14 and
those that relate to the secondary research questions. The available evidence for both
13 in the evidence base
14 The primary research question was: What evidence is there about the effectiveness of providing factual information
(including content, source and format) on the environmental sustainability of a product in influencing consumer
(individual and organisational) buying decisions?
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 20
primary and secondary research questions varies with product type, with secondary
evidence not extracted for some product types.
3.2.1 Appliances
Primary research question; appliances
Appliance confidence statements:
• There is evidence that EU Energy Label positively influences the purchase, choice,
or intentions, towards more energy efficient appliances (confidence = high).
• There is contested evidence for the influence of ENERGY STAR on consumer
behaviour.
• There is contested evidence for the addition of monetary running costs to the EU
Energy Label.
• There is contested evidence for the influence of monetary running costs
compared to energy consumption on consumer behaviour.
• There is evidence that carbon footprint information, in conjunction with energy
consumption information, positively influences willingness to pay for washing
machines. (confidence = medium).
• There is evidence that environmental impacts information (carbon footprint,
water use, eco-toxicity, and resource depletion), in conjunction with energy
consumption information, positively influences willingness to pay for washing
machines. (confidence = medium).
• There is evidence that product lifespan information increases consumer
preference for the labelled appliance (confidence = medium).
When surveyed, most consumers say they use energy labels when purchasing appliances.
For instance, 88% of 4,000 Australians reported using an energy label (which included the
Australian Energy Rating, Gas Energy Rating, or Water Conservation Rating labels and
ENERGY STAR) and 75% indicated the label is very important in their appliance purchasing
process (page 5, ArtcraftResearch, 2005). However, these types of studies, which ask if
people have used labels when making purchases, may give an over-optimistic picture of
how much people are influenced by labels. For instance, Feldman and Tannenbaum
(2000) specifically looked at USA consumers who had recently bought either a refrigerator
or a clothes washer (approx. sample size of 560 & 240, respectively). In their sample, only
16% for refrigerators and 7% for clothes washers both recalled the ENERGY STAR logo
being present on their appliance and reported being influenced by it.
The influence of the EU Energy Label on appliance purchases has been explored most
frequently out of the existing labels in the evidence base. Overall, there is strong evidence
that the presence of an EU Energy Label on a large appliance influences - or is related to
– selection of the more energy efficient product. For instance, the majority of respondents
in Greece (85% of 596) self-reported that the EU Energy Label had influenced their
purchase of refrigerators and the majority of respondents in Spain (72% of 500) reported
it had influenced their purchase of washing machines (Foudi, de Ayala, López, & Galarraga,
2018). Furthermore, WTP increased with higher grades of energy efficiency (indicated on
the label), with WTP increasing by 696 CHF (£547) for an A graded washing machine
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 21
compared to a C graded washing machine (347 CHF for B to A) (Langley et al., 2012;
Sammer & Wüstenhagen, 2006). Critically, in an experimental field trial, it was found that,
compared to no label, having an EU label increased the purchase of more energy efficient
vacuum cleaners, tumble dryers, and fridge freezers (Stadelmann & Schubert, 2018). As
such, overall, there is strong evidence that the presence of an EU Energy Label on a large
appliance influences - or is related to - more frequent choosing the more efficient product.
The US ENERGY STAR label has not been explored as frequently as the EU Energy Label
within the evidence base; there is some evidence for an influence of the ENERGY STAR
label on consumers. In a field trial, telesales provided some customers with information
on the energy efficiency and environmental benefits of ENERGY STAR water heaters. This
did not have a significant influence on purchases compared to when the telesales people
did not provide the benefits information (Allcott & Sweeney, 2017). However, this study
had multiple confounds as the salespeople varied in their delivery of the information.
These variations perhaps highlight the difficulty of conveying environmental sustainability
information to customers in a real-world setting, and make the findings difficult to
interpret or generalise.
In contrast, though, an analysis of transactions from a US appliance retailer, Houde (2018),
found that the ENERGY STAR label had a positive association with WTP for a full-size
refrigerator. The WTP also increased with the consumers’ income with a WTP of US $30 in
the low-income tertile, US $44.60 in the medium-income tertile and US $56.90 in the high-
income tertile. A refrigerator with the ENERGY STAR label (compared to no label) was also
found to be preferred and have a higher WTP of between US $249.82 (£195.93) and US
$348.30 (£273.95) compared to no label (estimations varied with model used) (Ward,
Clark, Jensen, Yen, & Russell, 2011).
The Australian Energy Rating label has had few studies explore its influence that were
selected in the evidence base for the REA. A field trial in an Australian online store found
that there was no statistically significant effects of the Energy Rating Label or of their own
running costs label on purchases of vented dryers, fridges or washing machines
compared to when there was no label (BETA, 2018). However, the observed, non-
statistically-significant changes did suggest a positive effect of labels – but there was no
statistically significant difference between the effectiveness of the two labels.
A newer energy label is the South Korean Energy Frontier label. The only study15 to test it
offers medium strength evidence that the presence of the label was associated with
greater preference and an increased WTP for refrigerators as the energy efficiency grade
(indicated on the label) increased (Jeong & Kim, 2015). A single study16 also explored the
China Energy Efficiency Label and found that higher ratings were associated with a higher
WTP for air conditioners and refrigerators (Shen & Saijo, 2009). Similarly, a WTP of
US$67.60 (£52.85) was found for an increase from a (study-specific) 3 star energy
efficiency rating to a 5 star rating on air conditioners in India (Jain, Rao, & Patwardhan,
2018).
15 in the evidence base
16 in the evidence base
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 22
Only one study17 (Barthel, Kaselofsky, & Madry, 2015) explored the provision of energy
information not in a label form. In a follow-up survey (n=399), it was found that 82% of
visitors to the Euro Topten website (website to provide information the most energy
efficient products on sale in a given country) bought one or more products listed on the
website (household appliance or a lighting device). These results cannot be considered
representative of all the visitors to the website, as respondents to the follow-up survey
will be highly motivated participants.
The EU Energy Label in its current form (displaying an energy efficiency rating and annual
energy consumption as kW/h) was often used as the control label in the large appliance
studies, with the relative effectiveness of adding monetary running costs to the EU Energy
Label being tested. The evidence from the experimental field trials is contested, with
either no significant effect found for the addition of running costs or an effect found, but
only when in conjunction with other factors (DECC, 2014; Kallbekken, Sælen, &
Hermansen, 2013; Stadelmann & Schubert, 2018). For instance, compared to just the EU
Energy Label, the addition of lifetime running costs to the EU Energy Label did not
significantly increase purchases of the more energy efficient vacuum cleaners, tumble
dryers, fridge freezers, or washing machines (DECC, 2014; Kallbekken et al., 2013;
Stadelmann & Schubert, 2018). For tumble dryers, there was a significant effect of adding
the running costs, however, it was only when the labels were combined with staff sales
training and because of the tumble dryer’s higher running costs (relative to the non-
significant fridge-freezer; (Kallbekken et al., 2013). Likewise, DECC (2014), also found a
significant effect of adding running costs, but only for the washer-dryer, where the high
running costs (compared to the washing machine and tumble dryers) may have made the
information more salient to the customer.
In contrast, studies using experimental surveys suggest that the addition of running costs
to the Energy Label or providing running costs alone, are more effective than the EU
Energy Label on its own (Andor, Gerster, & Sommer, 2016; Blasch, Filippini, & Kumar,
2017). For instance, compared to a simplified version of the EU Energy Label (only the
energy consumption information), the addition of annual running costs significantly
increased the participants’ choice of more efficient refrigerator (Andor et al., 2016).
Similarly, for lightbulbs and refrigerators, displaying the annual energy consumption on
the EU Energy Label as a monetary cost compared to a physical unit (kWh, as is standard),
led to a greater likelihood that the most cost-efficient appliance was chosen (Blasch et al.,
2017).
Further studies compared energy consumption information with monetary running costs.
Using variations of the EnergyGuide label, it was found in the USA that the monetary
savings information had the strongest influence on water heater choices, followed by the
amount of physical energy an appliance used, followed by information on CO2 emissions
(Newell & Siikamäki, 2014). In contrast another US study found that, providing information
(in monetary terms) on how much energy could be saved by each washing machine did
not have a significant effect on participant choices (compared to no monetary energy
saving information), whereas an study-specific energy efficiency score was associated
17 in the evidence base
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 23
with choosing the more energy efficient washing machine; (Niederberger & Champniss,
2017).
Three studies (two conducted by Langley et al., 2012) in the evidence base investigated
the effect of other forms of environmental sustainability information, rather than just
energy consumption. In a large cross-European study, (including UK, France, Germany,
and Italy), Langley et al. (2012) found that, compared to the current EU Energy Label, bids
in an experimental auction were higher for washing machines (average of €3.75) that had
a modified EU label that also displayed the products’ carbon footprint. The same effect
was found for adding environmental impacts (including carbon footprint, water use, eco-
toxicity and resource depletion), although the increase was smaller (€2.16). For all the
labels, the level of efficiency indicated by the label had a significant effect on bids as well,
with bids for higher efficiency washing machines being an average €2.99 higher than for
lower efficiency versions. In contrast, there were no significant differences in bids for
lightbulbs18. A choice experiment using the same labels supported their experimental
auction findings, with a WTP an average of 40% more for a washing machine with the
“energy and carbon footprint label”.
A further, large scale, European experimental survey tested the effect of lifespan
information. It was found that lifespan labelling influenced purchasing decisions in favour
of products with longer lifespans. Across the tested products (which included washing
machines, vacuum cleaners and coffee makers), products with a label showing a longer
lifespan than the competing products were chosen an average of 13.8% more (Jahnich,
Boulbry, & Dupre, 2016).
Secondary research question; appliances
Which other product characteristics shape decisions: Price is commonly found to be one
of the most important factors that consumers take into account when they are buying an
appliance. For instance, in several studies in the evidence base, price was found to be
most influential (Consumer Focus, 2012; Goucher-Lambert & Cagan, 2015; S. L. Heinzle &
Wüstenhagen, 2012).
Which audience factors are relevant: Age and gender have been found to influence
purchase decisions in a number of studies. For age, there is some evidence that the
influence of energy and lifespan labels decreases with age (Jahnich et al., 2016; Ward et
al., 2011). However, gender differences may exist for different labels; preference for
refrigerators with the ENERGY STAR label was higher among males than females (Ward et
al., 2011). However, in a separate study females were more likely than males to base their
choice of appliances on the lifespan label (Jahnich et al., 2016).
Understanding of the labels is also highlighted by some studies within the appliance
evidence. For instance, greater understanding of labels was associated with choosing the
more energy efficient appliances (Langley et al., 2012). Similarly, more general energy
18 The authors speculate that “the lower stakes associated with light bulbs in the bidding experiment, and the need to input bids which included decimal points, may have led to more manual errors creating noise in the light bulb bid data. This has resulted in observations for the light bulbs not being as clear as those for the washing machines and televisions.”
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 24
literacy and investment literacy was associated with choosing the more efficient appliance
(as indicated by the EU Energy Label; Blasch et al., 2017). Equally, a lack of information or
a lack of understanding of the label may prevent people being able to choose the most
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 53
Building ecolabels Textile labels
International LEED Oekotex 100
International BREEAM Global Organic Textile Standard
US EPA water sense Fairtrade Cotton
Bluesign
Information
Sustainable Procurement Guidance Other information
EU Green Public Procurement Guidance
(GPP)
Top Ten (for energy only)
US EPEAT (IT and consumer electronics) Top Runner (Japan)
Irish Government
Eco schools England
Energy Saving Trust retailer buyer’s guides
(UK)
TopTen professional procurement guides
(EU)
Cradle to Cradle product certification
(USA)
Table 2 Labels and information provision to be included in the search
Development of inclusion and exclusion criteria.
Inclusion and exclusion criteria for evidence were determined by the research team, in
conjunction with the Steering Group, based on the PICO components of the primary
research question and the initial trialling and refining of search terms. The draft criteria
are shown in table 3.
Inclusion and exclusion criteria for REA How implemented in the
REA
Exclusion criteria
Research which has only explored responses to the
information itself (e.g. comprehension, evaluation etc.) and not
its impact on purchase-related decision-making or behaviour.
Using search terms (see table
1).
Not to include product behaviours that are not purchasing
related (e.g. disposal). However these to be separately noted
and citations supplied to Defra/WRAP
Manually during data
selection
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 54
Not to include market impacts. However these to be
separately noted and citations supplied to Defra/WRAP
Manually during data
selection.
Studies related to health* Have “not health” in the
search string or manually
during data selection.
Inclusion criteria
Research published between the year 2000 and current date Search criteria (>1999).
Must have researched a form of information which relates to
environmental sustainability AND a product.
Search criteria (see table 1).
Any geographic focus Search criteria (no
geographic focus defined).
Must have assessed a behavioural/decision outcome of the
provision of a form of information.
Search criteria and manually
during data selection.
Table 3 Inclusion and exclusion criteria
* Note: It was agreed with the steering group that food labels would not be included. It
was found that using “not health“ was the most efficient way of eliminating evidence that
related to food. This will be used if there appears to be a high proportion of food focused
evidence within the search results. However, where specific instances of an overlap
between environmental sustainability and health are expected to occur (e.g. organic
labels), then a specific search may be carried out without using the ‘not health’ criteria.
The agreed search protocol will be followed throughout the evidence assessment.
However, the protocol is a working document. If changes to the protocol are deemed
necessary due to unexpected results identified during the evidence assessment, the
Steering Group will be consulted and any protocol changes that are made will be
recorded.
Sources of Evidence
Evidence will be sought from the following sources:
◼ Peer-reviewed literature will be sought using the database platforms Web of Science
and Scopus. These will enable multiple journal databases to be searched
simultaneously using the identified search strings.
◼ Grey literature will be sought through use of internet search engine “Google” (using
identified search strings) and specific searches of relevant institution websites (e.g.
the European Commission, ecolabel.eu). It should be noted that due to the large
number of search results Google is likely to produce, extraction will be limited to the
first 100 results.
◼ Unpublished evidence and grey literature that has been provided by the expert
interviewees and from the project steering group.
◼ Further evidence will be sought through a call for evidence. The content of the call for
evidence will be drafted by the research team before being agreed or amended by
the steering group. The agreed call for evidence will then be distributed through
reviewer contacts, expert interviewees and relevant, existing mailing lists comprising
academics, policy organisations, relevant government departments/agencies, other
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 55
stakeholders (e.g. Energy and Social Science Network; Sustainable Development
Research Network), and other publications as suggested by the expert interviewees.
This will specify unpublished evidence as a priority but welcome the highlighting of
relevant evidence from white literature.
Evidence search and search record
The agreed information search protocol will be implemented to conduct the evidence
search in a systematic and transparent manner. A record of searches will be maintained.
The search terms, the date, the database, the number of hits and any date limits for each
search will be tabulated. Details of individual pieces of evidence identified separately (e.g.
from interviewees or from individual websites) will also be tabulated in a separate
spreadsheet with the publication name, date, and source location (including a hyperlink)
and made available to the Steering Group. After the completion of the individual search
strings, results from each search will be combined to create a full list of evidence (with
duplicates removed). Reference managing software, Endnote, will be used to aid this
process.
Screening search results
The selected inclusion and exclusion criteria will then be used to identify the most relevant
evidence amongst the search results. Evidence will be evaluated in two stages:
◼ First stage: Using the title of the evidence, the evidence will be considered against the
inclusion/exclusion criteria. A rating of “clearly relevant”, “clearly not relevant” or
“uncertain” will then be given. A full text will be obtained for evidence evaluated as
“clearly relevant” or “uncertain”.
◼ Second stage: The abstract (or first paragraph) of the full text will then be read and
evaluated against the inclusion/exclusion criteria. Those evaluated as meeting the
inclusion criteria will be selected, noted and used at the extracting evidence stage.
The ratings of the evidence at the second stage will be recorded in an evaluation record
spreadsheet. Likewise, whether the evidence is subsequently included will be recorded.
These spreadsheets will be made available as supplementary information. In order to
reduce bias and increase consistency and objectivity, at the start of each stage a second
member of the research team will independently screen a sub-section (~5%) of the
evidence. The evaluations of each team member will be compared.
Inconsistencies between evaluations will be discussed to ensure that the
inclusion/exclusion criteria are applied consistently and with minimal bias. This process
will be recorded. Furthermore, if an aspect of environmental sustainability information
that was raised by the interviewees as being important is not present within the screened
search results, then the interviewee(s) may be contacted to provide evidence.
Extracting the evidence
The selected evidence will be read in full to extract the information critical to answering
the research questions. See Table 4 for a list of information that will be extracted. All
evidence read at the full text stage will be presented in an Excel file with its extracted data
and critical appraisal (see below). This will form the systematic map or database, which
will then be provided to the Steering Group.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 56
Citation Details
Author(s)
Year of publication
Title of paper
Title of publication (e.g. book, journal,
report)
Vol., Issue, Pages
Nature of study
The type of evidence
The research design used
The population studied
The product-type/sector
Details of the intervention applied
Outcomes measured
Evidence relating to the primary question
(e.g. evidence of impact/response
measured or observed)
Evidence relating to secondary questions
a) What other product information
(e.g. brand, type of product, price)
and consumer characteristics
influence purchasing of
environmentally sustainable
products?
b) What factors have played a role in
successful schemes?
c) What factors have played a role in
unsuccessful schemes?
d) What, if anything, is required in
addition to factual information to
positively influence purchase
decisions?
Table 4 Data extraction form
Critical appraisal of the evidence
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 57
To ensure the most relevant and high-quality evidence is given greater weighting in the
synthesis, the evidence will be subject to an appraisal. The appraisal will take into account
the following:
◼ Relevance: Each piece of evidence will be appraised for its relevance to the primary
research question. The relevance criteria are derived from Collins et al. (2015). A
numerical value of between 1 and 3 will be allocated for each criterion, with 1
representing lower and 3 representing higher relevance. An overall, summary score
using the same 1 to 3 scale will then be given based on the evidence’s performance
on all the criteria. See Annex B for the list of relevance criteria.
◼ Robustness: Each evidence type (e.g. experiment, review etc.) has its own criteria.
During the extraction phase, each piece of evidence will be coded for its type and how
well it meets the robustness criteria for its type, scored from 1 (few criteria met) to 3
(all/most criteria met).
[Note, if no information is provided in the evidence for one of the criterion, then a 1
will be awarded to the evidence on that criterion. However, if the study appears
highly relevant then, in a few isolated cases the author may be contacted and asked
to supply the missing detail.]
An overall, summary score using the same 1 to 3 scale will then be given based on the
evidence’s performance on all the criteria. See Annex C for a list of the different
robustness criteria.
◼ Combination: The numerical values for relevance and robustness will be multiplied
for each article to give a combined score (1 = weak evidence and 9 = strong evidence).
Evidence with a higher combined score will be given greater weight in the synthesis.
At this stage, very low scoring evidence may be excluded (to be decided by review
team). Any evidence excluded on this basis will be recorded.
Assigning Confidence and the Creation of Evidence Statements
Using the scores from the critical assessment, a confidence class will be assigned to
descriptions of what the evidence indicates during the synthesis phase. This will provide
an indication of level of confidence with which conclusions can be made from the existing
evidence. The confidence classes and their explanations are shown in table 5.
Class Description
High Evidence from many studies assessed as 6 and/or 1 or more studies assessed as 9
Medium Evidence from one or more studies that have been assessed as 6
Low Evidence from a small number of studies or studies assessed as ≤4
Contested Evidence that differs in its conclusions (present the assessment for each
study/evidence)
Table 5 Categorisation of confidence
Synthesis of the evidence and production of deliverables
All the selected evidence will be reviewed and used to answer the research questions (with
greater weight given to the more relevant and robust evidence) and to report findings on
the adequacy of the evidence base. The synthesis will result in a concise technical report
detailing the following elements:
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 58
◼ Description of the volume and characteristics of evidence base: This will include the
types of evidence, research designs, interventions used and outcomes measured.
Gaps and abundance in the evidence will then be highlighted. Results and
conclusions from the critical appraisal will also be discussed.
◼ Description of what the evidence indicates: this will be reported in relation to the
research questions in a narrative synthesis. Tables, figures and graphs will be used to
convey the review and assessment process (e.g. assessment criteria, flow diagram of
included/excluded evidence). In particular, points of evidence convergence and
divergence will be discussed, as well as mixed or uncertain findings. Confidence will
be assigned (according to the relevance x robustness score) and evidence statements
created.
◼ Implications: Indications from the evidence will be related to key policies and
practices (identified with the steering group) to determine if the evidence supports
them. Recommendations for future research or future review processes will be made.
In addition, a two page summary will be produced for publication purposes and a
slide deck will be prepared for the final project meeting to summarise the findings.
References
Collins, A.M., Coughlin, D., Miller, J., Kirk, S. 2015. The Production of Quick Scoping
Reviews and Rapid Evidence Assessments: A How to Guide.
Clayton, S., Devine-Wright, P., Stern, P., Whitmarsh, L., Carrico, A., Steg, L. Swin, J. &
Bonnes, M. (2015). Psychological Research and Global Climate Change. Nature Climate
Change, 5, 640-646.
Corner, A. Whitmarsh, L. & Xenias, D. (2012). Uncertainty, scepticism and attitudes
towards climate change: biased assimilation and attitude polarisation. Climatic Change,
114, 463-478.
Verplanken, B., & Wood, W. (2006). Interventions to break and create consumer habits.
Journal of Public Policy and Marketing, 25, 90-103
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally Sustainable Product Purchase Decisions: Expert Interviews and a
Rapid Evidence Assessment 59
Annex A Conceptual model of processes and influential factors in environmental sustainability information influencing consumer
buying decisions
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 60
Annex B- Relevance assessment criteria
Relevance of the selected articles
Criteria Low (1) Medium (2) High (3)
The relevance of the
method used to the
REA question*
No intervention
(observational)
Intervention in
laboratory or online
survey setting
Intervention in field
The relevance of the
intervention
assessed*
Environmental
sustainability was
only a small
component of the
information
Environmental
sustainability, in
general, was a large
or main component
of the information
A specific
environmental
sustainability aspect
was the main
component of the
information
Relevance of the
outcome
measurement*
Stated attitudes Stated intentions Purchases (actual or
experimental setting)
Provision of
information
Information was
provided out of
context (e.g. not in
relation to a product)
Information was
provided in relation
to a hypothetical
product
Information was in
relation to a specific
product.
*From the Defra NERC guide to scoping reviews and rapid evidence assessments.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 61
Annex C- Robustness assessment forms
Quantitative: Experimental Design
Title of Evidence Review/Statement:
Title of study reviewed:
Date and authors of study reviewed:
Name of quality assessor:
Date completed:
Criteria Score* Comments
Ge
ne
ral
Are the question(s) and hypothesis/hypotheses
addressed by the study clearly identified?
Are related existing research and theories
acknowledged?
Are sources of funding and vested interests
declared?
Me
tho
do
log
y
Is the sample population used in the study
representative of the overall population that is the
subject of the study and is it relevant in the
context of the evidence statement (e.g. relevant to
England/UK)
Were the experimental/management
interventions well described?
Me
tho
do
log
y c
on
t.
Was the allocation of the
management/experimental interventions random?
If not are confounding factors likely?
Was an adequate control group used? Was this
similar to the population receiving the
management/experimental intervention?
Were outcome variables/measures reliable? I.e.
were outcome variables/measurements objective,
was there any indication that measures had been
validated or subjected to another QA processes?
Were the experimental/management
interventions applied representative in the context
of the evidence statement (e.g. relevant to
England/UK)
An
aly
sis
Were the analytical methods appropriate?
Were the estimates of effect size given or
calculable?
Was the precision of the intervention effects given
or calculable? I.e. Were confidence intervals and or
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 62
Criteria Score* Comments
p-values for the effect estimates given or
calculable?
Su
mm
ary
Overall how well was bias minimised by the study
and how relevant is it to the evidence review/
statement? I.e. how well are the criteria above
met?
*1= Adequate, 2 = Good, 3 = Excellent.
Quantitative: Observational Design
Title of Evidence Review/Statement:
Title of study reviewed:
Date and authors of study reviewed:
Name of quality assessor:
Date completed:
Criteria Score* Comments
Ge
ne
ral
Are the question(s) and hypothesis/hypotheses addressed by
the study clearly identified?
Are related existing research and theories acknowledged?
Are sources of funding and vested interests declared?
Is the sample population used in the study representative of
the overall population that is the subject of the study and is it
relevant in the context of the evidence statement (e.g. relevant
to England/UK) -
Were the experimental/management interventions applied
representative in the context of the evidence statement (e.g.
relevant to England/UK)
Me
tho
do
log
y
Were the experimental/management interventions well
described?
How were the exposure and comparison groups selected? Was
bias minimised?
Was the selection of explanatory variables based on a sound
theoretical basis?
How well were likely confounding factors identified and
controlled? Were there likely to be any confounding factors that
have not been controlled for that could cause bias?
Were outcome variables/measures reliable? I.e. were outcome
variables/measurements objective, was there any indication
that measures had been validated or subjected to QA
processes?
Were the analytical methods appropriate?
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 63
Criteria Score* Comments A
na
lysi
s
Were multiple explanatory variables considered and accounted
for in the analysis?
Were the estimates of effect size given or calculable?
Was the precision of the intervention effects given or
calculable? I.e. Were confidence intervals and or p-values for
the effect estimates given or calculable?
Overall how well was bias minimised by the study and how
relevant is it to the evidence review/ statement? I.e. how well
are the criteria above met?
Su
mm
ary
Overall how well was bias minimised by the study and how
relevant is it to the evidence review/ statement? I.e. how well
are the criteria above met?
*1= Adequate, 2 = Good, 3 = Excellent.
Reviews e.g. literature reviews, systematic reviews etc.
Title of Evidence Review/Statement:
Title of study reviewed:
Date and authors of study reviewed:
Name of quality assessor:
Date completed:
Criteria Score* Comments
Ge
ne
ral Is the aim of the question/topic of the review
clearly identified?
Are sources of funding and vested interests are
declared?
Me
tho
do
log
y
Was a search strategy outlining key words and
sources to be searched identified a priori and used
consistently?
Was publication bias mitigated through the
identification of grey/unpublished literature.
Is there a clear rationale for the inclusion of
studies and is this applied consistently.
Sy
nth
esi
s
Has information from the review synthesised
information in a way that minimised bias.
Do the conclusions relate to the information found
by the review.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 64
Su
mm
ary
Overall how well was bias minimised by the study
and how relevant is it to the evidence review/
statement? I.e. how well are the criteria above
met?
*1= Adequate, 2 = Good, 3 = Excellent.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 65
Appendix 2 – Adaptions to the REA protocol
Despite testing during part 1 of the project, difficulties with implementing the search
strings still arose once the full search was conducted and some alterations to the search
protocol had to be made to ensure that the project was manageable and effective within
the scope of the REA and the research question. These changes and their rationale are
outlined below.
Database searches
Databases of peer reviewed literature were searched using Scopus and Web of Science.
Following the full instigation of the search protocol, it was determined that the original
search strings (Protocol Version 4) were inefficient and returning too many results to be
screened within the scope of a REA; a moderately high number of results for one search
string was then compounded by the overall high number of search strings that were
necessary to capture all the keywords of interest. As such, new search strings were
developed which utilised the databases’ proximity Boolean operators (W/x in Scopus and
NEAR/x in Web of Science). This forces the search to only show results where the
keywords appear near to each other. This increases the specificity and effectiveness of
the search by ensuring the words of interest are appearing close together (as opposed to
“AND” which allows the words to appear anywhere). Whilst some of the search strings still
returned high numbers of results, it was felt that, given the breadth of the research
question, further refinement might not be possible at the search stage.
Selecting the number of words that the terms must be within or near requires
consideration of the likely phrases that might be encountered in the relevant evidence.
For this search W/1 and NEAR/1 was adopted. To ensure that this was not overly strict,
the first 200 results for the organisational and individual search strings that used a more
relaxed “W/5” in Scopus were compared to the W/1 in Scopus. Where a relevant piece of
evidence was identified in the W/5 results, the W/1 results were checked to see if it was
present. In all cases they were which gives confidence that the W/1 was efficiently
capturing relevant evidence.
Google searches
For grey literature, Google was used to search the internet. Due to the uniqueness of the
Google algorithms and their move towards using “native language” as opposed to
Boolean operators, the pre-determined search strings were ineffective in Google.
Therefore, following consultation with the steering group, new search strings using fewer
of the keywords were adopted. Furthermore, a specification for .pdf files was added. This
was use in an attempt to focus on evidence reports as opposed to websites. To keep the
returned evidence to be screened at a manageable number for screening, only the first
100 results were captured for each search. Of these 100, approximately the first 30 will
be screened (in anticipation that some search terms results will have greater relevance
than others, more or less than 30 may be screened based on the research teams’
judgement).
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 66
Changes to Keywords
The terms “Bio*” and “Eco*”, added during the interim meeting, proved too broad when
searched. Consequently, Bio* was dropped from the list of sustainability terms and the
Eco* search string had the additional NOT criteria of “economi* and “economy” added.
“Life*” was also found to be too broad and so “Lifecycle”, “lifespan“ and “lifetime” were
specified in the search string instead.
Label searches
It was intended that the specific label searches would be conducted in Scopus, Web of
Science and Google, however, due to project time restrictions, the labels were only
searched for in Scopus and Google.
In addition to these search related changes the confidence statement classifications were
adjusted as noted in table 2 of the main report.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 67
Appendix 3 – Questions for expert interviews
1. In the context of our primary research question:
“What evidence is there about the effectiveness of providing factual information
(including content, source and format) on the environmental sustainability of a
product in influencing consumer (individual and organisational) buying decisions?”
a) Please describe the relevant areas you have worked on:
[Note: Information type (waste reduction, lifecycle, energy, ecolabel, carbon label,
procurement advice etc) and sector (electronic and electrical, textile, others…) ]
b) Please describe any relevant current work by other people/organisations,
published or unpublished, especially any ‘grey literature’ (i.e. produced by
organizations outside of the traditional commercial or academic publishing and
distribution channels) that we might not otherwise find:
2. In the context of our secondary research questions, please provide your insights
and any references to relevant research (published or unpublished) or
refinements to the questions themselves:
What other factors play a role in decisions, which have the most influence, and how
do they interact …?
a) Other product information (e.g. brand, type of product, price) and consumer
characteristics influence purchasing of environmentally sustainable products?
b) What factors have played a role in successful schemes?
c) What factors have played a role in unsuccessful schemes?
d) What, if anything, is required in addition to factual information to positively
influence purchase decisions?
e) To what extent is there divergence or convergence in buyers’ assessment of
environmental information versus its behavioural impact?
f) Is there any evidence of unintended consequences?
3. Which specific labels/schemes do you think we should include with reference to
the following audiences:
a. Organisational purchasers
b. Individuals
4. Please provide details of any textile related environmental information schemes
you are aware of:
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 68
5. Please provide details of anyone else you suggest we interview:
6. Please provide suggestions on how to judge the robustness/significance of
evidence, in particular any methods you consider more robust than others:
7. How would you suggest we best publicise a call for evidence (if issued)?
8. Would you like to get interim feedback on our findings?
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 69
Appendix 4 – Collated responses by experts to interview
questions and summary
Collated responses
The collated response of interviewees to our interview questions is shown here. (Note
that not all questions were answered by all interviewees.)
Role of information and other factors in influencing consumers’ decision-making
◼ The effect of information on behaviour depends on the context, in particular the
product bought. For example the experience from buying appliances or choosing
renewable supply of electricity do not transfer to buying cars
◼ Consumers’ behaviour depends on attitudes – for example on level of environmental
concern or in the importance they place on what happens in the future
◼ For cars, consumers first chose the model/class of car they are interested in – and
the environmental impact is not taken into account in this decision. Then within the
class, various factors are considered including, sometimes, sustainability
◼ Other factors considered include: price; safety; functionality; quality; running costs;
how long a product will last.
◼ For energy using appliances the scope for whole life payback is less than it used to be
- the ‘it will pay you back’ argument for energy efficient goods is becoming weaker –
when the difference becomes small consumers are not motivated by it.
◼ Durability was found to be much more important for large white goods and less
important for clothes and technology (high technological rate of turnover durability).
An unusual trend observed was that there was a relatively higher willingness for
second-hand replacement rather than a new phone upgrade to replace a broken
smartphones
◼ Circular economy considerations are less important for clothes, but a jacket is more
likely to be repaired if there’s an emotional connection
◼ For cars the decision isn’t generally made by an individual on their own – it is made
within a family, with other family members’ views taken into account
What factors have played a role in successful schemes?
◼ Presentation:
o Simplicity of presentation of information so that it can be understood
intuitively and not require mental processing (for household schemes)
o Simplicity of message: those (for householders) which have a simple
message (eg FSC)
o Consistency and comparability of information
o Potential for negative signals: ‘don’t buy’ may be more powerful than a
positive signal
o Schemes that allow choice editing by purchasers (eg EPEAT and TopTen)
◼ Policy:
o Coverage: A regulatory mandate so that all products have to provide the
information/carry the label.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 70
o Combination with minimum standards or any package of policies that
reinforce each other.
o Incorporation of schemes within requirements for public procurement
o Combination with a financial incentive: increases influence directly and
also provides a signal that the Government is serious about the topic –
they are not just providing information they are prepared to put money
into it.
◼ Credibility:
o Endorsement: by a respected authority (eg EC or national government)
o Verification: high credbility provided by certification and ongoing
verification
◼ Traceability:
o For professional procurers it helps to have traceability of the
sustainability of the purchase to show that they are meeting their
organisation’s specific sustainability goals.
◼ Promotion:
o Those which are persistently and consistently promoted
◼ Duration:
o Schemes which are long established build recognition and trust
◼ Motivation:
o Schemes driven by demand from consumers, or manufactures who want
to create a market advantage by using sustainable labels, may be more
effective at motivating the creation of sustainable labels than
government.
What factors have played a role in unsuccessful schemes?
◼ Presentation:
o Too much information or information that is too complex
o Linking to a distant, complex or abstract problem is less effective than
something more concrete and closer to hand (eg climate change labels vs
dolphin friendly tuna label).
o Where what is measured is not meaningful to customers (eg some carbon
labels)
o Consistency and comparability: Schemes that use inconsistent
benchmarks or make it difficult to assess or compare between models.
◼ Coverage:
o Non-mandatory schemes can be weaker - when some products don’t
have information it creates doubt around the validity of the information
that others have.
o Overlapping labels (ie labels relating to the same kinds of issues on the
same product eg EU ecolabel on products with the EU energy label) have
reduced impact
o Applicability: Schemes that were designed for one country and transfer to
another without customisation.
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 71
◼ Credibility:
o Schemes that have been associated with hypocrisy or unreliability or
otherwise lack credibility
◼ Promotion:
o Schemes that are not promoted clearly – that are associated with mixed
messages
◼ Ambition / stringency:
o Schemes which set the bar unrealistically high so no or very few products
qualify
What is required in addition to factual information to positively influence purchase decisions
Required factors were limited to:
◼ Credibility:
o Credibility of the information (consistently listed as essential)
◼ Presentation:
o Design is important – information must be provided in a form people can
use.
◼ Confidence:
o Consumers trusting themselves to be able to make the decision.
However, suggestions for other factors that are not required, but which can help
included:
◼ Cultural:
o Favourable public opinion towards the sustainable issue or innovation –
supported by coverage of the issue in the media
o Social norms can be effective – people can be influenced by what other
people are (reported) as doing.
o The sustainability feature being seen as ‘cool and trendy’ by consumers
◼ Policy:
o Financial incentives can help, if carefully designed.
◼ Delivery:
o Social media can be an important tool for communication – act as a
gateway to information
o Gamification through apps (teach people about the sustainability of
products/the labels)
◼ Presentation:
o How easy the ‘green’ product is to find – how it is presented in the
shop/online can make a difference. For example, if a retailer has an own
brand product which is labelled they will display this prominently and this
will sell better
Divergence or convergence in buyers’ assessment of environmental information versus its
behavioural impact
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 72
◼ What people think they want from information and what they respond to are
different (frequently they say they would like more information, but then negatively
react to large amounts of information). Consumers want an easier solution than
labels and information – if they can trust a brand/shop is sustainable across the
board, then they prefer that.
◼ A number of interviewees recognised the gap between intention and behaviour:
o The importance of the gap may depend on the product. Pleasure based
purchases, for example TVs, are less susceptible to an energy efficiency
message than more functional products such as freezers.
o Some studies found a clear link between self-declared attitude and
behaviour, and that it was easier to influence those who were already
environmentally minded.
◼ Labels can influence behaviour quite directly:
o Eco-labels can facilitate an intention becoming a behaviour and so be a
tool to bridge the intention-behaviour gap.
o Labels can serve to make sustainability a factor under consideration at
point of purchase, increasing the likelihood people will prioritise this.
o Information may have an indirect influence by making people aware that
there is a problem that caused the need for a label. The more information
is available and promoted, the more people will be exposed to the
arguments and understanding of the labels.
o Upstream sustainability information is often not a concern for consumers.
They focus on the performance or usage characteristics (eg. the energy in
use).
Evidence of unintended consequences
It is possible that in the drive to achieve environmental sustainability, there are
compromises in other aspects, for example moral or ethical. Whilst one problem may be
solved, another may be created.
◼ Neglect of unlabelled aspects: Labels/information (and the calculations behind
them) may lead to a focus on one sustainability issue of a product to the neglect of
others. For instance, indicating that electric vehicles are low emission only takes into
account their point of use and does not account for the (energy and emission
intensive) production of the battery and so calling electric vehicles “low emission”
may mislead consumers.
◼ Disproportionate focus on aspects vs impacts: For example, the focus on energy
efficiency, rather than energy use, in the EU energy label (and Ecodesign regulations)
for washing machines – as higher energy efficiency is easier to achieve with larger
machines this led to an increase in the capacity of washing machines sold despite
consumers using the same load size.
◼ Unintended wider positive influence: There can be a ‘halo’ or positive ‘spillover’
effect from labels – a product which is labelled for a particular feature may be
assumed by consumers to be higher quality in other respects. Information or labels
on one product or issue may lead to changed purchasing behaviour with other
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 73
products or issues. Information can reinforce other policies even if it has a limited
effect in isolation.
◼ Negative perception of information provision: The study for the EU on resource
efficiency found that there were no unintended consequences from giving
information on reparability – this was not interpreted by consumers as implying that
the product was easy to break.
◼ Information overload: Too many labels may put consumers off.
◼ Negative association between labels: It is possible that weak or ineffective labels
may ruin the reputation of other labels.
◼ Rebound effect: Some interviewees pointed to the rebound effect or moral licensing
(for example driving further in a low carbon vehicle) as a concern; others felt that its
impact was overstated and that it is not a significant problem.
Summary
The interviewees represented a wide range of organisations and experience relating to
environmentally sustainable product purchase decisions: some had been involved
through specific projects, for others this has been the work of most of their career; some
had worked on or studied established schemes such as ecolabels and energy labels,
others in newer areas such as the circular economy. This diversity provided a wide range
of responses to all the research questions.
The interview findings have contributed to the development of the draft protocol (that
has been drafted in parallel with the interview process) which will be used to undertake
Rapid Evidence Assessment (REA) on this topic. They have also provided direct evidence,
in the form of white and grey literature, were included in the REA.
It is difficult to condense such a rich response to a few key points, but we feel that some
highlights can be drawn from the findings:
Considerations for successful schemes:
◼ Presentation (simplicity, consistency, lack of abstraction/meaningful, comparability)
◼ Policy (mandated for comprehensive coverage, combination with financial incentives,
procurement requirements and other policies)
◼ Credibility (endorsement, certification and verification)
◼ Traceability
◼ Promotion
◼ Long duration
◼ Demand driven schemes
◼ Avoidance of overlap
◼ Tailoring to geographical / cultural context
◼ Appropriate level of ambition
Intention vs behaviour: This will vary depending on the product (e.g. pleasure vs
practical products), and it is easier to influence those who are already environmentally
minded.
Unintended consequences of information provision can include:
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 74
◼ Neglect of unlabelled aspects and disproportionate focus on aspects vs impacts
◼ Unintended wider positive influence
◼ Information overload
◼ Negative association between labels
◼ Rebound effect
Features of robust studies include:
◼ Wide range of information types and sustainability aspects addressed
◼ Incentivised choice experiments
◼ Focus groups
◼ Simulation of reality
◼ Long duration
◼ Avoidance of a focus only on intention
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 75
WRAP - The Effectiveness of Providing Pre-Purchase Factual Information in encouraging more Environmentally
Sustainable Product Purchase Decisions: Expert Interviews and a Rapid Evidence Assessment 76