CARBON LABELLING IN RETAIL GROCERY INDUSTRY - A study on consumer attitude and behaviour Nitai Chand Patra Mail: [email protected]MBA Full Time Ustinov College Word Count: 14995 Date: 6 th Sept 2010 Dissertation submitted as part requirement for the degree of Master in Business Administration of the University of Durham, 2010.
CARBON LABELLING IN RETAIL GROCERY INDUSTRY - A study on consumer attitude and behaviour
Welcome message from author
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
CARBON LABELLING IN RETAIL GROCERY INDUSTRY - A study on consumer attitude and behaviour
Dissertation submitted as part requirement for the degree of Master in Business Administration of the University of Durham, 2010.
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Declaration
“This dissertation is the result of my own work. Material from the published or unpublished
work of others, which is referred to in the dissertation, is credited to the author in question
in the text. The dissertation is 14995 words in length. Research ethics issues have been
considered and handled appropriately within the Durham Business School guidelines and
procedures.”
Nitai Chand Patra 2
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Abstract:
This research looks into - UK retail consumer attitude and behaviour towards carbon
labelling, with a view at providing practical recommendations for the enhancements of
carbon label’s effectiveness. Using the Theory of Planned Behaviour (TPB) framework (Ajzen
1991) an extended model for carbon labels has been developed to explain consumer
behaviour in the context of retail grocery. The model has been tested with structural
equation modelling. Next, consumer behaviour has been studied by analysing relative
importance of the carbon footprints in comparison to other major product attributes such
as brand and price. Further, effectiveness of traffic signal based carbon labels has been
evaluated. The empirical study was quantitative and was based on discrete choice based
conjoint analysis. Using orange juice as the instrument, from the responses of 208 UK
participants, consumer behaviour was studied. The choice based approach brings new
insights and empirical evidence on the carbon label.
No comprehensive research has been carried so far to understand the attitude and
behaviour of consumers towards carbon labelling. The study presented in this dissertation
systematically evaluates various factors influencing consumer behaviour towards the use of
carbon labels as a decision making tool. Results show that consumer’s knowledge, ease of
locating, interpreting and comparing the carbon footprints are major predicators of using
the carbon label as a decision making tool. Social norms and influence from others such as
family, friends and environmental groups have positive influence on intention of using
carbon labels, but the influence does not get translated to environmental friendly purchase
behaviour.
Further, results imply that consumers’ attitude differs from behaviour and there is no
association between eco-friendly behaviour and age, income and education. Further, it has
been found that the integration of traffic light label with the present label can enhance the
effectiveness of carbon labels. Based on these findings, some of the topical
recommendations are: print carbon labels on more products, integrate traffic light label
with present carbon label, present carbon labels next to the price and communicate more
about the labels.
Nitai Chand Patra 3
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Contents
2.1. CHOOSING THE RIGHT FRAMEWORK ......................................................................................................................... 14 2.2. CONCEPTUALISING A FRAMEWORK FOR CARBON LABELS ................................................................................................... 17
2.2.1. Attitude - Intention .............................................................................................................................. 18 2.2.2. Subjective Norm – Intention ............................................................................................................... 19 2.2.3. Perceived behavioural control - Intention ........................................................................................... 19 2.2.4. Perceived behavioural control - Behaviour ......................................................................................... 20 2.2.5. Comparison amongst SN, PBC & ACL .................................................................................................. 20 2.2.6. Attitude towards carbon labels ........................................................................................................... 20 2.2.7. Attitude-Behaviour gap: ...................................................................................................................... 21 2.3.1. Motivating Factors: ............................................................................................................................. 21 2.3.2. Role of communication ....................................................................................................................... 23 2.3.3. Awareness Level .................................................................................................................................. 24 2.3.4. Proposal of TLS based carbon label ..................................................................................................... 25
3.1. PART-I GENERAL QUESTIONNAIRE .......................................................................................................................... 27 3.2. PART-II STRUCTURAL EQUATION MODELLING AND QUESTIONNAIRE DEVELOPMENT .................................................................. 27
6.1.1. Direct influence on behaviour ............................................................................................................. 62 6.1.2. Influence on habits that control behaviour ......................................................................................... 63 6.1.3. Influence on the convenience of using carbon labels .......................................................................... 63
INCOME ............................................................................................................................................................... 90 EDUCATION ........................................................................................................................................................... 91 ATTITUDE TOWARDS CARBON LABELS (ACL) .................................................................................................................... 92
Nitai Chand Patra 5
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
List of Tables
Table No. Description Page no.1 Questionnaire 302 Coding 313 Demographic profile of respondents 374 Reliability and validity statistics 385 Score Interpretation 406 Model fitness indices of proposed model 447 Model fitness indices of extended model 459 Regression weights 4510(a) Effects on INT 4610(b) Effect on BEH 4611(a) Conjoint analysis 4611(b) Attribute utility from traffic light conjoint analysis 4712 Wilcoxin signed rank test results 4813 Attribute level utility from two conjoint analysis 5614 Consolidated hypothesis testing result 57
List of illustrations
FIGURE 1: A SAMPLE CARBON LABEL........................................................................................................... 10
FIGURE 2: THEORY OF PLANNED BEHAVIOUR (AJZEN 1991).........................................................................15
THE FIGURE 3: THE FINAL MODEL DEVELOPED FROM SEM...........................................................................45
Figure 13: NAT Proposed by Schwartz 1977......................................................................66
Contents of the attached CD
1. Complete data collected in this study.
2. Analysis of responses for individual questions.
Acknowledgement
Nitai Chand Patra 6
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
I would like to take this opportunity to express my gratitude to all who have helped me in
completion of this dissertation. Firstly, my heartfelt gratitude to my supervisor, Miss Mary I
Mundel, with her guidance only the dissertation has been successful. Secondly, I offer my
regards to my family members and friends to encourage and motivate me throughout the
study. Finally, I thank all the participants and other individuals, who have extended their
support.
Thank you all.
Nitai Chand Patra 7
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
1. Introduction
Grocery consumption contributes to almost one third of the total environmental impact and
emissions arising from EU economies (European Commission 2007). Researchers have
confirmed that, most of the consumer value environment friendliness and products derived
from ethical sources. However, consumers’ buying behaviour is often found to be
inconsistent with their attitude (Uusitalo 1990). It would be interesting to investigate how
companies’ ethical and social responsibility will pay off and the growing concerns of
environment get translated into a widespread purchase of eco-friendly products. The
objective of this dissertation is to contribute to the understanding of eco-friendly shopping
behaviour by examining in the context of carbon labelling and confirming the factors
affecting effectiveness of the label.
Bronwen Jones expressed that carbon literacy is increasing. Many people now understand
direct emissions from their air travels, using cars instead of public transports or household
energy consumption. And it won’t be very long before consumers understand indirect
emissions from their consumption of goods and services (Berry et al. 2008). Food miles, a
term coined by Tim Lang, Professor of Food Policy, City University, London in early nineties
represents the distance travelled by a product between points of production and
consumption. In recent past the concept of food miles became a strong marketing tool for
UK’s National Association of Farmers. In 2006, the association launched a campaign “Local
food is miles better”. Kemp et al. (2009) in their research on UK consumers’ attitude &
behaviour towards Food miles have reported that 21.5% consumers preferred not buying
products, which have travelled a long distance such as from New Zealand & Kenya.
In the use of the term food miles, it is the assumption that the longer a product has
travelled, the larger is its effect on the environment. Saunders et al. (2006) during life cycle
assessment of some products found that, for products those are transported from far New
Zealand, the green house gases emitted due to transportation are higher. However, the
total environmental impacts of those products are far less than the similar products
produced in the UK (Kemp et al. 2009). Therefore, the concept of food miles has some flaws
in the assumption. Further, with the rise in concern of global warming, the concept of food
miles has now evolved to the carbon label. In food miles only a product’s carbon emission
from transportation and distribution were matters of concern, whereas in carbon labels, the
sum total of green house gases emitted during production, transportation, distribution and
Nitai Chand Patra 8
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
consumption are considered. Carbon labels delineate a bigger picture of a product’s impact
on environment. Additionally, in 2002, World Summit for Sustainable Development in
Johannesburg, some leaders suggested for life cycle assessment of products and proposed
adoption of tools and policies for sustainable production and consumption (UN 2002).
The Carbon Label conveys the volume of the carbon footprints (carbon dioxide and other
green house gases) generated by a product or service during its complete life cycle starting
with raw materials to end user’s consumption and disposal. The Carbon Label is developed
by the Carbon Trust, a non profit organisation and the leading authority for reduction of
carbon in UK.
In November 2008, the UK government passed the legislation and adopted the Climate
Change Act, which sets a target of reducing green house emission levels to 80% below 1990
levels by 2050. Further an interim target of 34% reduction has been set for2020 (Carbon
Trust 2010). The Carbon Trust has been established by UK government in 2001 with a vision
to take UK towards a low carbon economy and help achieve the set targets in Climate
Change Act.
The two primary purposes of carbon trust are, firstly, to inform consumers about the
environment impact of the products and services they consume and help them in choosing
an environment friendly product, secondly, to help the businesses in measuring the carbon
emission of the products at every step of their supply chain, thus eventually helping
businesses to explore and evaluate the cost and energy saving opportunities by reducing
waste and enhancing efficiency in the production and distribution of the product.
Additionally, companies displaying carbon labels are committed to reducing the carbon
footprint of the respective products in a span of two years.
Some of the pioneering companies in carbon labelling are Tesco, Walkers, Boots and
Innocent Drink, who have printed carbon footprint on their own branded products to inform
consumers on the volume of carbon generated during the products' life cycle. These
companies work with Carbon Trust to calculate the carbon footprint of their selected
products range.
Nitai Chand Patra 9
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Figure 1: A sample carbon label
Now many more products are displaying a carbon label. E.g. Kingsmill breads, Walkers
shampoos. Appendix-A2 (page-85) presents a list of such companies and products. The
volume of green house gasses emitted is displayed on the label in gm or kg or tonnes. The
lesser the carbon emission of a product, the more environment friendly is the product. For
example, a product with a carbon footprint of 1000gm/unit is comparatively eco-
friendlier/greener than a product with a carbon footprint of 1200gm/unit.
As per the Carbon Trust (2010), the display of the carbon footprints on the products
enhances brand reputation and sales appeal. Further, as per the corporate social
responsibility (CSR) agendas, increasingly many companies are displaying the carbon
footprint of the products. Tesco, the leading supermarket in the UK measures and displays
the full carbon footprint of more than 500 of its products as one of its CSR- initiatives (Tesco
CSR 2010). Tesco CSR manager expressed “We will …begin the search for a universally
accepted measure of the carbon footprint of every product we sell … [to] enable us to label
all our products so that customers can compare their carbon footprint as easily as they can
currently compare their price or their nutritional profile.” (Leahy 2007)
To understand why consumers act as they do, marketers and policy makers require
understanding of more than just the attitude (Ramayah et al. 2009). Understanding the
underlying beliefs, values and other influencing factors that manifest towards the attitude
and intention will help in developing effective marketing proposition and environmental
policies. In order to design effective campaigns and policies; it has been a continuous
endeavour by marketers, environmentalists and government agencies to understand
consumers’ environmental friendly behaviour.
As the concept of carbon labelling recently gained momentum, not much research has been
done so far, so the number of publicly available literatures on carbon labelling is limited.
Nitai Chand Patra 10
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Few of the pertinent works were as such. Berry et al. (2008) conducted eight focus groups
with UK consumers to understand their perception on carbon labelling and recommended
some enhancements. Boots Plc and Tesco have referred to quantitative surveys with their
customers on carbon labels during 2007 in their CSR; however, the reports are not available
on public domain. Upham et al. (2009) tried assessing the public perception on carbon labels
in UK via three focus groups. All these studies were focused on consumer perception, but
none tried evaluating the attitude-behaviour link and factors influencing such behaviour and
the role of communication in such behaviour and hence the study was an attempt towards
filling that gap.
Berry et al. (2008) and Upham et al. (2009) recommended use of traffic light system for
carbon labels, this study evaluated the proposal. Vanclay et al. (2009) have studied
Australian consumers’ behaviour towards traffic light based carbon labels in a non-intrusive
way by monitoring sales at a convenience store. Vanclay et al. performed a comprehensive
study exploring the role of communication in the effectiveness of altering consumer
behaviour in the context of carbon labelling. This study extends Vanclay et al. study on the
role of communication in UK retail settings.
Theoretically, this study builds on ideas that environmental friendly shopping behaviour is
influenced by a number of factors and consumer’s environmental choice is a trade-off
between several choice criteria. Therefore this study not only tries to describe the
underlying attitudes, values and intention towards socially responsible buying behaviour,
but also tries to explore environmental friendly buying behaviour in a more realistic choice
situation, where consumers have to balance their purchase decision over various product
attributes. The aim is to evaluate the extent to which consumers value carbon footprint in
their buying behaviour compared with other major attributes, in a situation where the
purpose of carbon labels is made explicit to the consumers.
Therefore, the following were the objectives of this research:
i. To find the percentage of buyer population is aware of carbon labelling.
ii. To develop a model explaining various factors affecting consumer’s environment
friendly shopping behaviour by extending the theory of planned behaviour model
with structural equation modelling.
Nitai Chand Patra 11
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
iii. To evaluate the extent to which consumer value carbon labels in their product
choices in comparison to other important product attributes. This examines the
importance of carbon labels and its position in the hierarchy of major influencing
factors of consumers’ decision making process.
iv. To explore, demographic variables such as age, sex, education and income level and
their effect on behaviour towards carbon labels.
v. To examine, the effect of integration of traffic light signals with current labelling.
The empirical method employed was quantitative and the research was segregated into two
parts. In the first, using the TPB framework and structural equation modelling a model was
developed describing factors affecting consumers’ eco-friendly buying intention and
behaviour. In this part, consumers past behaviour, intention and attitude have been studied
based on self reporting questionnaires. However, as expressed by Fisher (1993), the human
tendency is to present oneself in the best possible light. So measuring consumer behaviour
from stated preference is a vulnerable method. In order to mitigate this short coming, in the
second part, consumer behaviour towards carbon labelling has been studied using discrete
conjoint analysis. In this method respondents were presented with various combinations of
attributes such as the price, brand and carbon footprint of a product. Based on respondent’s
product choice their behaviour was estimated. The possible impact of consumer
demographics such as age, gender, education level and income level are examined as an
extension.
Household consumptions make up to 45% of average UK consumers’ total carbon footprint.
These averages to 5 tonnes per person per year and around 300 metric tonnes per year for
total UK population (Berry et al 2008). So a step towards understanding consumer
perception towards carbon labelling will be immensely helpful in reducing the emission and
controlling climate change. This dissertation adds to the body of literature by evaluating
factors those influence consumers’ use of carbon labels in their purchase decisions.
Additionally, it suggests some enhancements for improving influence of carbon labels on
purchase decisions.
The remaining part of the dissertation is organised as follows. Section-2 presents the
development of the research hypotheses and research model from literature review.
Section-3 elaborates on the research design, method, questionnaire development, data
Nitai Chand Patra 12
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
collection and sample description. Section-4 presents the final verified model and statistical
results derived for testing of hypotheses proposed under section-2. Section-5 elaborates on
the result and testing of the hypotheses followed by additional discussion on the model and
extensions of research. Section-6 submits the recommendations to various stakeholders
followed by research limitations, recommendations for further researchers and conclusion.
Nitai Chand Patra 13
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
2. Literature review
2.1. Choosing the right framework
Attitudes play an important role in affecting behaviour and influencing decision making.
They strongly influence factors such as what products to buy, where and when. The decision
making process is complex and various factors influence them on various levels. There are
several theories explaining the relation between attitude and behaviour. Staats (2004) has
recommended use of Theory of Reasoned Action (TRA), Theory of Planned Behaviour (TPB)
and Norm Activation Theory (NAT) in the context of environmental behaviour.
In the economic and cognitive convention, it is assumed that consumers behave rationally,
in the sense they act consistently as per their preferences and beliefs. Most consumer
behavioural studies are based on this assumption and employ variations of attitude-
behavioural models such as TPB (Ajzen 1991) or TRA (Ajzen & Fishbein 1975). These models
have clearly demonstrated that attitude can influence or predict behaviour (intention).
However, the relation between attitude and behaviour has been found to be varying to a
large extent on the context of application (Ajzen & Fishbein 1980).
The TRA model postulates that behaviour is the result of three main factors: attitude
towards the behaviour, subjective norm and behavioural intention. Attitude and subjective
norms are postulated to influence intention, which in turn results in behaviour. TPB (Ajzen
1988, 1991) is an extension of TRA and includes an additional component of perceived
behavioural control. According to TPB, the immediate determinant of an individual’s
behaviour is his/her intention to perform the task. And the determinants of intention are
attitude, subjective norm and perceived behavioural control.
Nitai Chand Patra 14
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Figure 2: Theory of Planned Behaviour (Ajzen 1991)
Attitude is explained as a person’s general evaluation of an object or behaviour. Further it
can be described as an individual’s favourable or unfavourable evaluation of performing the
behaviour. Schwartz (1992) explained that attitude is a set of beliefs about an object or act
which may translate into intention of carrying out the act. Therefore, the more favourable is
an individual’s attitude towards behaviour, the stronger will be the intention to perform the
behaviour.
Subjective norm is explained as an individual’s perception towards the social pressure (from
family, friends and society at large) to perform or not perform the particular behaviour and
normative beliefs are its antecedent. If a consumer perceives that significant others approve
(disapprove) his behaviour, then he is more likely (less likely) to perform the intended
behaviour. A responsibility towards environment can also be linked with subjective norm; as
it is expected that consumers play an active role in solving environmental problems by
choosing environmentally friendly products or way of life (Uusitalo 1990).
Perceived behavioural control (PBC) is the measure of an individual’s perception on the
ability to perform the behaviour in the context and control beliefs are its antecedent.
Follows and Jobber (2000) classified it as perceptions of convenience. They have postulated
that inconvenience and additional efforts on the part of a consumer, towards purchasing a
green product, act as a deterrent for consumers to adopt a green practice. In the early
version of TPB (Ajzen 1988) there were no direct link between PBC and behaviour. However,
after meta-analysis of TPB model, Ajzen (1991) proposed that PBC can affect behaviour
directly. Ajzen argued that the increased feeling of control or convenience would increase
Nitai Chand Patra 15
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
the extent to which an individual would be willing to exert additional effort in order to
successfully perform the particular behaviour.
Intention is explained as the immediate determination of acting in a certain way. All the
factors discussed flow through intention and then influence actual purchasing behaviour of
a consumer. Intention has been found to be a good predictor of behaviour. Bagozzi et al.
(1990) has proposed that, in an attitude-behaviour relationship, intention is influenced by
the level of effort needed to exhibit the behaviour.
The underlined assumption for TPB is that people behave rationally in normal condition.
Conner & Armitage (1998) expressed that, this assumption sets limitations to the model, as
people often act spontaneously or habitually. Sutton (1998) review of nine meta-analysis of
TRA & TPB model, found that attitude, subjective norm and PBC account for only 40-50% of
variances in behavioural intention. And the power of these three variables of TPB are low
regarding prediction of behaviour and indicated that different studies found additional
variables affecting behaviour and intention depending on the context of study.
Nevertheless, researches confirm that extended models can help achieving better
predictability of the relationships (Rokka & Uusitalo 2008).
Further, Sutton (1998) found that only 19-38% of variances in behaviour can be explained by
intention and Armitage & Conner (2001) found that (meta-analysis of 185 TPB based
independent studies) TPB could explain only 27-39% of variances in behaviour and
intention. The lower prediction power of the TPB model regarding behaviour has raised
scepticism of the model, as a major part of the behaviour is still unexplained. This implies
that, the TPB model cannot satisfactorily explain the difference between the intention and
behaviour or why consumers behave differently than they originally intended.
Additionally, the attitude concept in TPB is meant for a general evaluation of behaviour, that
is, the attitude is just a decision criterion which guides such behaviour. Many researchers
found that, in the social dilemma character of pro-environmental behaviour, moral
dimension plays an important role. However, these moral dimensions are not covered in
TPB’s attitude. Schwartz (1977) norm activation theory (NAT) helps this issue by explaining
the personal obligation to perform a specific behaviour using situational and personality
trait activators (Harland 2007). However, I felt that without understanding the general
attitude & behaviour, exploring moral dimensions would be less useful.
Nitai Chand Patra 16
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Furthermore, behavioural models such as TPB rely on self reports, which suggest
vulnerability of data, mostly because of presentation biases due to social desirable
responses. This threatens the validity and reliability of the models (Armitage & Conner
2001).
Although there are some limitations of TPB as discussed, as the carbon labelling is an
emerging concept, and currently no framework exists for evaluation of its effectiveness, I
considered this framework as a stepping stone in the process of development of a
comprehensive framework. Further in order to achieve desired validity and reliability of the
model and minimise error, necessary statistical tests and structural equation modelling
were employed for the analysis.
Eco-friendly buying behaviour is a complex phenomenon and there will be several factors
influencing the behaviour. Many researchers have adopted the TPB to examine consumer’s
behaviour and factors influencing such behaviour in the environmental friendly
consumerism context. Some of the useful works are Ho (2002) for waste recycling behaviour
of Singaporean household, Cheung et al. (1999) for waste paper recycling behaviour, Oliver
& Lee (2010) for buying hybrid cars in Korea, Tsay (2010) green products in Taiwan, Gupta &
Ogden (2009) for green consumerism in USA and Della Lucia et al. (2007) for organic coffee
buying behaviour in Brazil (Ramyah et al. 2009).
Staats (2004) expressed that the TPB has been chosen by many researchers for investigating
specific environmental behaviour. And it has been proven to be useful over other models
such as TRA, in understanding the behaviour with the contribution of perceived behavioural
control. Following the footsteps & recommendations of previous researchers, TPB was
chosen to systematically explore various influencing factors of behaviour towards carbon
labels. Further, this was a confirmatory research, and the TPB provided the necessary
support to develop the right framework thereafter.
2.2. Conceptualising a framework for carbon labels
Following the TPB model, in the proposed model attitude toward the carbon labels (ACL)
and subjective norm towards carbon labels (SN) are positioned as indirect predictors of
behaviour towards the use of carbon labels (BEH). Intention is the immediate determinant
of behaviour. Perceived behavioural control (PBC) has both direct effect and indirect effect
(via intention) on BEH.
Nitai Chand Patra 17
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Figure 3: Proposed model for carbon label.
Researchers have found that the three factors discussed in TPB have varied degrees of
influence on intention, which ultimately influences behaviour (Armitage & Conner 2001).
Ajzen (1991) confirmed the same that the effects of individual factors are situational. In
certain case some factors might have very low or insignificant influence. In this research, it
was hypothesised that all the three factors of TPB have a significant influence on behaviour
via intention.
2.2.1. Attitude - Intention
As discussed earlier, attitude towards behaviour reflects an individual’s positive or negative
evaluation towards performing the particular behaviour. So, a favourable attitude will lead
to a stronger intention of performing the behaviour. In their research on recycling
behaviour, Tonglet et al. (2004) found that positive attitude towards recycling is the most
significant predictor of recycling behaviour, and has the strongest correlation with the
intention. Further, Armitage and Conner (2001) from meta-analysis of 115 studies reported
positive relation between attitude and intention. Consequently, it can be expected that an
individual’s positive or negative attitude towards carbon labels would influence individual’s
intention of using carbon footprints during the purchase of any products or services
accordingly. Thus the following hypothesis (H1) was derived:
H1. There is a positive relationship between attitude towards carbon labels and intention
towards using carbon labels as a decision making tool.
Nitai Chand Patra 18
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
2.2.2. Subjective Norm – Intention
Conner and Armitage (2001) discussed that several authors have considered the subjective
norm to be the weakest component and have deliberately removed them from analysis.
However, Trafimow and Finlay (1996) from analysis of 30 studies found that individual’s
attitude to behaviour relationship were primarily driven by subjective norm.
Further, Uusitalo (1989) suggested that favourable attitude towards environmental friendly
products alone cannot predict behaviour, if the social norms and individual’s awareness of
such norms are weak. One’s perception of how friends, relatives and society at large
behave, what they feel about one’s environmental friendly behaviour and social benefits of
such behaviour encourage one to behave favourably or unfavourably. If people important
to a person expect the person to choose environment friendly products then it might
translate to green consumerism. Further, various incentives while buying greener products
(E.g. green club card points by Tesco) will further strengthen the subjective norms. So, it
implies that the subjective norms related to environment supportiveness, incentives from
the retailers and society’s expectations exert a positive influence on an individual’s intention
of using carbon labels. Thus the following hypothesis (H2) was proposed:
H2. The subjective norms have a positive influence on consumer’s intention of using carbon
labels as a decision making tool.
2.2.3. Perceived behavioural control - Intention
Pro-environment PBC measure is operationalised by combining statements related to
amount of effort required in performing the specific environment related act. The easier
and convenient specific environment behaviour is to perform, the higher will be the
intention of performing the behaviour (Dahab et al. 1995, Ajzen 1991, Uusitalo 1989,
Armitage and Conner 2001). Consequently, if an individual perceives the use of carbon
labels as easy & convenient and there are convenience & enough products with carbon
labels to choose from, then the individual has a higher control over the act. Therefore, it
implies that higher is the PBC towards carbon labels the stronger will be the intention of
using them in purchase decision making. Thus the following hypothesis (H3) was derived:
H3. There is a positive relationship between perceived behavioural control towards carbon
labels and the intention towards using carbon labels as a decision making tool.
Nitai Chand Patra 19
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
2.2.4. Perceived behavioural control - Behaviour
Ajzen (1991) argued that under conditions where the behavioural intention alone could
account for a small variance in behaviour; PBC would act directly on behaviour and predict
behaviour. This is based on the rationale that the increased feeling of behavioural control
will influence the subject to exert additional effort, in order to successfully execute the
behaviour. As at this point of the study, the intention-behaviour relation was unknown, so
following Ajzen (1991), it was proposed that perceived behavioural control has positive and
direct influence on behaviour towards carbon labels. Thus the following hypothesis was
proposed:
H4. There is a positive relationship between perceived behavioural control towards carbon
labels and the behaviour of using carbon labels as a decision making tool.
2.2.5. Comparison amongst SN, PBC & ACL
Contrary to Uusitalo (1989) as discussed in section-2.2.2, page.19, Armitage and Conner
(2001) mentioned that many meta-analyses suggest that the subjective norm is the weakest
predictor of intention. Even some authors have removed subjective norm from their
analysis. This reflects lesser importance of normative factors as determinants of intention.
So, it is hypothesised here that subjective norms have lesser influence on intention of
considering carbon labels as one of the decisive factors. Thus the following hypothesis (H4)
was derived:
H5. The subjective norm has comparatively lesser influence on intention than attitude and
perceived behavioural control have, for considering carbon labels as a decision making tool.
2.2.6. Attitude towards carbon labels
Tesco CSR (2010) reported that, in their carbon label survey most of the consumers who
were aware of carbon label expressed positive attitude towards the labels (no supporting
result in the report). Berry et al. (2008) focus group research reported that 59% of the
participants expressed positive attitude towards eco-labels and were interested to know
how their purchase decisions may impact the climate change. Though, no quantitative
attitude measure had been carried out, the previous researches indicate that consumers
have an overall positive attitude towards carbon labels. Thus the following hypothesis (H5)
was derived:
Nitai Chand Patra 20
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
H6. Overall, consumers have a positive attitude towards carbon labels.
2.2.7. Attitude-Behaviour gap:
It is well known in social psychology and consumer behaviour literatures that there is a
substantial gap between consumer intention and their behaviour (Young et al. 1998). It has
been intriguing topic for study how attitude influences behaviour via intention in different
contexts. Young et al. expressed that the empirical evidences suggest that most of the time
the intention provides a biased (under or over) estimation of behaviour or purchase
propensity. Further, there are various other factors influencing the intention to behaviour
estimation, such as habit and impulse buying or unplanned shopping. Kemp et al. (2009)
research on motivational factors in UK supermarket purchases found that 8% of
respondents expressed that they bought a product or brand because they usually buy that.
Next, marketers and consumer psychologist agreed that a large proportion of supermarket
purchases is impulsive or unplanned. Phillips and Bradshaw (1993) found that almost 50%
purchases are impulsive, whereas Bellenger et al. 1978 found it to be 38.7% (Jones et al.
2003). It indicates that there is a significant influence of impulse buying on behaviour.
Schiffman and Kanuk (1994) mentioned that intention is a good predicator of behaviour.
However, Sutton (1998) review of various studies found that only 19-38% of variances of
behaviour are explained by intention. All these previous empirical evidence suggests that
there might be a similar gap between intention and behaviour in the context of carbon
labelling. It could be because of strong attitude and social norms, consumers have the
intention of considering carbon footprint while shopping, but their pro-environmental
intention is not efficiently converted into green purchase behaviour. Thus the following
hypothesis (H6) was derived:
H7. Consumer intention to consider the carbon footprint as a decision making tool
substantially differs from actual behaviour.
2.3.1. Motivating Factors:
Staats (2004) expressed that consumer’s pro-environmental behaviour can be understood
only if the knowledge of the social dilemma is applied simultaneously. Pro-environmental
behaviour is explained as the behaviour that is relatively favourable towards the
environment in comparison to another behaviour that serves the same primary purpose or
Nitai Chand Patra 21
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
function (e.g. using public transportation instead of using a personal car). Staats referred
social dilemma as the tension between the individual interest and collective or social
interest. As per the economic and cognitive theories, this social dilemma drives a person’s
trade-off situation between pro-environmental and non pro-environmental options. An
individual sacrifices his own interest in some degree in favour of the interest of the
community. E.g. a consumer buys a pro-environment product at £1.10, while he had the
choice to buy a comparatively less environment friendly product with equal other product
attributes at £1.00. Here the consumer sacrificed £0.10 in favour of interest of environment
and this is the trade-off.
It is established that consumers hold high regards for eco-friendly products; however, their
buying behaviour are often inconsistent with their values (Uusitalo 2008). Many people
consider themselves as environment conscious consumer and express willingness to buy
products with minimum environmental effects, however, the link between their intention
and behaviour is weak. However, their purchase decision is often guided by other factors
such as brand, quality, price and individual buying habit (Horne 2009). Even the most
environment conscious customers do not choose a product, only considering a product’s
environment aspects, they trade-off between several attributes of the product. So it has
been difficult for marketers and policy makers to determine whether environment
friendliness is an important product attribute for consumers. (Uusitalo 2008)
In the economic and cognitive psychology convention, it is accepted that consumers behave
rationally based on their preferences and beliefs. Further, studies confirm that, consumers
have high preferences for eco-friendly products; however, the link between attitude and
behaviour is weak. Customers make trade-offs in their everyday purchasing, for example, on
price versus quality/utility or availability. Nevertheless, sustainability trade-offs such as
choosing an eco-friendly product is often complicated.
Consumer behaviour varies a great deal depending on the product and its function. While
making purchase decision customers evaluate perceived value using extrinsic cues such as
price, promotion, brand & label information and intrinsic cues such as quality, specifications
and physiological characteristics (Zeithaml 1988). Further, Zeithaml expressed that extrinsic
cues are used more often when there is low product differentiation or there is a difficulty in
evaluating the quality of products.
Nitai Chand Patra 22
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
In their research, Kemp et al. (2009) found that price and brand are the two major
motivating factors for UK supermarket consumers. During a pre-survey of this research, 80
UK supermarket buyers were asked to rank various product attributes. Among the extrinsic
cues, the price and brand received the highest ranking with relative importance 82.56% and
70% respectively. The results are consistent with the findings of Kemp et al (2009). The
detailed results of Kemp et al. (2009) and the pre-survey are presented in appendix-A4,
page.87.
Rokka & Uusitalo (2008) expressed that environmental friendliness is an important product
attribute and worth evaluating its position in consumer multi-attribute motivation models.
The urgency for controlling climate change and consumer’s role in that suggest that carbon
label should be added into consumer choice models as a relevant product attribute.
Considering Zeithaml’s (1988) extrinsic cues, a comparison among the price, brand and
carbon labels was an interesting experiment to learn consumers’ decision making process.
Environment consciousness does not automatically lead to pro-environment behaviour
(Horne 2009). According to research finding by Morris (1997), eco labelling might guide
consumers toward environmental friendly product is a weak assumption (Horne 2009).
According to the economic theory (Lancaster 1966), humans are rational and their decision
depends on the maximisation of their utility. So it implies that, price and brand will maintain
the leading position as prime motivators in decision making process and carbon label will be
next to them. Thus the following hypothesis (H8) was derived:
H8. Price and brand would maintain their leading positions as prime motivators followed by
carbon label.
2.3.2. Role of communication
Over the last 2-3 years, the concern over climate change has driven consumers towards
greener products (products with a lesser carbon footprint) and sustainable lifestyle (Horne
2009). Although there is a widespread concern about the environment damages, many
consumers fail to understand the damages caused by their own consumption. The eco-
labels such as carbon labels play an important role in informing consumers the
environmental consequences of their purchase decisions. Bech-Larsen (1996) expressed that
one of the possible reasons for the attitude behaviour gaps is that not too many products
Nitai Chand Patra 23
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
carry any eco-labels. Another possible reason Uusitalo (1989) expressed that it is a general
tendency that consumers undervalue their marginal contributing to the problem.
Berry et al. (2008) expressed that until now, the climate related product information have
been used as a brand distinguishing factors mainly to appeal the already climate conscious
consumers, rather than as a mainstream evaluation tool. Further, carbon labelling is just a
beginning to fill this gap. Moreover, once the link among the carbon labels, products and
environment impact explained, a majority of the focus group participants expressed positive
interest and engagement with carbon labels. Moreover, they suggested further research to
assess the role of carbon labels and to find out whether communicating carbon impact of
products make any difference in consumer behaviour, and to know if they make use of the
tool.
Vanclay et al. (2009) found that the carbon labels have potential to become mainstream
toolkit if consumers are reminded of the importance & use of labels at the time of purchase.
In their study in an Australian convenience store, the customers were presented with
leaflets on carbon label, explaining its significance and role in climate change. The result
shows that there was an overall 6% decrease in sales of high carbon emitting products and a
4%increase of low carbon emitting products. It implies that, communication plays an
important role in consumer pro-environment behaviour.
According to behavioural science, consumer’s decision is driven by several factors such as
personal, marketing mix, psychological, socio-cultural, social and situational. The
communication of importance of carbon label at the time of shopping can act as a strong
social marketing mix and/or social and/or situational and drive consumers to use the carbon
label as an important product evaluation tool. The findings suggest that there is a high
probability that UK consumers would also act eco-responsibly and use the carbon labels as a
product evaluation tool if communicated adequately and given enough options. Thus the
following hypothesis (H8) was derived:
H9. When informed about the significance of carbon footprint and labelling, consumers use
carbon labels as an important tool in their buying decision.
2.3.3. Awareness Level
EU Flower Label, an eco-label, signifies a product’s kindness on the environment. Only the
best products which are kindest to the environment are entitled to carry the label. In a
Nitai Chand Patra 24
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
study of awareness of EU Flower Label, it was found that nearly 48% of 24,000 respondents
didn’t know about the label, despite a well-funded information campaign. (European
Commission 2007, cited by Horne 2009)
In a survey conducted in Tesco, half the respondents responded that they understand what
carbon label represents (Tesco CSR 2010). As this study was limited to Tesco consumers only
and numbers of respondents in the survey were also unknown, a review of this finding will
be helpful to confirm the percentage of population in general is aware of carbon labelling.
Thus the following hypothesis (H9) was derived:
H10. Half of the retail consumers are aware of carbon labelling.
2.3.4. Proposal of TLS based carbon label
Focus group research by Berry et al. (2008) and Upham et al. (2009), indicate that the most
popular label format would be traffic lights (where green indicates ‘low-carbon’, amber
‘medium-carbon’ and red ‘high-carbon’). Some participants expressed that the current
numerical presentation of the carbon footprint, does not make much sense for them, as
they do not know what value is good, bad or optimum. Further some participants had the
misconception that the presence of carbon label suggests that the product is environment
friendly. However, it is a fact that the presence of carbon labels do not confirm environment
friendliness, it only informs the amount of green house gas emitted. The findings of these
studies imply that, the present carbon label is not effective enough and integration of TLS
can improve its effectiveness.
The main advantages of TLS are that: they are simple, consumers are already familiar with
such labels, and they are intuitive in nature. When consumers are presented with complex
product choices with lots of product information, they are less able to make informed
purchase decision. And in ideal shopping scenario, it is a fact that consumers are
continuously bombarded with a large amount of product information starting from brands,
nutritional values, price, source, carbon footprints, organic, fair-trade and many more. All
these information makes the decision making process further complex. Black and Rayner
(1992) confirmed the consumers struggle when they are presented with lots of nutritional
information. It implies that simplification of product information will be beneficial.
Lang (2006) mentioned that in many researches it has been found that consumers find it
easier to follow the traffic light system in comparison to other labelling formats (Fraser et al.
Nitai Chand Patra 25
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
2009). In their research on the effectiveness of traffic light labels on food products, Fraser et
al.(2009) found that participants avoided foods with red lights from a mixed basket of food
products. Further they expressed that the use of label information has been observed to
alter overall food purchase behaviour.
On the other hand, Verbeke (2005) expressed that although consumers may prefer simple
label formats on the pack of products, however, it does not mean they behave to it in the
manner it was indented (Fraser et al. 2009). Because when it comes to using of the label,
the taste, brand, price and other attributes may override the purpose of the label and
making it ineffective. Further, TLS alone is ineffective in delivering the complete information
and lack of clarity. What is good and how is it compared? However, a focus group
participant expressed “I think the traffic lights are the best because it is really clear and
obvious. I know it doesn’t give stats, figures and numbers, but I don’t think we understand
the stats, figures and numbers anyway.” (Berry et al. 2008)
Consequently, the combination of TLS and current carbon label can be effective and address
all the concerns as it provides necessary information & ease of interpretation. All these
findings suggest that the TLS based carbon labelling format would be more effective in
influencing consumer decision making in a realistic choice environment. This leads to the
final hypothesis (H11) of this study:
H11. Integrating traffic light signals with the present carbon label would enhance
effectiveness of the carbon label.
Nitai Chand Patra 26
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
3. Research Design & Method
This section presents the development of the questionnaire, discussions on the adopted
quantitative and cross-sectional methods of research, analysis technique, data collection
and description. Structural Equation Modelling (SEM) was adopted for testing and analysis
of relationships between different elements of the model. Choice based conjoint analysis
was used to calculate the part-worth utility of different product attributes.
The questionnaire was developed in four parts. Part-I contained a general question on
carbon labelling followed by a detailed explanation about carbon labels and guidance for
interpretation. Part-II contained questions related to TPB model, Part-III contained
questions related to conjoint analysis and conjoint question for evaluating usefulness of
traffic lights with the carbon labels. Part-IV contained demographic questions and open
ended feedback.
3.1. Part-I General Questionnaire
A short qualitative survey (20 respondents) was conducted with few volunteers and
supermarket customers to find what consumers think the carbon label represents. The
responses were classified and categorised into five definitions (one of which was the correct
definition). In the development of the final questionnaire, for the question what carbon
label represents, these five definitions, along with “I do not know” and “others” options
were used.
3.2. Part-II Structural Equation Modelling and Questionnaire development
This part discusses structural equation modelling and questionnaire development for testing
of the proposed model and related hypotheses. The constructs for measurement were
attitude, behaviour, intention, subjective norm and perceived behavioural control. These
elements, like other psychological constructs are latent and cannot be observed directly.
3.2.1 Structural Equation Modelling (SEM)
“SEM is a powerful statistical technique that combines measurement model or
confirmatory factor analysis (CFA) and structural model into a simultaneous statistical test”
(Loon 2008). SEM is a multivariate statistical modelling technique, which allows researchers
Nitai Chand Patra 27
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
to examine more than one relation simultaneously using multiple regression equations
concurrently, whereas the regression analysis looks at only one relation or equation at a
time. Since no factor exists alone and often there are complex interactions among the
factors in attitude-behavioural studies, use of SEM is more realistic, effective and efficient.
SEM is more versatile than other multivariate analysis as it allows simultaneous analysis of
relationships between variables (Maxwell 2009, Loon 2008). Further, SEM takes potential
measurement errors into account, while regression does not.
The factor analysis can be performed in two ways: Exploratory factor analysis (EFA) and
Confirmatory factor analysis (CFA). In EFA no prior restrictions are set and the relation
between the observed and latent variables are explored. Whereas in CFA, the researcher
needs to specify the factors, expected relations and pattern of indicating factor loadings,
which is generally done based on previous research findings. SEM is a confirmatory method,
as the modeller is required to define the relationship between variables. The results
obtained from EFA are exploratory in nature and are often unreliable; use of CFA provides
better and reliable results (Maxwell 2009). In this study, a CFA using SME had been adopted
to test the proposed model.
SEM employs covariance analysis method for estimation. The goodness-of-fit tests are used
to determine whether the research model is consistent with the variance-covariance
pattern in the data. Further SEM specifications and criteria help the researcher in
determining an optimal model from a set of competing models. Though SEM is a relatively
new technique developed during 1970s, it has been used extensively in the study in
psychology, sociology, biological sciences and market research (Golab 2003).
Few of the several advantages of SEM over the linear-in-parameter statistical methods
presented by Golob (2003) are:
i. Test of the overall model by considering multiple equations, rather than just
computing the coefficients individually.
ii. Modelling of intervening or mediating variables.
iii. Direct, indirect and total effects are important distinguishing features of SEM. The
direct effects are the regression weights, referring one variables direct effect on
another variable along the specified path as per the model. Indirect effects represent
the sum of all effects between two variables that involve some intervening variables.
Nitai Chand Patra 28
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
General regression models compute only the direct effect of one variable on
another; however, SEM provides information on the total effect, which includes both
direct and indirect effects.
iv. Testing of coefficients across multiple groups in a sample.
v. As most of the behavioural research data are non-normal, use of SEM is very
effective in those cases.
vi. Separation of measurement errors from specification errors.
Therefore, considering all these benefits of SEM, in this research SEM was used to test the
models.
3.2.2. Questionnaire Development
Pro-environmental attitude measure is typically operationalised by blending statements
concerning a variety of environmental issues (Follows & Jobber 2000). Samuelson and Biek
(1991) argued that a significant correlation between attitude and behaviour could only be
obtained if both attitude and behaviour measures correspond to a specific issue or object.
Hines et al. (1987) found that the attitude-behaviour relationship is weaker when general
attitude towards the environment operationalised instead of specific environment related
behaviour. So in this case as the behaviour is related to a specific act of considering carbon
footprint while shopping, the attitudinal measure should relate to carbon labelling. As no
pre-validated scales were available for measuring attitude-behaviour relationship towards
carbon labels, a scale was designed and made operational.
The questionnaire was developed following the literatures on environmental behaviour and
previous application of TPB such as Ajzen 1991 and Tonglet et al. 2004. In order to develop
an effective scale multiple pilot tests were performed. In the initial rounds, a wide variety of
questions with a mix of negatively worded questions were presented to increase the validity
of responses. However, many respondents reported fatigue and dropped out. Further, in
the reliability analysis lower reliability of scale (Cronbach alpha) were discovered. To
increase the reliability of the scale, multiple changes were made and all the questions were
made positively worded to increase participation and decrease fatigue.
The final questionnaire for measuring elements of the model contained 20 items. According
to the model, the questions broadly fell into five categories as presented in the table-1
below.
Nitai Chand Patra 29
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Table-1: QuestionsAttitude towards carbon label (ACL) (Reliability coefficient > 0.7) ACL1: I would prefer to buy products with low carbon footprints. ACL2: I am most likely to pay a marginally higher price for an eco-friendly product. ACL3: I consider the carbon footprint as a major product attribute in my purchase decision. ACL4: The carbon label provides satisfactory information about a product’s impact on environment. ACL5: I appreciate retailers’/manufacturers’ initiative for carbon labelling of products.
Subjective Norm(SN) (Reliability coefficient > 0.7)I choose/ will choose a low carbon emitting product because:
SN1: Climate change is a global concern and a collective responsibility.SN2: Some retailers are providing additional incentives. (e.g. Tesco Green Club Card Points)SN3: People who are important to me expect me to use low eco-friendly products.SN4: I am contributing to a higher purpose.
Perceived Behavioural Control (PBC) (Reliability coefficient > 0.7)PBC1: It is convenient to compare carbon footprints on products.PBC2: There are reasonable options to choose a low carbon footprint product.PBC3: I know where to look for the carbon label on the products.PBC4: I know how to interpret the carbon label on a product.PBC4: The higher price of eco-friendly products does not abstain me from buying them.
Intention(INT) (Reliability coefficient > 0.7)Considering my last five shopping trip intention,
INT1: I had intentions for considering carbon footprint while buying products.INT2: I had intention of comparing the carbon footprints of the products before buying them.INT3: I intended to buy at least one product with a comparatively lower carbon footprint.
Behaviour (BEH) (Reliability coefficient > 0.7)In the course of last five shopping trips,
BEH1: I did consider the carbon footprint of products while buying them. BEH2: I have compared the carbon footprints of products before buying them.BEH3: I have bought some products with a comparatively lower carbon footprint.
3.2.3. Instrument
Nitai Chand Patra 30
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
As recommended by Ajzen (1991), a seven point bipolar scale was used for measurement.
Krosnick (2010) has expressed that bipolar scales perform best with seven points, whereas
unipolar scales performed best with five. For each question, the respondents were asked to
choose an option from a Likert-type scale with choices “Strongly Agree” to “Strongly
Disagree”. The following coding was used for statistical analysis, where a higher score
represents more favourable rating towards the concept being measured.
35 respondents expressed that they had no intention of considering carbon labels for their
purchase decisions, whereas 35 respondents expressed neutrality. The remaining 138
respondents expressed that they had some intentions of considering carbon label in their
purchase decisions, while 31 respondents expressed strong intentions. The mean intention
score was 0.8607, which shows an overall moderately positive intention of respondents for
considering carbon label as one of the deciding factors in their purchase decisions.
Nitai Chand Patra 42
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
4.2.6. Behaviour score (BEH) distribution
Figure 11: Score distribution of BEH
102 2
26
8 513 12 11
32
125
16 167
19
6 2 405
101520253035
No of respondents
89 respondents expressed negative behaviour. 32 respondents expressed neutrality. Only
87 respondents expressed that they have considered carbon footprint of products in their
last five shopping trips. Only twelve participants expressed strong positive behaviour. The
mean behaviour score is -0.0533 ≡ 0. The overall behaviour score of participants in the
survey reflects neutral behaviour in using carbon labels as a decision making tool.
Nitai Chand Patra 43
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
4.3. Structural Equation Modelling Output
Structural equation modelling using Amos-18 was used to estimate the measurements and
model fit for proposed model. Garver and Mentzer (1999) have suggested that in SEM the
indicators of a model fit are the Chi-square normalized by degrees of freedom (χ2/df),
goodness of fit index (GFI), adjusted goodness of fit index (AGFI), non-normed fit index
(NNFI), comparative fit index (CFI) and root mean squared error (RMSEA).
A low or non-significance χ2 signifies good fit, as it confirms that there is no difference
between actual and predicted matrices. Χ2 is sensitive to sample size, so for larger samples,
Kline (1998) suggested that Chi-square normalized by degrees of freedom (χ2/df) should not
exceed 3 for good model fit (Loon 2008). Garver and Mentzer (1999) have recommended
that GFI, AGFI, NNFI and CFI should exceed 0.9 and RMSEA should not exceed 0.08 (Loon
2008). Following these recommendations, the fitness of the proposed model was evaluated.
Goodness of fit of proposed model:
Table-6 : Model fitness indices of proposed modelCriteria Recommended value Obtained value Fitnessχ2/df <=3 (p<0.05) 13.29/2 Not achieved.GFI >0.9 0.9757 Achieved.AGFI >0.9 0.8180 Not achieved.NNFI >0.9 0.9786 Achieved.CFI >0.9 0.9815 Achieved.RMSEA <0.08 0.1651 Not achieved.
The goodness-of-fit indicators indicate that the proposed model was not completely fit as all
the minimum requirements of a model fit were not achieved. To improve the fitness, the
estimates were recalculated with modification indices. Then AMOS recommended an
additional possible path and relation between ACL and BEH. By evaluating the
recommendation, the modified model was established.
Nitai Chand Patra 44
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
The Figure 3: The final model developed from SEM.
Table-7 : Model fitness indices of extended modelCriteria Recommended value Obtained value Fitnessχ2/df <=3 (p<0.05) 0.0217/1 (p<0.001) AchievedGFI >0.9 1.0000 AchievedAGFI >0.9 0.9994 AchievedNNFI >0.9 1.0000 AchievedCFI >0.9 1.0000 AchievedRMSEA <0.08 0.0000 Achieved
A χ2 value of 0.0217, df=1, p<0.001 of the modified model confirms that the sample data fits
the model, i.e. there is no difference between actual values and predicted values. Further,
all other indicators (Table-7) confirm the goodness-of-fit of data with the extended model.
Regression weights:
The table-8 presents the regression weights of different paths in the final model.
Table-8: Regression weights
Estimate Std. Estimate S.E. C.R. P
INT <--- ACL .1923 .1464 .1083 1.7757 .0758
INT <--- PBC .3508 .3163 .0768 4.5668 ***
INT <--- SN .4279 .3169 .1044 4.0974 ***
BEH <--- INT .2672 .2381 .0654 4.0864 ***
BEH <--- ACL .3390 .2299 .0916 3.7019 ***
BEH <--- PBC .5226 .4199 .0783 6.6784 ***
***p< 0.001.
Nitai Chand Patra 45
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
The R-square value for intention (INT) is 0.4741. This suggests that, the predictors of
intention: attitude, subjective norm and perceived behavioural control can explain 47.41%
of the variances in intention. The R-square value for behaviour (BEH) is 0.6004. This value
suggests that, the predictors of behaviour: attitude, intention and perceived behavioural
control can explain 60.04% of the variances in behaviour.
In the final model (Figure-12), attitude has direct effects on behaviour and also indirect
effects via intention. Similarly, PBC has both direct & indirect effect on BEH, whereas, SN has
only indirect effect. The total effects represent the sum of direct and indirect effects. Tables-
9 & 10 present direct, indirect and total effects of different variables on INT & BEH
Table-11 (a): Attribute utility from conjoint analysis
AttributeRelative Importance
Attribute level wise utility
Brand 4.34%Tropicana Del Monte Princes2.237 2.383 2.151
Price 44.66%£0.80 £1.00 £1.203.530 2.788 1.136
Carbon footprint
50.99%960gm 1000gm 1200gm3.759 1.947 1.027
Table-11(b): Attribute utility from traffic light conjoint analysis
Attribute Relative Importance
Attribute level wise utility
Brand 46.77Tropicana Del Monte Princes1.336 2.250 3.812
Price 53.23%£0.80 £1.00 £1.203.816 1.913 0.998
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Nitai Chand Patra 50
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5. Analysis and Hypothesis Testing
5.1. Hypothesis Testing
Using the statistical results presented in the chapter-4, this section elaborates the testing of
the hypotheses proposed in literature review (section-2).
5.1.1. Testing H1. There is a positive relationship between attitude towards carbon labels
and intention towards using carbon labels as a decision making tool.
The ACL->INT regression estimate (0.1923, p=0.0758) suggests that the attitude-intention
relation is insignificant. However, this is against Armitage and Conner (1998) meta-analysis,
which reports significant attitude-intention relationships (mean r=0.49, N=115).
Nevertheless, the discrepancy can be explained in the light of an experiment and the
extended model.
In the experiment, the model was tried without the attitude-intention path. SEM could not
achieve the desired model fit. The model fitness results indicate the presence of attitude-
intention association, though weak. Hence, H1 is supported. In the extended model, SEM
suggested a new path between attitude and behaviour. The relation between attitude and
behaviour (r=0.3390, p=.000) is significant. Ajzen (1991) argued that in circumstances where
the intention-behaviour relationship is not strong enough, some factors may act directly on
behaviour. Depending on the context, they may act together or in different combinations. In
this model, attitude is acting directly upon behaviour as intention-behaviour relation is weak
(0.2672, p=0.000).
All these results indicate that consumer attitude towards carbon labels is not being able to
develop a strong intention of using the carbon labels. Two of the plausible explanations
could be as such. First, consumers do not feel morally responsible for their consumption and
perceive that their contribution towards climate change is marginal or insignificant. Second,
consumers may not have the right means or behavioural controls to convert their attitude
to intentions.
Nitai Chand Patra 51
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5.1.2. Testing H2. The subjective norms have a positive influence on consumer’s intention of using carbon labels as a decision making tool.
The SN-> INT regression estimate 0.4279 with p=0.000, suggests that subjective norm has a
positive and significant influence on intention. Hence H2 is supported. The relation between
SN and INT suggests that social pressure or normative beliefs have influence on one’s
intention for considering carbon labels while shopping.
The SN was estimated by asking questions such as participants’ responsibility towards
climate change (SN1), perception of retailers’ incentive on green products (SN2), approval or
disapproval of people important to the participant (SN3) and purpose (SN4). All the items
had significant (>0.7) individual loading (correlation) on SN. It implies that when consumers
feel that they are responsible for climate change, society would appreciate their eco-
conscious behaviour and retailers would provide additional incentives, they significantly
intent to consider carbon labels in their purchase decision.
5.1.3. Testing H3. There is a positive relationship between perceived behavioural control
towards carbon labels and the intention towards using carbon labels as a decision making
tool.
The PBC->INT regression estimate 0.3508 with p=0.000, suggests that perceived behavioural
control has a positive and significant influence on intention. Hence the H3 is supported.
The relation between PBC and INT implies that consumers’ perception of control over their
behaviour has a significant influence on their intention. The consumers who perceive that
comparing carbon footprints is easy and there are enough products with carbon labels, have
a significant intention of using carbon labels as a decision making toolkit. Those consumers,
who perceive that comparing carbon footprints is a difficult task and there are not many
options to perform the task, do not have an intention of considering carbon labels as a
decision making toolkit.
Nitai Chand Patra 52
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5.1.4. Testing H4. There is a positive relationship between perceived behavioural control towards carbon labels and the behaviour of using carbon labels as a decision making tool.
The PBC->BEH regression estimate 0.5226 with p=0.000, suggests that perceived
behavioural control has positive influence on behaviour. Hence, H4 is supported. This
relation is in-line with Ajzen’s (1991) argument that, when prediction of behaviour from
intention is hindered, PBC facilitates the behavioural intention into action and predicts
behaviour directly. In the case of carbon labels, clearly the prediction of behaviour from
intention is low (r=0.2672), so the PBC facilitates the intention to action and acts directly
upon behaviour.
The result implies that, increased feeling of control and convenience of using carbon labels
will increase the extent to which consumers are willing to exert additional efforts on using
the labels as a decision tool.
5.1.5. Testing H5. The subjective norm has comparatively lesser influence on intention than
attitude and perceived behavioural control have, for considering carbon labels as a decision
making tool.
The SN->INT regression estimate (0.4279, p=0.000) is greater and significant than the
regression estimate of both ACL->INT (0.1923, p=0.0758) and PBC->INT (0.3508, p=0.000).
This suggests that subjective norm has higher influence on intention than attitude and
perceived behaviour control. Hence, H5 is not supported.
This result along with the covariance result of ACL<->SN (0.7935) and SN<->PBC (0.7658),
imply that normative belief and retailers’ incentive can affect intention, attitude and
perceived behavioural control. That is, when the social pressure is strong and incentives are
high, consumer intention grows accordingly, along with strengthened attitude and mental
preparedness to overcome the perceived difficulty of using carbon labels.
Contrary, Armitage and Conner (2001) meta-analysis found that SN is the weakest predictor
of INT. However, Armitage and Conner (2001) argued that, many studies failed to get the
right measure between SN-INT because of use of single item scale instead of multi-item
scales for SN. Trafimow and Finlay (1996) meta-analysis reported that there are significant
numbers of studies where the subjective norm was the primary driver of behavioural
Nitai Chand Patra 53
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
intention. So, these discussions imply that the relation between SN and INT found in this
study is valid.
5.1.6. Testing H6. Overall, consumers have a positive attitude towards carbon labels.
For 168 out of 208 (80%) respondents, the attitude score was positive. Further, 113
respondents had an attitude score higher than the mean attitude score (0.99). Thirteen
respondents had a neutral score and remaining twenty seven have a negative score. This
result indicates that majority of participants were positive about carbon labels. Hence, H6 is
supported.
Despite this overall positive attitude, some participants expressed extremly negative feeling
towards carbon labelling. One participant expressed that “ There is no point of England
making such a fuss over the carbon footprint because no other country bigger than England
is doing much to help. We really are just going around with a dustpan and brush after an
earthquake”.
Another respondent commented “what is the point in buying a low carbon footprint crisp,
while I drive miles to get them, rather using public transports and air travel thousands of
miles for holidays”. These comments imply that, though consumers in UK are positive
towards carbon labels, they believe that the relative contribution from their use of carbon
labels in alleviating climate change is insignificant, so carbon labels are not worth enough
for their attention.
5.1.7. Testing H7. Consumer intention to consider the carbon footprint as a decision making
tool substantially differs from actual behaviour.
A Wilcoxon Signed Ranks Test revealed a statistically significant difference between
intention (INT) and behaviour (BEH), Z=-8.68, p=0.000, with a large effect size (r=0.42). The
ranks suggest that out of 208 samples, in 127 cases BEH < INT, 26 cases BEH > INT and 55
cases BEH = INT. This reflects the inconsistency between INT and BEH. Further, the INT ->
BEH regression estimate was 0.2672, P=0.000, shows that the influence of intention on
behaviour is weak. So the intention is not a good predictor of behaviour. The mean scores
(Table-4, p.34) suggest that participants had a moderately positive intention of considering
carbon footprint of products in their shopping but in their actual shopping, they did not
exhibit the same. So there is a significant difference between consumer intention and
behaviour. Hence H7 is accepted.
Nitai Chand Patra 54
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Young et al. (1998) expressed that, in consumer psychology tradition it is a known fact that,
there is always a considerable difference between consumers’ intention and subsequent
behaviour. Uusitalo (1989, 1990) expressed that studies confirm that consumers have a high
attitude towards eco-friendly products, but the consistency between the attitude and
behaviour is bleak. The finding of this study on carbon labels implies the same.
Nevertheless, during SEM analysis the model was tested without intention mediating
between the ACL, SN, PBC and BEH. However, the required goodness-of-fit of the model
could not be achieved. This suggests that intention has a role in predicting the behaviour in
case of carbon labels, though the influence is bleak.
5.1.8. Testing H8. Price and brand would maintain their leading positions as prime
motivators followed by carbon label.
According to the conjoint analysis result (Table-11(a) on page.49), the carbon footprint had
the highest relative importance (50.99%) followed by price (44.66%) and brand (4.34%). It
implies that carbon footprint was the higher motivating factor in the study. Hence the
hypothesis is not supported.
The result indicates some bias. Because Kemp et al. (2009) and the pre-survey results
suggest that price and brand are primary motivating factors (Appendix-A3 and A4, p.86).
The possible reasons for the brand receiving such a low importance could be the use of
packaged orange juice as the research instrument. As it is a very common product and the
brands used in the experiment have significant market presence, so participants didn’t pay
much importance to the brand. The attribute level utilities of all the brands also suggest the
same, as all the brands received almost equal utility score (Table-11(b) on page.49:
Tropicana-2.237, Del Monte-2.383 and Princes-2.151).
Next, price received significant importance but next to carbon label. The possible
explanation could be as follows. Firstly, as the respondents were aware of the research
purpose, they were more conscious of the carbon footprint. This kind of inclination has been
reported in many psychological experiments. Secondly, as the price levels of the products
were comparatively low (£0.80, £1.00 and £1.20) and participants’ income levels suggest
that most participants can afford the price variation, so participants paid a comparatively
higher importance to the footprint.
Nitai Chand Patra 55
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
The discrepancy can be further explored in the light of one respondent’s comment “I do
compare the carbon footprints for similar products provided, when they are costing the
same. However, if the cost of one is lower than the other, then I do not think for the carbon
footprint.” There were two more similar comments. This implies that, for many consumers
price will remain the most important motivator. Nevertheless, the carbon label can also
become an important motivator if the price difference between products is marginal and
brand differentiation is bleak.
5.1.9. Testing H9. When informed about the significance of carbon footprint and labelling,
consumers use carbon labels as an important tool in their buying decision.
The insignificant mean behaviour score (-0.0533) suggests that currently on an average
carbon label is not an important tool in the consumer decision process. However, in the
conjoint research, participants exhibited pro-environmental behaviour by giving carbon
footprints the highest relative importance. In the beginning of the experiment, the
importance of carbon labels and the way of interpreting the labels were explained to the
participants in a simple manner using examples, which is believed to be the prime reason
for the carbon footprint’s high relative importance score. This implies that if consumers are
informed or reminded of carbon labels, during their shopping, they consider it as an
important attribute. Hence H9 is supported.
One respondent commented “In fact, this survey helped me to understand the
representation of carbon footprint. Furthermore, this gave me a feel of basics on which I
should purchase the product.” There were four other comments on the similar line of
thought. Another respondent expressed “Now I have better understanding of carbon labels,
and I will consider them henceforth. However, in order to sustain the good intention,
regular reinforcement is required.”
The findings imply that regular reinforcement through communication at the right time and
place can increase the influence of carbon labels and help achieve the purpose of carbon
labels.
5.1.10. Testing H10. Half of the retail consumers are aware of carbon labelling.
In the response to the question “What does carbon label represent?” 115 out of 208
participants chose the right and complete definition. This shows that almost the half of the
Nitai Chand Patra 56
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
participants were aware of the meaning of carbon label. Hence the H9 is supported.
However, there are chances of guess work and possibilities of hoodwink as the survey was
administered online, participants can easily do a web-search to find the right answer and
hence the H9 could also be inconclusive.
A significant number of participants (49) chose the option “Eco friendly”. Though currently,
carbon labels do not confirm the eco-friendliness of a product, many consumers perceive
that the presence of the label signifies products’ eco-friendliness. The similar misconception
is also reported by Berry et al. (2008). In the focus group, many participants perceived the
products or services with carbon labels are green or environmentally safe.
5.1.11. Testing H11. Integrating traffic light signals with the present carbon label would
enhance effectiveness of the carbon label.
Table-13 Attribute level utility from two conjoint analysisAttribute level wise utilityTropicana Del Monte Princes
First conjoint analysis 2.237 2.383 2.151Second conjoint analysis with TLS 1.336 2.250 3.812All the effects were statistically significant, yielding x2 with probabilities less than 0.01.
In the first conjoint experiment, all the three brands received an almost similar utility score
(Ref: Table-13). However, in the second conjoint analysis, there is a clear differentiation in
the utility scores. The Princes brand which was assigned a green label received the highest
utility score (3.812), followed by Del Monte which was assigned with a yellow label. The
Tropicana brand was assigned with red traffic label and received the least utility score
(1.336). This implies the influence of traffic light label systems on consumers’ behaviour.
Participants expressed distinctive support to the environment friendly product, due to the
presence of a green or yellow label. Hence, H11 is supported.
Vanclay et al. (2009) research on traffic light based carbon labels in Australia, confirms
similar findings. Sales for products labelled green in the experiment increased by 4%.
Whereas, the sales for products with black labels (black label represented the carbon villains
in the research) decreased by 6%. Further, no significant change in sales of products with
yellow label was reported.
Hence, traffic light integrated carbon labels are more effective than present carbon label.
Nitai Chand Patra 57
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5.1.12. Consolidated hypothesis testing results
Table-14 below presents the consolidated results of hypothesis testing.
Table-14: Consolidated hypothesis testing resultsNo Hypothesis Statistics ResultH1 ACL +ve related to INT B=0.1923, p=0.0758 SupportedH2 SN +ve related to INT B=0.4279, p=0.000 SupportedH3 PBC +ve related to INT B=0.3508, p=0.000 SupportedH4 PBC +ve related to BEH B=0.5226, p=0.000 SupportedH5 SN-INT < ACL-INT or PBC-INT Above statistics Not supportedH6 Most UK consumer +ve ACL 80% participants Supported
H7 INT differs from BEHWilcoxon Signed Ranks, Z=-8.68, p=0.000
Supported
H8Among Price (P), brand (B) and Carbon footprint(C), P and B are leading motivators.
P= 44.66%, B= 4.34%, C= 50.99%
Not supported
H9 Communication can improve BEH H7 statistics SupportedH10 Half the consumers aware of carbon label 115 out of 208 Supported
H11Inclusion of TLS in current carbon label will enhance effectiveness
Table-13Supported
Nitai Chand Patra 58
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5.2. Additional Discussion
This section (5.2) discusses additional findings from the SEM and association of background
variables with carbon label related behaviour.
5.2.1. The model
Among the various factors influencing intention and behaviour, perceived behavioural
control has the highest influence on behaviour (r=0.5226) and second highest on intention
(r=0.3508). In this research, the items measuring PBC, tried to access whether consumers
are comfortable with locating & interpreting carbon labels and comparing carbon footprint
of products. The significant association between PBC and purchase behaviour suggests that
consumer behaviour is largely influenced by their knowledge about the carbon labels, and
then followed by attitude. Whereas, the intention is majorly influenced SN, and then
followed by PBC.
Direct and Indirect effects: As discussed earlier the direct & indirect effect results are one of
the distinctive features of SEM. The results (Table: 12) suggest there is no direct effect of SN
on BEH. However, SN has indirect effects on BEH (0.1143, p=0.000). This suggests that
consumers’ normative beliefs do not directly lead to pro-environment behaviour.
Next the results suggest that ACL has both direct (0.3390) and indirect (0.0514) effect on
BEH. In the proposed model as per TPB (Ajzen 1991) there is no direct path between ACL
and BEH. However, the direct effect is statistically significant. The results indicate that ACL,
SN and PBC are responsible for intention and ACL, PBC and INT lead to behaviour. Further
these effects suggest that, intention is fully mediating SN towards behaviour related to
carbon labels and partially mediating ACL and PBC.
The developed model suggests that environmental consequences and subjective norms
have a significant influence on purchase intention. However, SN has minimal influence on
behaviour as discussed. Pickett-Baker and Ozaki (2008) also stated the same that, the values
and beliefs about environmental issues and consequences have no direct link with
environmentally responsible behaviour. Staats (2004) expressed the same that the link
between environment concern and pro-environment behaviour is generally weak.
Nitai Chand Patra 59
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Finding of Follows and Jobber (2000) research on environmental responsible purchase
behaviour supports the results of this study. According to their model, behaviour is
influenced by intention and intention is influenced by environmental consequences and
individual consequences. Environmental consequence refers to various environmental
impacts of a purchase decision; whereas the individual consequence refers to measures of
convenience, range of product sizes, cost and efforts to follow certain behaviour. The study
found that individual consequence (B=0.63, t=8.57) has marginally higher influence on
purchase intention than by environmental consequences (B=0.55, t=7.49). Follows and
Jobber’s individual consequence measure is equivalent to PBC in this study. The findings of
this study matches with Follows and Jobber’s study that, PBC has higher influence on
behaviour than attitude and subjective norm have.
Some of the other studies where the PBC individual consequence found to have a significant
influence on behaviour are as follows. Domina and Koch (2002) in a study on textile
recycling behaviour found that the convenience and closeness of drop-off centres is closely
related to recycling behaviour (Ramayah et al. 2009). Therefore, the results of this research
supported with the finding of above cited researches, confirm that the convenience of using
carbon labels is the greatest predictor of use of carbon labels as a decision making tool.
Nitai Chand Patra 60
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
5.2.2. Discussion on background variables
Uusitalo (2008) expressed that background variables such as age, gender and education
level are very weekly associated with the pro-environment attitude. However, a study
conducted in Finland by Uusitalo & Rokka (2008) for consumer preference for eco-packaging
confirms that on an average older member and females choose eco-friendly packaging.
Further, there has been some consensus between researchers that environmental friendly
consumers tend to occupy certain demographic characteristics as highly educated,
knowledgeable, relatively high income and more likely to be female and younger (Carrigan
and Attala 2001, De Pelsmacker et al. 2005 and Uusitalo 2008).
The table-15 below presents the statistical results in relation to exploration of difference of
carbon label associated behaviour across various background groups.
Table-15: Results of association of background variablesGrouping variable
Groups Test performed Test result
Gender
Gender: Male, FemaleBEH: Continuous to seven groups with equal range. Chi-Square
Mitchell,K.M., Stewart,M.D. and Yates,J. (2009). Customer response to carbon
labelling of groceries. Submitted to Journal of Consumer Affairs
Verbeke, W.,(2005). Agriculture and the food industry in the information age. European
Review of Agricultural Economics, Vol- 32, No-3, pp. 347–368.
.
Weatherell, C., Tregear, A. and Allinson, J. (2003). In search of the concerned consumer: UK
public perceptions of food, farming and buying local. Journal of Rural Studies Vol-
19, pp. 233–244.
Wüstenhagen, Katharina S. and Rolf. (2005). The Influence of Eco-Labelling on Consumer
Behaviour – Results of a Discrete Choice Analysis. Business Strategy & the
Environment, Switzerland: Business Strategy & the Environment, Sept 2005.
Young, M.J., Desarbo, W. and Morwitz, V. (1998). The stochastic modeling of purchase
intentions and behavior. Management Science, Vol- 44, pp. 188–202.
Zeithaml, V. (1988). Consumer perceptions of price, quality, and value: a means-end model
and synthesis of evidence. Journal of Marketing, Vol-52, pp.2–22.
Nitai Chand Patra 76
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Appendix
A1. Sample Questionnaire
Hello,
The climate change has been a global concern and as per Carbon Trust (2006), 45% of green house gasses generated in UK are from what people buy and use. The present survey is a part of an academic research (Durham Business School) to understand the current attitude and behaviour of consumers towards eco-friendly consumerism. You are invited to participate in our survey.
This survey includes simple questions, and does not require any specific knowledge or skills. It takes approximately 10 minutes to complete the questionnaire. Your participation in this study is completely voluntary. You can withdraw from the survey at any point. We are interested in your personal opinions regarding green consumerism. There are no correct or incorrect responses; we are merely interested in your personal point of view. Your survey responses will be strictly confidential and data from this research will be reported only in the aggregate.
If you have any questions or comments about the survey or the procedures, you may contact Nitai Patra by email [email protected] or can fill the feedback form at the end of this survey. Please click Continue to proceed to the survey. Thank you. We appreciate your participation.
What does the Carbon Footprint label represent?1. I don’t know.2. Fair trade.3. Eco-friendly.4. Organically Produced.5. Something that is recyclable.6. Amount of green house gases left by this product during its production and transport.7. Other __________________________________________________
Carbon footprint represents the total carbon emitted during the production, transportation and consumption of a product. Many products now-a-days carry a carbon label. E.g. Tesco brand washing detergents, orange juice, potatoes & light bulbs, Walkers Crisps and Boots shampoos. The lesser the carbon emission of a product, the more eco-friendly is it. For example, a product with a carbon footprint of 1000gm/unit is comparatively more eco-friendly/greener than a product with a carbon footprint of 1200gm/unit. Please consider this definition for answering the following questions.
Please select a suitable option in the context of your retail grocery/departmental shopping such as day-to-day shopping in stores such as Tesco/ Boots/ Iceland/ Asda/ Lidl/ M&S etc. There are no correct or incorrect responses; we are merely interested in your personal point of view.
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I would prefer to buy products with low carbon footprints. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I am most likely to pay a marginally higher price for an eco-friendly product. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Nitai Chand Patra 77
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I consider the carbon footprint as a major product attribute in my purchase decision. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
The carbon label provides satisfactory information about a product’s impact on environment.
❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I appreciate retailers’/manufacturers’ initiative for carbon labelling of products. ❏ ❏ ❏ ❏ ❏ ❏ ❏
I choose/ would choose a low carbon emitting product because:
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
Climate change is a global concern and collective responsibility. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
Some stores provide additional incentives. (e.g. Tesco Green Club Card Points) ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
People who are important to me expect me to use low eco-friendly products. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I am contributing to a higher purpose.❏ ❏ ❏ ❏ ❏ ❏ ❏
Nitai Chand Patra 78
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Please choose an option which you would most likely to buy, considering that these are the only choices and information available for you. Further consider that all other functional attributes are same for all the products.
Nitai Chand Patra 79
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Nitai Chand Patra 80
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Please select a suitable option in the context of your retail grocery/departmental shopping.
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
It is convenient to compare carbon footprints on products. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
There are reasonable options to choose a low carbon footprint product. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Nitai Chand Patra 81
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I know where to look for the carbon label on the products. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I know how to interpret the carbon label on a product. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
The marginally higher price of eco-friendly products does not abstain me from buying them.
❏ ❏ ❏ ❏ ❏ ❏ ❏
In the course of last five shopping trips,
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I did consider the carbon footprint of products while buying them. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I have compared the carbon footprints of products before buying them. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I have bought some products with a comparatively lower carbon footprint. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Nitai Chand Patra 82
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Nitai Chand Patra 83
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Considering my coming three months shopping intention,
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I had intentions for considering carbon footprint while buying products. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I had intention of comparing the carbon footprints of the products before buying them.
❏ ❏ ❏ ❏ ❏ ❏ ❏
Strongly Agree
Agree Moderately Agree
UndecidedModerately Disagree
Disagree Strongly Disagree
I intended to buy at least one product with a comparatively lower carbon footprint. ❏ ❏ ❏ ❏ ❏ ❏ ❏
Which best describes your gender?1. Male2. Female3. Prefer not to say
Nitai Chand Patra 84
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
Which best describes your age?1. 19 or younger2. 20 - 353. 36 - 504. 51 or older5. Prefer not to say
Which best describes your current annual family income?1. £15,000 or less2. £15,001 - £30,0003. £30,001 - £45,0004. £45,001 or more5. I dont know.6. Prefer not to say.
Which best describes your education level?1. Primary2. Secondary3. High school diploma4. Bachelors5. Masters & above6. Prefer not to say7. Other ______________________________
Please use this space if you have any questions/comments/feedback on this survey.
Nitai Chand Patra 85
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
A2. Product Directory
Company Product description
Walkers All varieties of standard crisps sold in single packets
Tate and Lyle 1kg bag of granulated cane sugar
Tesco Range of toilet paper and kitchen roll
Tesco Milk: Skimmed, Semi-skimmed, Whole
Tesco Range of own brand laundry detergent
Tesco Range of chilled and long life orange juice
Tesco Range of light bulbs
Tesco / MMUK Jaffa Oranges / soft fruit
PepsiCo Quaker oats and Oat so Simple
Morphy Richards Range of Irons
Allied Bakeries Kingsmill wholemeal, white and 50:50 loaves
British Sugar A range of white granulated sugar - British Sugar - B2B
British Sugar A range of white granulated sugar - Silver Spoon - B2C
Levi Strauss A one off promotional bag
Haymarket Magazines – Marketing and ENDS report
Continental Clothing A range of over 800 t-shirts and other cotton apparel
Continental Clothing Woven bags (USA and Japan) and t-shirt internet retailing service
Aggregate Industries 3 varieties of paving products - Bradstones
Source: Carbon Trust, web: http://carbon-label.com/individuals/product.html Accessed on 11th July 2010.
Nitai Chand Patra 86
CARBON LABELLING IN RETAIL GROCERY INDUSTRY
A3. Kemp et al. (2009) survey result
The following table summarises the findings of the research of Kemp et al. (2009). The table presents purchase motivating factors reported by respondents in revealed preference survey as primary, secondary or tertiary factors they considered.