Sustainability and the Supermarket Shopper (Project FO0401)
Final Report
Claire May (PhD candidate)
Professor Andrew Fearne (Supervisor)
Dr Susan Hornibrook (Supervisor)
Centre for Value Chain Research
Kent Business School
University of Kent
Canterbury
Kent
CT2 7PE
August 2011
Table of Contents
Executive Summary………………………………………………………………………………………2
1: Introduction and research objectives 8
2: Carbon labelling – the evidence to date 8
2.1: Effectiveness of carbon labelling 9
2.2: The role of situational factors 10
3. Conceptual framework 13
4. Methodology 15
4.1 Exploratory research……………………………………………………………………16
4.2 Pilot intervention……..……………………………………………………………….. 17
4.3 Evaluation……………..…………………………………………………………………. 20
5. Key Findings 21
5.1: Focus groups 21
5.2: Pre-store intervention 24
5.3: In-store intervention 25
5.4: dunnhumby data analysis 25
6. Conclusions 27
7. Limitations and Further work 29
7.1 Limitations of the study……………………………...... …………….. 29
7.2 Further work………………………………...... ……… 30
References 32
Executive Summary
1. Introduction
Climate change is an increasingly prevalent and urgent topic of debate and there is little dispute that urgent reductions in greenhouse gas emissions are required. The food system contributes almost 20% of the UK’s greenhouse gas emissions, making it a significant area on which to focus in terms of reduction targets, and changes in food consumption are fundamental (Garnett 2008). This research, which was undertaken as part of a Defra-funded PhD studentship, sought to explore ways to change consumer food purchasing behaviour to become more environmentally sustainable. The launch of a carbon labelling trial of own-label products by the UK’s largest supermarket represented an opportunity to investigate situational factors in the context of carbon labelling and their role in facilitating behaviour change. This report presents the key findings of the research, the details of which are presented in the PhD (May, 2011).
2. Literature Review
The extant literature on ethical purchasing is far from conclusive. There is some evidence that the propensity to purchase ethical products increases with age (Hines & Ames, 2000) and income (Barnett et al, 2005) and is higher amongst women than men (Parker, 2002) and people with lower levels of formal education (Dickson, 2005), but other studies have found no significant correlation between demographic factors in general and interest in ethical products (De Pelsmacker et al., 2005; O’Fallon and Butterfield, 2005).
Research into the role of situational factors within the ethical purchase decision making process is limited but suggests that barriers to ethical purchasing include the limited availability of ethical products (Nichols & Lee, 2005), an excessive bombardment of ethical messages (Boulstridge & Carrigan, 2000) and scepticism towards ethical labels (Nicholls and Lee, 2006). There is also evidence that consumers are more likely to purchase ethical products when they are not required to pay more, sacrifice quality or make a special effort (Carrigan & Attalla, 2001, Vanclay et al, 2011).
There are two main arguments in relation to the effectiveness of carbon labelling, one relates to the efficacy of the approach and the second is the presentation/format of the carbon label itself.
Upham and Bleda (2009) highlight that one of the main concerns connected to carbon labelling is that it is of limited potential in reducing carbon emissions relative to alternative actions. Their view is that the footprint logo should be used as a badge of approval - indicating that the product is under-going a reduction in carbon. However, they still stress that changing the higher impact behaviours of consumers and production processes should be the priority.
Other research suggests that the current format of the carbon label is not optimal. There seems to be a consensus that a ‘traffic light’ type rating system would be preferred by consumers (Berry, Crossley et al. 2008). Boardman (2008) however highlights that the relative rankings can only be made if the carbon footprints of all products within a given range are known, which is currently not the case.
A report commissioned by the European consumer organization ANEC (Quack 2010), is particularly critical of carbon labelling and argues that the focus on reducing carbon may mean that already ‘good’ performers are not easily able to reduce their footprint any further but ‘bad’ ones could, meaning effectively that ‘wrong’ is rewarded. In contrast, the House of Commons Environmental Audit Committee report (2009) calls carbon labelling “crucially important” and suggest that there is the opportunity for it to change decisions made by consumers if it is implemented as part of a comprehensive, universal environmental label.
Upham, Dendler et al. (2010) suggest that there could be wider, indirect consequences from carbon labelling amongst consumers, for example by enabling conditions whereby more significant reductions in emissions are considered acceptable. However, they also warn that consumers might mistakenly think that such reductions are happening already.
3. Research Methodology
The primary research undertaken in this research comprised three phases: exploration, intervention and evaluation.
The first phase involved nine exploratory focus groups with women from three distinct lifestages (young adults, young mothers, older mothers) with varying degrees of interest in environmental issues – strong, moderate and none at all. The focus groups were designed to generate an understanding of food purchasing behaviour and an appreciation of current levels of awareness, understanding and use of carbon labelling.
Phase two comprised two interventions - pre-store (carbon footprint week in selected primary schools) and in-store (merchandising and signage in selected stores) - targeted towards young families and designed to raise awareness and understanding of carbon labelling and facilitate the purchase of low carbon products. Young families were chosen as the target group because they are a diverse segment that is amenable to change (due to the influence of their children), facing a range of constraints (stay at home versus working mums, shopping with/without children) and often demonstrating a higher level of involvement in food purchasing (Slama and Tashchian 1985, Verbeke and Vackier 2004, Sridhar 2007).
The impact of the interventions was evaluated using multiple measures – interviews with primary school teachers from the 15 schools who participated in the pilot, an on-line survey of 90 parents/guardians who participated in the pilot, an exit survey of 786 shoppers at the 8 pilot stores (4 in the North of England 4 in the South of England[1]) and analysis of dunnhumby (supermarket loyalty card) data pertaining to all carbon labelled products purchased before, during and after the intervention in the pilot stores. Access to dunnhumby data provided the opportunity to explore situational factors through the study of both self-reported/intended behaviour and actual purchasing behaviour. This is an important methodological innovation given the growing evidence of the gap between ethical purchasing intentions (and claimed purchasing behaviour) and actual purchasing behaviour (Nichols & Lee, 2006, Carrington et al, 2010, Bray et al, 2011).
4. Limitations of the study
This study is the first to combine measures of claimed and actual behaviour in the evaluation of a pilot intervention designed to change the awareness, attitudes and purchasing decisions of individuals (supermarket shoppers). This was made possible by the access the student was given to the dunnhumby (supermarket loyalty card data). However, this methodological innovation, combined with the limited resources available for the design and implementation of the pilot interventions, give rise to a number of limitations, which need to be born in mind when drawing conclusions and making recommendations from the key findings.
In terms of the overall research design the two major limitations are that: a) the study involves just one supermarket, albeit the largest in the UK, and b) the pilot interventions were implemented in only fifteen schools and four stores. The choice of supermarket was driven by the fact that a) they were the only one with a fully operational (and expanding) carbon labelling program at the time of the study and b) the dunnhumby data was available, enabling the analysis of actual purchasing behaviour. The limitation in the number of schools participating was the result of two constraints – financial and time. More schools could have been recruited had we a) been able to offer larger incentives to more schools spread across a wider geographical area, b) started the process of engagement towards the end of the preceding school year rather than in the middle of the year in which the pilots were planned and c) created a much higher level of awareness of the carbon footprint week, possibly with input from local and national media.
In addition to the limitations associated with the overall research design, there are many lessons that can be learned regarding the operationalisation, execution and evaluation of the interventions. Working with retailers represents a great opportunity but can also impose restrictions and reduce control. Given a supermarket’s sheer scale in terms of size and the host of competing situational factors that are likely, attention to should be given to making any materials or activities as prominent as possible. Similarly there should be recognition of the time and bureaucratic restraints that schools are under and make participation in any intervention as simple as possible. Finally, the complexity of designing, executing and evaluating pilot interventions involving multiple stakeholders (Government agencies, commercial businesses, schools) with divers interests (public policy, commercial, academic) should not be under-estimated and requires a substantial commitment of time and effort on the part of the researcher to ensure all stakeholders are sufficiently engaged and committed from an early stage and throughout the pilot process.
5. Key Findings
5.1 Focus Groups
The focus groups confirmed the importance of situational factors in determining food choices in the context of supermarket shopping. Most participants reported being highly influenced by what they see in store and in particular by offers and promotions. The shopping mission clearly influences what is purchased and where from, with many respondents highlighting the habitual nature of ‘bulk’ shopping for the household and the contrast in approach when purchasing food for a special occasion. The presence of children is also a differentiating factor, in terms of what is purchased and the key factors considered – most of the young mothers confessed to being predominantly driven by their children’s needs and wants and the presence of children during the shopping mission is major deterrent for exploring new products or giving additional consideration to brand choice as “pester power” all too often renders the supermarket trip a nightmare for young mothers.
The overall consensus was that the environment was not a primary consideration for participants when actually in the store, although many reported that it was something they thought about at home, particularly in relation to recycling. Where participants said that they were thinking of the environment in the store, this was often in terms of a dislike of excess packaging, rather than relating to specific types of products they would purchase.
When the discussion moved on to explore awareness of and attitudes towards carbon labelling the majority of participants reported no recognition of seeing the carbon label on products in Tesco, thus awareness of the existence of the carbon label was very low. The few participants that reported seeing the carbon label were often vague or hesitant as to what product they had seen it on.
Similarly, although there was some understanding of the carbon label’s connection with carbon footprints and the environment generally, understanding of what the carbon label meant and how participants were expected to respond to it was equally low. The concept of grams of carbon was particularly problematic for the participants, especially when trying to relate this to something quantifiable that they could understand. It therefore follows that not one participant reported the carbon label having any influence on their purchasing behaviour.
5.2 Pre-store Intervention
Fifteen schools from across the four geographical areas participated in the carbon footprint week (representing a response rate of just 6%), involving approximately 700 children. This resulted in 90 usable responses to the parent/guardian questionnaire, of which two thirds were from schools located in New Malden (Surrey).
The general consensus from the teacher interviews was that the children were enthusiastic and interested in the topic (the environment, climate change). For most children, the study of environmental issues was not a new subject but the carbon footprint was a new concept. Overall, teachers reported a reasonable level of understanding, although some children found the scientific elements or specific vocabulary challenging to grasp, particularly those in lower year groups.
Two of the teachers commented on children bringing carbon labelled products into school. One of these said: “They’ve got it into their heads to sort of look for it and are quite excited about it” . Another reported that children would have made parents aware either through looking for carbon labeled products when shopping or because of an additional homework task they were set requiring them to find carbon labeled products at home. This teacher also said that the children were talking more about the environment and the things that they could do to protect it.
The results from the parent/guardian survey indicate that respondents as a whole had a reasonable level of concern about the environment, a fair understanding of carbon emissions and a positive disposition towards carbon labelling. However, the majority reported that environmental issues had a limited impact on their purchasing behaviour.
When asked what supermarkets could do to encourage them to make (more) use of carbon labels, the majority supported the views expressed in focus groups, highlighting the need for greater awareness, better understanding and some direct incentivisation to purchase the lower carbon variants. Significantly, only a third of respondents supported the idea of choice editing – removing the highest carbon variants from the supermarket shelves.
5.3 In-store intervention
A total of 786 usable store exit questionnaires were completed and, without prompting, almost a third (249) of respondents said they noticed something different in the store. However, almost two thirds of these respondents cited something that had nothing to do with the carbon labelling! One in five respondents cited the leaflets and 10% recalled the staff wearing carbon footprint T-shirts.
When prompted, only 20% of respondents claimed awareness of Tesco’s carbon labelling initiative and when asked what it was there was widespread ignorance or confusion, with some making the link to food miles and local foods and others confusing carbon labelling with excess packaging, waste and recycling.