Design a coloured shelving system for grocery products based on carbon footprint,to help consumers make better-informed choices.
The proposal is to develop a database/interface which could be used to develop an in-store shelving system or on retailer's websites (which could allow consumers to filter/group grocery products) based on their carbon footprint into four basic colour groups: green, yellow, orange and red (with green having the lowest GHG impact and red the highest). It is suggested that this colour coding could help promote the message that for a lower-carbon diet, “green” products should be used dominantly, freely, whereas “red” products should be selected consciously/sparingly.
The carbon intensity/impact of people’s diets can widely vary, depending on its composition (including factors such as level of meat consumption, origin of goods, locality of produce). There is, therefore, significant opportunity to actively reduce the greenhouse impact of the food sector by promoting a more conscious and environmentally-friendly approach to selection of food products.However, it is often the case that consumers are poorly-informed of the carbon impact of their food choices.
Surveys have suggested that the majority of consumers seem willing to make more conscientious food choices if given sufficient, clear information on the carbon footprint of supermarket products. A survey by the European Commission reported that 72% of consumers questioned were in favour of carbon labelling on foods, and that over the last 10 years, 42% had changed their shopping habits due to environmental concerns. This same survey also found that 89% of consumers found existing footprint information hard to understand, making product comparison “difficult and confusing”.
A simple colour-coded system would allow time-strapped customers to quickly evaluate the carbon intensity of their purchases.
Category of the action
What actions do you propose?
Current carbon labelling projects are sparse, with very few products included in the system. For those which do feature on some labels, such as those run by the Carbon Trust, it’s easy to see how information is confusing to consumers (see images below). The Carbon Trust scheme provides the footprint of some products as g CO2e which often has little significance to shoppers. These measures are also inconsistent, with some footprints being given “per serving”, “per 250ml”, “per pint”, “per wash”, “per garment”. Other schemes have tried to inform customers of transport and product packaging impact, but again, this is given in an inconsequential measure such as fuel consumption in gallons or lb of packaging.
For consumers to make more informed food decisions, it seems necessary to develop a more consistent, standardized system of supermarket shelving, with easy clear classifications of food products depending on their overall footprint. This system should be universal, such that it’s easy for supermarket sellers to apply to their full product range, and most importantly, be quick and simple for shoppers to follow.
The main actions would be to develop a step-by-step process through which retailers could introduce a coloured shelving or grouping system for food and drinks products, to inform/educate customers on the carbon footprint of their purchases.
The progressive steps for development and integration that would be implemented is as follows:
1) Develop a broad colour categorization for food and drink products based on life-cycle analysis (LCA) data, with the ability to group products based on their carbon-intensity- ideally on a single-page sheet which supermarkets could widely distribute across their stores for consumers to consult and familiarize themselves with to make more informed choices (brief example shown below)
2) Develop a more in-depth model (with the steps to do so, discussed later), such that every individual product within a store can be colour-categorised.
3) Use this classification to set up a virtual carbon-shelving system on the retailer's websites. Where customers can currently filter or sort products by factors such as "price", this integrated application would allow customers to view by carbon-intensity. For example, they could filter for all the lower-carbon "green-coded" items. In this way, they could fill most of their shopping baskets with low-carbon, sustainable items, and choose sparingly/conscientiously only a few "red-coded" carbon-intensive products per shop.
4) Scale this colour system into a real carbon-shelving or labelling system in supermarket stores. This would be the maximum consumer exposure, whereby the carbon-intensity of every product is clear to customers in a simple, easy-reference way.
Behind-the-scenes steps to building the model
The steps involved in developing the model, algorithm and interface are described below. After developing an effective user-friendly application, the proposal would be pitched to a range of supermarket retailers with the aim of persuading them to implement the adoption steps as discussed above.
The steps for building the model would be, as follows:
- Using life-cycle analysis data, evaluate different shelving criteria to provide sensible differentiation of products to arrive at a defined shelving criteria (e.g. >6kgCO2e/kg product= “red”). This may be different for different food/drink groupings (which could vary based on volume, e.g. high volume drinks vs. small portion condiments).
- Identify the fewest number of variables, which can allow for the differentiation between groupings for be determined e.g. meat content, transport distance and shelf-life (i.e. whether air-freighted or not), refrigeration, packaging etc.
- Derive an algorithm which could be applied to any product in order to allocate it to a coloured virtual shelf. This should require the minimum number of input details about the product (e.g. distance travelled, shelf-life, packaging weight) such that a seller can quickly allocate a new product to one of the virtual shelves.
- Design a preliminary, working user-interface as an example of how the model could be used in practice by a user.
The seller would have a simple, quick database entry to carry-out for each product, with an example of relevant information retailers would enter/answer, shown below:
The output of the model would simply tell them the colour-coding of the product.
For consistency and fairness, all products within the store would be classified via the same simple checklist system. If a producer disagreed with their footprint rating, they would be required to carry out their own complete life-cycle analysis of their product to prove its emissions fall within a lower threshold. If consumers were more inclined to choose products with a lower footprint rating, it could also help motivate producers to reduce their environmental impact, either by completing an in-depth life cycle analysis themselves or implementing reductions in areas such as ingredient sourcing or packaging.
Who will take these actions?
The model could be developed by myself as part of an MSc dissertation project (or done outwith the dissertation as a side-project) using LCA data for food and drinks products which have been developed in sufficient detail for an effective algorithm to be derived and interface designed.
The other key players would be supermarket retailers themselves. The main task of the proposal would be to convince sellers to include this type of interface in their stores and online.
Such projects have been tried/implemented by organizations previously. For example, GoodGuide rates products, with one of their classifications being "Environmental impact". This rating does however include a range of environmental factors such as water resources, air pollution etc. which would not be factored into the proposed classification- it would be based on a much simpler carbon basis. GoodGuides also rates food items on a company-level basis rather than a product-basis. For example, if a farmer implements good environmental practices, the resultant beef products could rate highly e.g. scoring a 9 or 10. This is somewhat misleading to consumers who might assume that buying large quantities of beef is sustainable, when in reality, it has a high carbon-intensity and should be purchased consciously/sparingly.
Where will these actions be taken?
This proposal could essentially be carried out anywhere across the world where sufficient LCA data for food and drinks products is available. This would predominantly exist in the developed world, where LCA analysis would be more widely available (and consumers' diets are more carbon-intensive due to high meat consumption and imported goods).
LCA data is already available and has been done in-depth for at least one UK retailer. The main UK retailers such as Tesco, Asda, Sainsbury's, Co-operative, also show some signs of trying to inform their customers of the impact of their choices through ethical and environmental labelling (such as welfare appraisements on egg, dairy and meat products). This has primarily been introduced from consumer drive and pressure for ethical and environmental assurances. If a slightly more ethically-driven retailer (such as the Co-operative) were to start providing such information on carbon-intensity and it was well-used by some consumers, this could put significant pressure on other retailers to follow.
How much will emissions be reduced or sequestered vs. business as usual levels?
The contribution of agriculture and food production to greenhouse gas emissions is often underestimated and/or overlooked in the context of carbon reduction and mitigation strategies, despite its crucial role. The global food system accounts for more than 30% of total anthropogenic GHG emissions.
The emissions resulting from a person's diet varies by 2-3 multiples depending on the products they consume. The amount of GHG emissions reduced would depend on how much of an influence it had on consumer choices. But even taking a estimate of suggesting it might reduce a meat-lovers person's footprint by 15% would reduce their consumption by 0.5 tonnes CO2e/yr.
Multiplying by the population of the UK (63 million), would result in a saving of about 32 million tonnes CO2e/yr.
This is obviously based on many assumptions (not everyone is a meat-lover and it won't affect the buying habits of everyone!), but gives a sense of the scale that savings could be made (of millions tonnes).
What are other key benefits?
The implemented proposal would help promote a more general awareness and conscientious in consumers on their buying habits and choices, which could spill over into other retail sectors.
If consumers were to become more conscious of the sustainability of their purchases via exposure in some of the initial ethically-conscious stores (e.g. Co-operative), then this could put significant pressure on the other large supermarket retailers to follow this lead and provide the same carbon information to their customers too- leading to quick widespread uptake in the majority of stores across a given country. Societal pressure has driven such changes in a variety of ways (e.g. many retailers have now committed to sourcing only free-range eggs due to consumer pressure for them to do so)
If consumers were more aware of the environment impact of food and drinks products, and buy accordingly, it may encourage producers to consider ways in which they could reduce the carbon emissions of their production.
What are the proposal’s costs?
The fundamental Life-Cycle Analysis (LCA) data for a full range of food and drink items already exists and would be accessible, at no cost. Derivation of an ideal algorithm and set-up of a user-friendly interface and database could be achieved at no cost, as the basis of an MSc dissertation project.
The key hurdle in implementation of this proposal would lie in getting supermarkets and retailers on-board, with integration into their online shopping experience (and possible development into stores). But with statistics showing that consumers would like to make more educated decisions about the sustainability of their purchases, there does seem to a significant drive and societal/consumer pressure for supermarkets to do more to inform on the carbon footprint of their products (which customers can choose to use or simply ignore).
With much of the carbon footprinting data already available, development of the virtual database and interface could be generated within 1-2 months next summer as part of an MSc dissertation project. Alternatively, if there is significant support for it to be developed sooner, the model and interface could be developed in the same time period before next summer.
It's difficult to put a timeframe on putting this into application on retailer's online stores- it would depend entirely on how willing sellers would be to uptake and implement the proposal. If supermarkets viewed the proposal positively, there is no reason this could not be effectively run within a year.