General Enquiries on the form should be made to:
Defra, Procurements and Commercial Function (Evidence Procurement Team)
E-mail: [email protected]
Evidence Project Final Report

Note
In line with the Freedom of Information
Act 2000, Defra aims to place the results
of its completed research projects in the
public domain wherever possible.
The Evidence Project Final Report is
designed to capture the information on
the results and outputs of Defra-funded
research in a format that is easily
publishable through the Defra website
An Evidence Project Final Report must
be completed for all projects.

This form is in Word format and the
boxes may be expanded, as appropriate.

ACCESS TO INFORMATION
The information collected on this form will
be stored electronically and may be sent
to any part of Defra, or to individual
researchers or organisations outside
Defra for the purposes of reviewing the
project. Defra may also disclose the
information to any outside organisation
acting as an agent authorised by Defra to
process final research reports on its
behalf. Defra intends to publish this form
on its website, unless there are strong
reasons not to, which fully comply with
exemptions under the Environmental
Information Regulations or the Freedom
of Information Act 2000.
Defra may be required to release
information, including personal data and
commercial information, on request under
the Environmental Information
Regulations or the Freedom of
Information Act 2000. However, Defra will
not permit any unwarranted breach of
confidentiality or act in contravention of
its obligations under the Data Protection
Act 1998. Defra or its appointed agents
may use the name, address or other
details on your form to contact you in
connection with occasional customer
research aimed at improving the
processes through which Defra works
with its contractors.
Project identification
1.
Defra Project code
2.
Project title
FO0430
Evidence to define the sustainability of a healthy diet
3.
Contractor
organisation(s)
Institute of Food Research
University of East Anglia
SIK – the Swedish Institute of Food and
Biotechnology
Gene Rowe Evaluations
University of Kent
54. Total Defra project costs
(agreed fixed price)
5. Project:
EVID4 Evidence Project Final Report (Rev. 06/11) Page 1 of 27
£
86,591
start date ................
October 2010
end date .................
November 2011
6. It is Defra’s intention to publish this form.
Please confirm your agreement to do so. ................................................................................... YES
NO
(a) When preparing Evidence Project Final Reports contractors should bear in mind that Defra intends that
they be made public. They should be written in a clear and concise manner and represent a full account
of the research project which someone not closely associated with the project can follow.
Defra recognises that in a small minority of cases there may be information, such as intellectual property
or commercially confidential data, used in or generated by the research project, which should not be
disclosed. In these cases, such information should be detailed in a separate annex (not to be published)
so that the Evidence Project Final Report can be placed in the public domain. Where it is impossible to
complete the Final Report without including references to any sensitive or confidential data, the
information should be included and section (b) completed. NB: only in exceptional circumstances will
Defra expect contractors to give a "No" answer.
In all cases, reasons for withholding information must be fully in line with exemptions under the
Environmental Information Regulations or the Freedom of Information Act 2000.
(b) If you have answered NO, please explain why the Final report should not be released into public domain
Executive Summary
7.
The executive summary must not exceed 2 sides in total of A4 and should be understandable to the
intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together
with any other significant events and options for new work.
Main findings




This project has collated and evaluated numerically and qualitatively the nature, quality and
robustness of a large amount of evidence relevant to assessing the sustainability of a healthy diet, as
described by the eatwell plate.
A much greater volume and depth of robust evidence appears to be available with respect to
environmental sustainability than for economic and social sustainability.
o This reflects the development of life cycle assessment (LCA) and related research activities,
the four most commonly evaluated attributes being global warming potential (GWP),
eutrophication (EP), acidification (AP) and reported energy (RE). Additionally, the
environmental evidence predominantly relates to the production end of the food chain (prefarm gate).
o Evidence for economic attributes (see Methodology below) is also much greater at the
production end of the chain, with limited evidence downstream. The evidence that does exist
is also highly aggregated and of variable robustness in terms of methodological and analytical
rigour.
o The indicators of social sustainability are also not particularly product specific and at a higher
level than the environmental and economic indicators.
The eatwell groups are, by their very nature, heterogeneous in composition and therefore broad in
impact.
o As an example, the ‘Meat, fish, eggs, beans and other non-dairy sources of protein’ sector
includes vegetable protein such as pulses but also red meat (with relatively greater impacts
in global warming potential per kg).
o The variation in reported environmental impacts between food types within each eatwell food
group is further compounded by variation in reported impacts observed within the food
production chains of specific foods or food types (e.g. tomatoes, fish, etc.).
o Given these variations, it may be concluded that the groups in the eatwell plate are too broad
particularly for comparing environmental sustainability and probably for the other Pillars, and
will create significant challenges in communicating appropriate information on sustainability to
the consumer.
The current topical focus on greenhouse gas (GHG) emissions may promote behaviour change which
is environmentally beneficial from the GHG emissions viewpoint, but unintentionally overlook other
very important impacts, both positive and negative. Hence, a “consequential” approach to life cycle
analysis is considered to be the most appropriate route to evaluating the impact of dietary change,
EVID4 Evidence Project Final Report (Rev. 06/11) Page 2 of 27


augmented by using a “checklist” to ensure that change between elements of the diet does not lead to
unintended consequences in other attributes of sustainability e.g. changes in biodiversity or working
conditions.
A consequential lifecycle assessment approach is best used if the goal is to understand the impact (or
consequences) of a change in consumer diet, i.e. not the total (attributional) impact of a static
characterised set of food chains. Where a whole set of foods are to be promoted by policy (such as
eat more pulses and vegetables) over another set (e.g. reduce meat consumption), an assessment of
the impact of the change in the demand for their respective food chains will be very complex. The
impacts may result from consequences that are both direct and indirect. For example, substantial
changes in the demands for these foods could alter direct land and resource use, but also changes in
market price signals may indirectly impact upon commodity demand and land use resources
elsewhere as a consequence. These indirect environmental impacts are not captured by attributional
methods. Characterising and evaluating the marginal difference in impacts from many individual food
chains even if crudely grouped into various ‘eat less’ and ‘eat more’ (‘meats’, ‘vegetables’, etc.)
scenarios, to estimate the cumulative impact would be very time demanding and not without
significant methodological challenges.
Finally, it is imperative to ensure that food losses and waste disposal impacts along the food chain are
accounted for in all life cycle studies of food products particularly in the light of the recent FAO
(Gustavsson et al, 2011) and WRAP (Waste & Resources Action Programme, 2009; 2010 a&b)
reports, which highlight the high proportion of food wasted.
Introduction
Our food system and diet strongly influence our health and the environment and there is great pressure to
ensure that our food production and consumption is both health-promoting and sustainable. It is also
necessary to provide consumers clear and consistent information on the impact of our food choices.
Although there is no clear cut definition of “sustainability”, it is generally accepted that there are three
“pillars” of importance: environmental, economic and social. A sustainable diet has many attributes across
these pillars and there may be synergies and trade-offs between them.
Policies on sustainable consumption need to be underpinned by research to provide the necessary
evidence base. The aim of this study has been to identify gaps in the knowledge base by defining and
evaluating food chains that contribute to healthy diets (as described by the eatwell plate) and to evaluate
the nature, extent and robustness of evidence for their attributes of sustainability.
Methodology
The Project Consortium and an external panel of experts have reviewed information sources to identify
evidence, data, guidance and metrics readily available to define the environmental, social and economic
sustainability attributes of a healthy diet (as represented by the eatwell plate). Sources included academic
theses, reviews and reports, book chapters and company information/websites as well as personal
communications. [It is important to note that the project was not evaluating established guidance on
healthy eating].
On the basis of published literature and expert opinion from members of the steering group and members
of the project team’s research experience, the following sustainability attributes were chosen for
consideration:
Sustainability attributes examined
Environmental
Economic
Social
 Global warming potential
 Land use
 Health and welfare
(perceived and real), with
 Eutrophication potential
respect to a) food safety and

Production
 Acidification potential
b) nutrition
 Ozone depletion potential

Ethics, with respect to a)

Prices
 Photochemical ozone creation
inequity (amongst social
potential
groups) and b) animal

Value
of
output
 Biotic resource use / depletion
health/welfare
 Abiotic resource use /
 Working conditions, with
 Employment
depletion
respect to employment
 Ecotoxicity potential
conditions and worker safety

Trade
 Human toxicity potential

Societal dynamics, with
 Land use
respect to community
 Reported energy
cohesion, education,
 Water resource
landscape
 Biodiversity
EVID4 Evidence Project Final Report (Rev. 06/11) Page 3 of 27
Results and discussion
Evidence relevant to assessing the environmental sustainability of a healthy diet
The study focused predominantly on evaluating evidence from life cycle assessment studies, a large and
growing body of evidence. The results demonstrate that
 for all eatwell food groups, the system boundaries are generally positioned towards the farm
gate, and very few studies extend down the supply chain as far as the consumer.
 There is a shortage of food specific LCA studies which extend the boundary to include
information on distribution, retail and consumption and waste as part of a full life cycle of a food
product. Data on food losses at retail and consumer habits specific to food types would be useful
in understanding the impacts of 1kg of food type consumed, rather than just 1kg food type
produced.
 The four most commonly evaluated attributes were GWP, EP, AP and RE. The widest
consideration of attributes was evident in the “Fruit and Vegetables” food group.
The results are consistent with a commonly accepted view that generally larger values are reported by
studies on red meat production for GWP than those reported for vegetable production. However, since
each food group in the Eatwell Plate contains a considerable diversity of food products, a
considerable range in the reported values of environmental burdens are evident in the eatwell food
groups. This is most apparent in the non dairy protein food group. To add another layer of complexity,
wide ranges in reported values were found in studies on specific foods, this is apparent for meats but also
some examples are evident in fruit and vegetables. The lack of transparency of the studies – details of
methodologies and data sources etc. - made exact reasons for variations in reported values difficult to
interpret but some of the key reasons are summarised below.
Variation in reported burdens for specific food commodities
The relatively large variation in reported burdens for specific meats appear to be due to methodological
factors. For example, variations within the impact values reported for studies on beef production were due
to differences in how burdens were allocated to expensive cuts of meat, whilst some studies use
methodologies for soil emissions that are not used by other studies. The range would make conclusions
for quantifying environmental burdens of meat in healthy diets problematic. In fruits and vegetables
however, though some variation is still likely to be caused by differences in the goals and scopes of the
studies, methodologies, the setting of boundaries, assumptions, and data quality, relatively greater GWP
reported appear to be from differences in production systems (e.g. comparing seasonal production with
the growing fruits out of season under glass or air freighting fresh produce).
Variation in reported burdens within food groups
In addition to methodological differences the relatively large differences in reported burdens between
some types of meat & meat products (e.g. for chicken and beef) would appear to be explained by intrinsic
differences in animal species production characteristics, (i.e. extensive vs. intensive production, ruminants
vs. non ruminants, different feed conversion ratios and different feed components and their production
impacts).
It is important to reiterate that the inherent variation within the eatwell plate food groups regardless of
methodological issues. The eatwell groups are too broad in their compositions to quantify and convey the
relative levels of environmental sustainability. Aligning dietary intake to the proportions of the broad food
groups recommended by the eatwell plate allows for a multitude of different food options and therefore a
potentially wide variation in the environmental burden of a healthy diet. This issue along with the lack of
studies following a consistent or harmonised LCA methodology in reporting environmental burdens of
specific food chains within the eatwell plate (i.e. most foods) will create significant challenges in
communicating environmental information to the consumer.
Evidence relevant to assessing the Social sustainability of a healthy diet
Apart from a few exceptions, investigations indicated that there was a paucity of research on social
sustainability in relation to specific food chains. Indeed, social sustainability indicators are linked mostly to
how a food product is produced, processed, distributed and consumed, rather than the specific type of
food. This is a very different approach to that used in LCAs and other methods for assessing
environmental sustainability, and hence there may be a need to approach and conceptualise social
sustainability in a different way. Therefore, information searches on each dimension were targeted
generically towards each of the components in the food chain, from which findings could subsequently be
considered in relation to the different eatwell plate food groups.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 4 of 27
Unsurprisingly, the most highly informed dimension was that of health, reflecting the enormous effort and
interest in supplying the population with safe, health–promoting nutrition. The ethical dimension was also
highly informed, particularly regarding the impact of biotechnology e.g. GM for both plant- and animalbased foods, animal (including fish) welfare through to the knowledge base concerning social inequality in
relation to consumption. In contrast, the other two dimensions (working conditions and societal dynamics)
appear to be much less informed.
Evidence relevant to assessing the Economic sustainability of a healthy diet
The literature is of a generic nature, not specific to food group or to stage of the supply chain. Very few
studies that explore the impact of shifting towards a healthy diet were identified which corroborated the
indicators chosen for this study. It is evident that data availability is concentrated at the production end of
the chain and, for value, the retail/consumption end. For storage, transport and distribution, data is
fragmented at best. Of particular significance is the observation that food supply chains are highly
differentiated in their structure and operations – between food groups and distribution channels – and thus
much more complex, in terms of impact analysis, than the generic production – to – consumption chains
that are typically assumed in LCA studies. This makes the comparison and integration of economic
analysis with environmental analysis particularly difficult as the former is highly context-specific and
descriptive whilst the latter tends to be more generalised and prescriptive.
The healthy diet and interrelationships between attributes from across the three pillars
Recent high level reports and publications that consider food consumption and sustainability issues tend
to focus on the opportunities for reducing GHG in the food chain (Garnett 2011), and the impact of
changing diet on GHG emissions (Audsley et al 2009; MacDiarmid et al, 2011). Whilst this undoubtedly
reflects the topical interest in climate change, it also reflects the limitations in evidence across the other
environmental attributes as well as the lack of readily available evidence from the economic and social
directions. This may promote behaviour change which is environmentally beneficial from the GHG
emissions viewpoint, but unintentionally overlook other very important impacts, both positive and
negative (as acknowledged by Garnett 2011).
The eatwell plate gives a visual representation of the types and proportions of foods needed for a healthy
and well balanced diet. This suggests the consumption of more vegetables, fruits and starchy foods, and
less meat / dairy foods compared with current UK dietary behaviour. A significant reduction in the demand
for meat and dairy products by UK consumers might be expected to result in a decline in UK meat
production, assuming that the current output could not be supported by demand from overseas. This, in
turn, would be expected to reduce the environmental impacts of cattle production (namely methane
emissions) and of importing feed. However, increased consumption of vegetables and fruits may result in
increased imports to provide supplies out of season. It may also lead to changes in land use in order to
increase the production of food crops such as fruits and vegetables, although it might also lead to a
reduction in area of crops used in animal feed or perhaps energy crops. From the social perspective, the
changes in landscape may be deleterious if cattle husbandry was to be replaced by monotonous crop
monocultures. From an economic perspective, such changes would have substantially negative impacts
on the meat and dairy food chains, reducing commercial and employment opportunities in these sectors
particularly in agronomy, processing, transport and distribution and retail. There would also be an
associated impact on supply and employment related to non-food industries using by-products from
livestock. Increased unemployment would be expected to have negative social connotations. From the
environmental perspective, changes in land-use would be expected to have significant impacts on
biodiversity. Hence consuming less meat and dairy products may have positive benefits to health and to
some environmental attributes, but may result in numerous changes to impacts on other environmental,
economic and social sustainability attributes which need to be considered.
The current study demonstrates that for such an evaluation there is probably sufficient information to
evaluate the economic impacts, for example, on employment and production and prices. From the social
perspective, there is probably sufficient information to evaluate aspects of health but little on societal
impacts or working conditions. From the environmental perspective, it is clear that from LCA reports, only
three of the attributes (GWP, AP & EP) are adequately covered. However, there may be sufficient nonLCA information concerning biodiversity, water footprint, and reports on UK food waste to be able to make
a reasonable inference as to the effects of dietary change.
In conclusion
This project has collated and evaluated a large amount of evidence on attributes relevant to the
sustainability of food & drink consumption (Annexes A, B & C).
1) A much greater volume and depth of robust evidence appears to be available with respect to
environmental sustainability than for economic and social sustainability. This reflects the development
of LCA and related research activities, the four most commonly evaluated attributes being GWP, EP,
AP and RE. These generally relate to the production end of the food chain.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 5 of 27
Evidence for economic attributes is also much greater at the production end of the chain, with limited
evidence downstream. The evidence that does exist is also highly aggregated and of variable
robustness in terms of methodological and analytical rigour.
2) The indicators of social sustainability are not particularly product specific and at a different level than
the environmental and economic indicators (the focus being on processes rather than food types). The
definition of social sustainability is also more uncertain than for the other sustainability types, and it is
important to derive a more coherent and acceptable definition of this.
3) There is a pressing need to evaluate the above information in relation to UK consumption data with
particular reference to the most commonly consumed foods. Furthermore, additional studies are
required which adopt a more targeted and consequential approach as follows:
a. It is necessary to objectively prioritise a list of key sustainability attributes across the three pillars.
This is because the perceived importance of these attributes may differ according to, for example,
their topicality (e.g. climate change is much more heavily featured in the press than other
environmental attributes such as water use or resource depletion) or the country of production
where working conditions, employment or local biodiversity (e.g. farmland birds) may be an issue.
b. Secondly, using the agreed key sustainability attributes and selected, commonly-consumed
products, to examine in detail a change in consumption from one commonly-consumed product to
another. Indeed, the current Defra project FO0427 ‘Future implications of trends in healthy eating
on existing food production and manufacture’ will model possible future UK diet scenarios and
determine their environmental, social and economic impacts. Such a study should assess the
potential impact on the agreed sustainability attributes across all three pillars and then evaluate
how and where changes measured by the environmental attributes might impact on the social and
economic attributes (and vice versa).
The approach should also include assessing the possibility of “spillover” where sustainability
associated with production and consumption activity is displaced into other sectors. For example,
as pointed out by Garnett (2011), even if consumers change to eating less meat, any financial
savings made may just be used to carry out other GHG-emitting activities such as recreation and
travel. Furthermore, since the food supply is global, impacts of changing UK dietary patterns may
influence food production in other countries.
It may also be appropriate to extend studies such as the Livewell study (Macdiarmid et al, 2011)
to evaluate and as far as possible estimate the impacts of dietary change on a greater range of
sustainability attributes than just greenhouse gas emissions. This approach may be augmented by
using a “checklist” to ensure that change between elements of the diet does not lead to
unintended consequences in other attributes of sustainability e.g. changes in biodiversity or
working conditions.
c.
Finally, it is imperative to ensure that food losses along the food chain are accounted for in all life
cycle studies of food products because they are often missed out. The importance of this is given
weight by the recent FAO (Gustavsson et al, 2011) and WRAP (Waste & Resources Action
Programme, 2009; 2010 a&b) reports which highlight the high proportion of food wasted.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 6 of 27
Project Report to Defra
8.
As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of
the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra
to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information
obligations. This short report to Defra does not preclude contractors from also seeking to publish a full,
formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively
encourages such publications as part of the contract terms. The report to Defra should include:
 the objectives as set out in the contract;
 the extent to which the objectives set out in the contract have been met;
 details of methods used and the results obtained, including statistical analysis (if appropriate);
 a discussion of the results and their reliability;
 the main implications of the findings;
 possible future work; and
 any action resulting from the research (e.g. IP, Knowledge Exchange).
1.0 Introduction
Our food system and diet have a huge impact on our health and the environment and there is great
pressure to ensure that our food production and consumption is both health-promoting and sustainable.
In order to give consumers clear, consistent information on the impact of our food choices, we will need to
develop a robust definition of a diet that is both healthy and sustainable. Indeed there has been an
explosion of policies, instruments and initiatives during the past decades (from both governmental and
market actors) to achieve sustainable development outcomes across a broad range of economic sectors,
including food supply chains.
The most widely quoted definition of sustainable development is probably that of the Bruntland
Commission, namely “Sustainable development is development that meets the needs of the present
without compromising the ability of future generations to meet their own needs”. Although there remains
no clear cut definition - it is generally accepted that there are three areas of importance: environmental
sustainability, economic sustainability and social sustainability. A sustainable diet has many attributes
(attributing factors) and there will be synergies and trade-offs between these. The subsequent
development of detailed policies must also be underpinned by research to provide the necessary evidence
base.
This report provides an overview and brief evaluation of the nature and extent of evidence that may be
used in defining the sustainability of a healthy diet. The information within this report was compiled as part
of a Defra-funded study between August 2010 and July 2011.
The objectives of the study, as agreed with Defra and the Project Steering Group, were:
1) To define and evaluate food chains underpinning healthy diets which are described by current
Government guidelines with special reference to the eatwell plate.
2) To develop, where possible, a matrix approach and classify attributes of sustainability (social,
economic, and environmental) in relation to key parts of the representative food chains identified
in Step 1.
3) To integrate and evaluate the robustness of evidence from a broad range of disciplines with expert
opinion from an invited group, identify gaps in our knowledge, and provide recommendations to
inform advice to consumers and identify further research required to characterise a sustainable
healthy diet.
Data, evidence, metrics and guidance concerning the sustainability attributes of a healthy diet are
presented in relation to the component food chains.
2.0 Methodology
The Project Consortium and an external panel of experts have reviewed information sources to identify
evidence, data, guidance and metrics readily available to define the environmental, social and economic
sustainability attributes of a healthy diet. Information sources included academic theses, reviews and
reports (both peer and non-peer reviewed, from the public and private sector), book chapters and
company information / websites as well as personal communications.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 7 of 27
2.1 Food chains underpinning healthy diets: the eatwell plate
In accordance with the Project Specification, the study has used the Department of Health’s eatwell plate
and food groups therein as a basis for the composition of a healthy diet and has not attempted to reexamine this.
Figure 1. The eatwell plate. Image: Crown Copyright. Department of Health in association with the Welsh
Assembly Government, the Scottish Government and the Food Standards Agency in Northern Ireland.
The eatwell plate is a policy tool that defines the Government’s recommendations on healthy diets. It
makes healthy eating easier to understand by giving a visual representation of the types and proportions
of foods needed for a healthy and well balanced diet.
The eatwell plate is based on the five food groups:

Bread, rice, potatoes, pasta and other starchy foods

Fruit and vegetables

Milk and dairy foods

Meat, fish, eggs, beans and other non-dairy sources of protein

Foods and drinks high in fat and/or sugar
The eatwell plate encourages the choice of different foods from the first four groups every day, to help
ensure the population obtains a wide range of nutrients needed to remain healthy.
Choosing a variety of foods from within each group will add to the range of nutrients consumed. Foods in
the fifth group – foods and drinks high in fat and/or sugar are not essential to a healthy diet.
In each case, and for each pillar of sustainability, the food chains were generically considered as
comprising:

Agronomy / production

Processing

Storage and Transport (throughout the chain)

Distribution and Retail

Consumption
EVID4 Evidence Project Final Report (Rev. 06/11) Page 8 of 27
The methodologies associated with evaluating the evidence base were tailored to suit the different
sustainability attributes associated with the three pillars.
2.2 Classification of attributes of sustainability (social, economic, and environmental) in relation to key
parts of the representative food chains and the evaluation of evidence.
2.2.1 Environmental Sustainability
The evaluation of environmental sustainability is a broad and rapidly growing field of scientific research.
Over the last 25 years, a wide range of methods and attributes have been developed and explored in
relation to environmental impact of human and other activities. On the basis of published literature and
expert opinion, the following sustainability attributes were chosen for consideration:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
global warming potential (GWP)
eutrophication potential (EP)
acidification potential (AP)
ozone depletion potential (ODP)
photochemical ozone creation potential (POCP)
biotic resource use / depletion (BRD)
abiotic resource use / depletion (ARD)
ecotoxicity potential (ETP)
human toxicity potential (HTP);
land use (LU)
reported energy (RE)
water use (WU)
biodiversity (BD)
Whilst no one method for evaluating environmental impact is all encompassing, there is no doubt that the
most common approach involves the use of life cycle assessment (LCA), underpinned by a growing
number of software tools and databases e.g. SimaPro (Pré Consultants), Ecoinvent (ecoinvent Centre) &
GaBi (PE International). Therefore, for the purposes of this study, published LCAs (mostly attributional)
were chosen as the most robust sources of evidence for quantifying environmental attributes of specific
foods. LCA results of all food items found in extensive literature searches were recorded, where
applicable, for comparison under categories of the eatwell plate food groups along with other data on:






The food product;
The kind of production system;
The functional unit reported;
The data source and country of study;
The extent of the food chain covered (system boundaries);
The reported results and metrics of environmental impact.
Where possible the impacts reported were converted to common units of impact characterisation and
scaled to functional units of 1kg of food product. A summary of this evidence is presented in Annex A.
In addition to LCA, there are additional sources of data that can provide evidence for evaluating the
environmental sustainability of a healthy diet. These have also been considered and the findings
summarised.
2.2.2 Social Sustainability
Unfortunately, ‘social sustainability’ has no fixed, universally accepted definition. Further, the evaluation of
social sustainability is more qualitative in nature when compared with the quantitative methodology
associated with environmental evaluation and economic evaluation. When social sustainability has been
defined, it has tended to be according to the immediate and diverse needs of different practitioners. As
such, numerous indicators / attributes have been alluded to, for example, “internal social sustainability”
(relating to working conditions) and “External social sustainability” (relating to food safety, animal welfare
and animal health), by Van Calker et al (2005). In contrast Olesen, Groen and Gjerde (2000) in their
definition of sustainability, included social viability, and ethical aspects. Some studies have focused on
individual issues, for example, Broom (2010) has stated that ‘a system that results in poor (animal) welfare
is unsustainable because it is unacceptable to many people’. More recently the FAO (2011) highlighted
‘social well-being’ as including labour rights, non-discrimination and equity, education, health and safety,
nutritional products and safe working conditions while including ‘good governance’ as a fourth pillar of
social sustainability.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 9 of 27
Therefore, after consultation with the Expert Group, the following four key “dimensions” (broad attributes)
were selected:




Health and welfare (perceived and real), with respect to a) food safety and b) nutrition
Ethics, with respect to a) inequity (amongst social groups) and b) animal health/welfare
Working conditions, with respect to employment conditions and worker safety
Societal dynamics, with respect to community cohesion, education, landscape
Initial investigations indicated that there was, apart from a few exceptions, relatively little research on
social sustainability elements in relation to specific food products. Indeed, social sustainability indicators
are linked mostly to how a food product is produced, processed, distributed and consumed, rather than the
specific type of food. This is a very different approach to that used in LCAs and other methods for
assessing environmental sustainability, and hence there may be a need to approach and conceptualise
social sustainability in a different way. For present purposes, information searches on each dimension
were targeted generically towards each of the components in the food chain, from which findings could
subsequently be considered in relation to the different eatwell plate food groups. Evidence was gathered
through international web-based databases (Web of Science; mostly peer-reviewed articles) and expert
recommendations from the Steering Group and Expert Panel.
Information was compiled in tabular form (Annex B) and a subjective judgement on the level of knowledge
for each food component was scored numerically.
2.2.3 Economic sustainability
As a starting point, the basic factors of production (land, labour and capital) were selected and then other
possible dimensions added, these being trade, pricing and managerial capacity.
After preliminary literature studies, the following list was created:
1)
2)
3)
4)
5)
6)
Land (area)
Production
Prices
Value of output
Employment
Trade
Information on the availability of information and evidence for these indicators was obtained through the
evaluation of published literature, and guidance from contact with organisations and individuals with
expertise. This was performed for selected products and their food chains within each of the five food food
groups of the eatwell plate. Whilst it is recognised that there may be some gaps, some of the data
sources identified will be common to other products within each food group.
Table 1: Food products chosen to represent different food groups of the eatwell plate
eatwell plate food group
Product(s) chosen to represent
food group
Bread, rice, potatoes, pasta and other starchy foods
Potatoes, cereals, flour, bread
Fruit and vegetables
Carrots, lettuce, strawberries,
dessert apples
Milk and dairy foods
Milk, cheese (Cheddar)
Meat, fish, eggs, beans and other non-dairy sources of
protein
Red meat
Foods and drinks high in fat and/or sugar
Ice cream, biscuits
For each of the five food groups, an attempt was made to identify data sources for each of the six
economic indicators. This information was compiled in tabular form (Annex C) and a subjective judgement
on the level of readily available knowledge for each food component was scored numerically.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 10 of 27
3.0 Results
3.1 Evidence relevant to assessing the environmental sustainability of a healthy diet
A comprehensive synopsis of all the LCA and related literature evaluated was compiled (over 180
published articles). A simple numerical quantification of incidences (i.e. a count of all occurrences of all
attributes in all references) was graphically presented across the food chain for each of the five eatwell
plate food groups (Figure 2). An additional food chain step, “post-consumer disposal”, was included in this
analysis.
Figure 2. Comparative frequencies of incidences of environmental evaluation for different stages of the
food chain
Incidence frequency
300
250
Agronomy impacts
200
Processing impacts
Distribution / transport
Retail
150
Consumption
Post-consumer disposal
100
50
0
Bread & starchy foods
Foods high in fat and/or
sugar
Fruit and vegetables
Milk and dairy
Meat, fish etc
The results demonstrate that for all food groups, the system boundary is generally positioned towards
the farm gate end of the food supply chain, and very few incidences extend as far as the consumer.
There is a shortage of studies which include information on distribution, retail and consumption and
waste.
The summarised information on all the literature was further evaluated in order to assess the availability
and quality of the information in more depth with reference to the individual sustainability attributes. A list
of food products for which evidence on environmental sustainability was found is shown in Table 2. The
overall assessment of the evidence relevant to all selected sustainability attributes versus the eatwell plate
food groups is shown in Figure 3, whilst a more detailed assessment of each food group examining the
attributes against food product is shown in Annex A. The four most commonly evaluated attributes were
GWP, EP, AP and RE. The widest consideration of attributes was evident in the “Fruit and Vegetables”
food group.
Table 2.
List of foods organised by eatwell plate food group for which evidence of environmental
sustainability was found.
Food group
Bread, rice,
potatoes, pasta
and other starchy
foods
Fruit and
vegetables
Products for which evidence was found
 Bread

 Potatoes

 Oat flakes
Pasta
Rice





Apples
Pineapple
Strawberries
Carrots & root vegetables
Brassicas




Lettuce & cucumber
Tomatoes
Citrus fruits & juices
Alliums
Milk and dairy
foods


Milk
Cheese

Yoghurt
EVID4 Evidence Project Final Report (Rev. 06/11) Page 11 of 27
Meat, fish, eggs,
beans and other
non-dairy sources
of protein






Beef
Pork
Lamb
Chicken
White fish
Oily fish




Shellfish
Eggs
Legumes & pulses
Mycoprotein & processed meat
analogues
Foods and drinks
high in fat and/or
sugar





Cakes, biscuits & pastries
Sugary drinks
Chocolate & sweets
Ice cream
Jam & honey





Crisps
Butter, margarine & spreads
Cream
Edible oils
Sugar
Figure 3: Overall assessment of quality and robustness of environmental attribute data (mostly from LCA)
for all eatwell plate food groups.
Key:
GREEN
ORANGE
or
RED
Good evidence available for the majority of food products in food group. No
gaps in coverage of food group in terms of the most commonly consumed
products.
Some evidence / limited evidence across food group although coverage of
some products may be very good.
Gaps in coverage of food group in terms of consumption of products.
Insufficient evidence / no evidence on majority of food products in food
group.
Attribute
Bread, rice,
potatoes,
pasta & other
starchy foods
Fruit &
vegetables
eatwell plate food group
Milk & dairy
Meat, fish,
foods
eggs, beans
and other
sources of
non-dairy
protein
Foods and
drinks high in
fat and / or
sugar
GWP
EP
AP
POCP
ODP
HTP
ETP
ARD
BRD
RE
LU
BD
WU
Abbreviations: GWP, global warming potential; EP, eutrophication potential; AP, acidification potential; POCP, photochemical ozone
creation potential; ODP, ozone depletion potential; HTP, human toxicity potential; ETP, ecotoxicity potential; ARD, abiotic resource
depletion; BRD, biotic resource depletion; RE, reported energy; LU, land use; BD, biodiversity; WU, water use.
Even though there would appear to be a good level of evidence for these four attributes (shown by the
green highlighted boxes), considerable variation was often found within the published data, and between
data from different eatwell plate food groups. In order to present this variation as clearly as possible, the
data from all the identified literature for these four attributes has been graphically presented (Figure 4).
EVID4 Evidence Project Final Report (Rev. 06/11) Page 12 of 27
Figure 4: Plots of reported values for the four most commonly reported attributes for main categories of
foods within each food group of the eatwell plate.
Eutrophication Potential
g PO43- eq. kg-1
700
600
500
400
300
200
100
2
Pork
3
4
5
6
7
8
Sheep Poultry Farmed Farmed Wild Wild
meat
Oily Fish White White Oily
fish
Fish
Fish
9
10
Farmed Wild
Shell Shell
Fish
Fish
11
12
13
Non- Eggs
Meat
Protein
Eutrophication Potential g PO43- (ATTRIBUTIONAL)
14
15
Cheese Yoghurt Milk
16
Fruit
17
18
19
Bread Potatoes
Eutrophication Potential g PO43- (CONSEQUENTIAL)
EVID4 Evidence Project Final Report (Rev. 06/11) Page 13 of 27
20
21
Pasta
22
Rice
High in fat and sugar
1
Beef
Whole grains
(oats and wheat)
0
Vegetables
0
23
Acidification Potential
g SO2 eq. kg-1
800
700
600
500
400
300
200
100
0
3
4
5
6
7
8
Sheep Poultry Farmed Farmed Wild Wild
meat
Oily Fish White White Oily
fish
Fish
Fish
9
10
Farmed Wild
Shell Shell
Fish
Fish
11
12
Non- Eggs
Meat
Protein
Acidification Potential gSO2eq (ATTRIBUTIONAL)
13
14
15
Cheese Yoghurt Milk
16
Fruit
17
18
19
Bread Potatoes
20
21
22
Pasta
Rice
23
High in fat and sugar
2
Pork
Whole grains
(oats and wheat)
1
Beef
Vegetables
0
Acidification Potential gSO2eq (CONSEQUENTIAL)
Reported Energy
MJ kg-1
600
500
400
300
200
100
0
3
4
5
6
7
8
Sheep Poultry Farmed Farmed Wild Wild
meat
Oily Fish White White Oily
fish
Fish
Fish
9
10
Farmed Wild
Shell Shell
Fish
Fish
11
12
Non- Eggs
Meat
Protein
Energy Use MJ kg-1 (ATTRIBUTIONAL)
13
14
15
Cheese Yoghurt Milk
16
Fruit
17
18
Energy Use MJ kg-1 (CONSEQUENTIAL)
EVID4 Evidence Project Final Report (Rev. 06/11) Page 14 of 27
19
Bread Potatoes
20
21
Pasta
22
Rice
23
High in fat and sugar
2
Pork
Whole grains
(oats and wheat)
1
Beef
Vegetables
0
The results are consistent with generally accepted conclusions. For example, the production of red meat
often exhibits a much higher GWP than production of vegetables. However, an important observation was
that each food category had a considerable range of data, highlighting problems in making general
comparisons between specific food types (beef, pork, etc.) or food groups. Such wide ranges were due to
a number of factors. For instance, the main variation in reported burdens within the meat food chains were
often due to differences in life cycle assessment methodology and boundaries, (for example economic
allocation of GWP emissions to expensive cuts of meat at the retail stage, whilst other studies include
methodologies for soil emissions not commonly included in other studies). In addition, there are wide
differences in the values reported for specific burdens between different types of meat & meat products.
It should be stressed that there is also inherent variation in the environmental impact data within the
eatwell plate food groups regardless of methodological issues. In fruits and vegetables, very high GWP
reported for some products resulted from production system characteristics, (e.g. growing fruits out of
season under glass, longer cold storage compared to importing in-season produce, or air freighting fresh
produce).
The eatwell groups are, by their very nature, wide – for instance, the ‘Meat, fish, eggs, beans and other
non-dairy sources of protein’ food group includes a wide range of interchangeable products which vary in
their impact from very low to very high. . So, a key conclusion is that the groups in the eatwell plate are too
wide for this purpose and will provide significant challenges in facilitating communication of environmental
information to the general public.
In addition, variation was also due to the differences in the goals and scopes of the different studies, the
methodological approaches, the setting of boundaries and assumptions as well as data quality &
availability. It is likely that they were not designed with such comparative evaluation with other studies of
similar or different food products in mind. Other review papers have compared different lifecycle
assessments, often highlighting methodological differences. Transparency of studies, data and methods
were also an issue in this respect – many studies did not give enough details to allow studies to be
repeated. This is partly because of the widespread use of commercial databases but also because
journals do not always provide enough space for all details. In addition, some data are provided by
commercial operators on a confidential basis.
Further differences arise between studies according to whether they apply attributional or consequential
analysis. Attributional (or “accounting”) LCA describes an existing supply chain; it is used, for example, in
estimating the “carbon footprint” of a product. Consequential (or “prospective”) LCA attempts to explore
the system effects of changes in economic activities; a conspicuous use of consequential LCA is in
exploring the effects of changes in land use due to switching from production of food to fuel crops in one
location, compensated by increased food production elsewhere. The methodological differences between
the two forms of LCA are a matter of current debate but, for food products, consequential LCA usually
gives significantly higher impacts, particularly for climate change, because it includes changes in carbon
stock resulting from indirect land use change.
In order to provide a basis for comparison of evidence and data between the three pillars of sustainability,
the availability of environmental data (i.e. a count of all occurrences of the four most commonly reported
attributes) in relation to the eatwell plate food groups and component food chains is provided in Figure 5.
The colour scheme adopted is not judgemental, but provides a graduated indication of the extent of the
information. The “robustness” of the information is presented in Annex A. The results are consistent with
the conclusions above, and additionally demonstrate that most LCA-related studies concern GWP
reported to the farm gate.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 15 of 27
Figure 5. Comparative frequencies of incidences of the four most commonly reported attributes of
environmental sustainability as a function of stage of the food chain for the food products investigated
within different eatwell plate food groups, and for the total eatwell plate.
Bread, rice, potatoes, pasta
& other starchy foods
Fruit and vegetables
Milk and dairy foods
Meat, fish, eggs, beans and
other non-dairy sources
of protein
Foods and drinks high
in fat and/or sugar
Total eatwell plate
GWP
AP
EP
RE
Agronomy
97
46
47
37
Processing
80
36
32
31
Distribution & transport
72
30
26
29
Retail
45
28
24
22
Consumer
49
22
18
24
Post consumer disposal
14
3
2
0
Agronomy
181
84
80
103
Processing
108
43
37
52
Distribution & transport
108
46
40
57
Retail
22
3
3
8
Consumer
17
0
0
3
Post consumer disposal
9
0
0
0
Agronomy
52
31
28
21
Processing
29
14
15
6
Distribution & transport
18
7
8
5
Retail
16
7
8
5
Consumer
12
4
5
4
Post consumer disposal
8
1
2
2
Agronomy
229
132
135
115
Processing
113
52
52
42
Distribution & transport
64
38
40
28
Retail
39
25
27
16
Consumer
21
6
8
15
Post consumer disposal
8
5
7
3
Agronomy
83
40
48
28
Processing
67
24
32
19
Distribution & transport
53
16
20
14
Retail
30
5
9
7
Consumer
20
1
1
3
Post consumer disposal
27
8
8
8
Agronomy
642
333
338
304
Processing
397
169
168
150
Distribution & transport
315
137
134
133
Retail
152
68
71
58
Consumer
119
33
32
49
Post consumer disposal
66
17
19
13
Additional sources of evidence
Whilst LCA data provides the most comprehensive evidence base for most of the sustainability attributes
investigated, notwithstanding the limitations identified, there are also a number of other key sources that
may provide supplementary data.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 16 of 27
The following topics are discussed in Annex A:
1) Biodiversity
2) Waste
3) Alternative methods of environmental assessment.
a. Ecological footprinting
b. Water footprinting.
Food production systems and marques
Many management or product standards exist for stages of the food chain, though chiefly they are more
applicable to the food production stage.
A substantial number of UK farm assurance schemes have been carefully considered by Lewis et al
(2010) as part of a Defra-funded study. The study concludes that the proportion of content that sets out to
seriously tackle environmental issues is very limited in all but a few of the schemes considered, although it
should be recognised that this was not necessarily the focus of the schemes. There is often a mismatch
between the most commonly-reported attributes in LCA studies (e.g. GWP above) and those addressed
by farm assurance schemes (e.g. food safety, quality & traceability / animal welfare / high production
standards). Few sources of independent research were found specifically evaluating the environmental
attributes of management standards and marques from a lifecycle perspective. Where such studies were
found, these were not UK management standards (Cederberg & Stadig, 2003; Basset-Mens & van der
Werf, 2005; Basset-Mens et al, 2007; KRAV, 2010).
3.2 Evidence relevant to assessing the Social sustainability of a healthy diet
The evidence that has been obtained reflects that which can be readily identified. As some social science
provide qualitative data there may be significant additional relevant data that will not have been
recognised within the confines of this study. Furthermore, there are several caveats that should be borne
in mind: First, within the Social Sustainability Pillar, the potential sustainability of any one attribute may
potentially conflict with that of another, making it difficult to state with any clarity the overall social
sustainability of a healthy diet, or component thereof. Second, even one attribute may be affected in
conflicting ways by a specific action or behaviour. For example, increasing transport might be seen to
have a negative impact on human health as a result of greater traffic pollution; on the other hand, such
increased transport might also increase the availability of healthy foods thereby having a positive impact
on health. This issue reflects the limitations of a relatively high-level evaluation of Social Sustainability in
this study. Third, papers relevant to social sustainability can be found in sources ranging from medical
and nutrition journals to those on psychology, sociology, geography, law and occupational matters (to
name but a few). This project has not been able to evaluate fully the plethora of research from all of these
social science disciplines.
The key findings have been summarised in tabular form, and numerically scored on the basis of simple
criteria to reflect the availability and robustness of the information. A major caveat is needed here,
however, which is that considerable interpretation has taken place in order to discriminate the relevance of
the identified data to the different eatwell plate food groups. It may well be that this interpretation is
suspect, given that the data on social sustainability does not generally speak to specific food products
themselves, but rather to how the food is produced, with academic interest being in broader questions
about food processes and their effects on people and society. This data is presented in detail in Annex B
along with a detailed critique. It is summarised here in Figure 6, colour coded according to the key. In
evaluating evidence relevant to social sustainability, the extra component of the food chain, “breeding and
genetic modification”, was included (as this was a stage identified in the initial project proposal).
EVID4 Evidence Project Final Report (Rev. 06/11) Page 17 of 27
Figure 6. Schematic representation of the availability of evidence concerning the social sustainability of
foods within the different eatwell plate food groups.
Key:
GREEN
ORANGE
RED
TOTAL
Good evidence available for the sector in the form of 3 or more reliable
publications. Evidence is relevant across the sector.
Some evidence / limited evidence across sector (e.g. 1-2 reliable publications).
Insufficient evidence / no readily identifiable evidence in sector.
Health
Ethical
Working
conds.
Societal
Bread, rice,
potatoes,
pasta and
other starchy
foods
Breeding
Breeding
Agronomy
Agronomy
Processing
Storage &
Transport
Distribution &
Retail
Processing
Storage &
Transport
Distribution &
Retail
Consumption
Consumption
Fruit and
vegetables
Health
Ethical
Working
conds.
Societal
Milk and
dairy foods
Breeding
Breeding
Agronomy
Agronomy
Processing
Storage &
Transport
Distribution &
Retail
Processing
Storage &
Transport
Distribution &
Retail
Consumption
Consumption
Meat, fish,
eggs, beans
and other
non-dairy
sources of
protein
Health
Ethical
Working
conds.
Societal
Foods and
drinks high in
fat and/or
sugar
Breeding
Breeding
Agronomy
Agronomy
Processing
Storage &
Transport
Distribution &
Retail
Processing
Storage &
Transport
Distribution &
Retail
Consumption
Consumption
Health
Health
Health
Ethical
Working
conds.
Societal
Ethical
Working
conds.
Societal
Ethical
Working
conds.
Societal
3.2.1. Health and Welfare dimension: Unsurprisingly, the most highly informed dimension was that of
health, reflecting the enormous effort and interest in supplying the population with safe, health–promoting
nutrition. At the breeding and production end of the food chain, genetic modification (GM) provides much
discourse with a large body of evidence to show that GM processes do not pose a risk to human health
even though a significant proportion of UK consumers believe they do. The reverse situation is presented
for organic production. There is limited evidence that differing food processing methods have, or are
perceived to have, more or less health impacts. Regarding distribution and retail, and consumption, there
EVID4 Evidence Project Final Report (Rev. 06/11) Page 18 of 27
is evidence that better access to supermarkets relates to better access to fruit and vegetables, hence
greater consumption and hence better health. Generally access to supermarkets is more limited for certain
disadvantaged communities (see ethical dimension). Many consumers believe they eat healthily enough
and do not need to change their diets, and they are motivated by price and taste rather than claims of the
healthiness of foods. Impact on health is one aspect on which there is liable to be food-specific evidence,
and there is indeed a growing body of research that does identify the necessary components and
proportions of a healthy diet.
3.2.2. The Ethical dimension: This was also highly informed, particularly regarding the impact of
biotechnology e.g. GM for both plant- and animal-based foods, animal (including fish) welfare through to
the knowledge base concerning social inequality in relation to consumption. However, there appears to be
a paucity of information concerning aspects of processing, distribution and retail although the concept of
food miles has emerged as an issue (although considered to be utterly flawed as an indicator of
environmental sustainability). Social inequality is relevant with regards distribution and retail and
consumption, with evidence of reduced availability of fruit and vegetables in various disadvantaged
communities, and that a healthy diet is more expensive than an unhealthy one, so also impacting poorer
communities. In contrast, the other two dimensions (working conditions and societal) appear to be much
less informed.
3.2.3. Working conditions dimension: This seems to be relatively devoid of information. However, it is
likely that since they are probably dominated by legal requirements such as health and safety, working
hours directives etc., little funding would be forthcoming in this area for research. Furthermore, this
attribute is dependent on country of origin and, therefore, possibly more relevant for imported products
than those produced in the UK and Europe.
3.2.4. The societal dynamics dimension: This contains only limited evidence of relevance (which is not to
say that there isn’t a lot of published literature on aspects of the problem: there is, but little of this is
empirical or quantitative, and much is qualitative and open to interpretation), with the main contributing
knowledge relating to agriculture and employment and landscape. There are claims that organic farming
employs more workers than conventional farming, and so has potential sustainability benefits for
employment. Employment seems to cross social sustainability dimension boundaries, related to working
conditions too. There are also claims that organic farming appears better for biodiversity and landscape
diversity (positive social attributes). Interestingly, if the public do prefer patchwork landscapes over
uniformity, this suggests a more acceptable approach might be to have a mixture of eatwell components
produced in a particular landscape, rather than replacing one uniform component with another. Regarding
retail and distribution, it is claimed that farmers’ markets can help foster feelings of ‘community’ (though
evidence is weak). And regarding consumption, it has been demonstrated that decreased meat
consumption can increase available land for other activities – such as leisure.
3.3 Evidence relevant to assessing the Economic sustainability of a healthy diet
To examine the economic sustainability of ‘healthier’ diets, indicators were used to describe current
resource allocations and the impact of changes in demand. The indicators researched were land (area),
production, prices, value of output, employment and trade. This was carried out for selected products
within each of the five food groups on the eatwell plate. Products were selected to represent each group
but in many cases, the data sources identified would be relevant to other foods in the same group.
An extensive literature search was carried out to look for research on the economic impact of a shift to
more healthy diet, using key economic indicators identified above to measure impact effects.
All databases were searched according to product group. However, of around 65 references recorded in
the initial sifting, only 14 were product specific (of which 6 related to studies on fruit and vegetables).
Generally speaking, the literature is of a generic nature, not specific to food group or to stage of the supply
chain.
The list of 65 references that appeared relevant at first glance was then siphoned down to only 14 that
examine the effect of a shift to more healthy diets on specific economic indicators. A further 8 references
were selected that examine the economic effect of a shift in diets, but are less specific regarding the
impact on economic indicators. Reasons for exclusion included: studies published prior to 2000 that might
otherwise have been considered more relevant, studies that are related to the project but not necessarily
to the economics component, or studies that have a geographic basis considered less relevant to the UK
context.
The 14 specific references are summarised in Annex C. Those with more general findings regarding the
impact of a shift in diets are also summarised here. Where economic indicators were used in the studies
EVID4 Evidence Project Final Report (Rev. 06/11) Page 19 of 27
to measure impact, these are recorded in the Annex C. These tended to vary around the six indicators
identified earlier, namely land (area), production, prices, value of output, employment and trade.
As regards the selected economic impacts, our literature search revealed very few studies looking
specifically at the economic impacts of a shift in food consumption patterns. The key findings from the
most useful and methodologically robust studies, the majority of which focus on the United States, are
summarised in Annex C.
For each of the five food groups, an attempt was made to identify data sources for each of the six
economic indicators. The results of this search are tabulated in the main report. A summary of the
availability and robustness of the evidence is shown in Figure 7, colour coded according to the key.
3.3.1. Stage of supply chain
It is evident from Figure 7 that data availability is concentrated at the production end of the chain and, for
value, the retail/consumption end. For storage, transport and distribution, the majority of cells contain
sparse or no data sources. It would need specific research to calculate the product specific costs and
labour input for transporting a load of carrots, for example, from A to B. This then opens the debate as to
whether there is anything to gain from the calculation of some indicators for some stages of the chain.
3.3.2. Economic Indicators
3.3.2.1. Land (area) and production: Defra holds much of the data on the economic indicators specified
for the production stage of the chain. Contact with the relevant individuals at Defra also suggested that in
some cases it might be possible to draw on primary data sets to select data more specific to research
needs.
3.3.2.2. Prices: For prices, some data is available from Defra and for some products this is available from
the various representative trade associations or the ONS Focus on Consumer Prices publication.
Whereas data further back in the chain and for certain indicators is available at the product group level
only, prices tend to be product/variety/season/quality specific. Defra made reference to an indicator “price
per calorie” of food as a useful measure of comparing “healthy” with “less healthy” foods. This indicator is
referred to in The Food Statistics Pocketbook 2009, p. 42 (“Healthy foods often cost more per calorie”).
Sufficient information is available to calculate price per calorie.
3.3.2.3. Value of output: Value of output for the production stage of the chain is largely given in Defra’s
annual publication, “Agriculture in the United Kingdom”. For other stages of the chain, this data is largely
absent, except for consumption where value of consumption for some product groups is available from the
various representative bodies (e.g. AHDB, EBLEX, etc.).
3.3.2.4. Employment: Employment data is available from the Business Register and Employment Survey
(previously from the Annual Business Survey) specified by industry. It is difficult to break this down to
employment at the product level.
3.3.2.5. Trade: For trade, the primary source is HM Customs and Revenue. The difficulty here with
specifying products is that these are often grouped together. For example, while strawberries are
separately specified, lettuce is grouped with endive, making it difficult to give an exact figure for lettuce
imports and exports.
3.3.3. Product Groups:
Product groups were specified as in the eatwell plate food groups. A random selection of products within
each food group was selected to use as a basis for retrieving data. For some indicators, the data source
would apply to all products within the group (such as trade, employment), whereas for the more diverse
eatwell plate food groups such as “Meat, fish, eggs, beans and other non-dairy sources of protein”, for
some indicators the data sources were more varied.
In addition, supply chains were found to be much more complex than the basic production through to
consumption chain used as a basis for research. For a product such as potatoes there will be various
branches to the chain for seed potatoes, main crop, earlies, potatoes for processing, potatoes for fresh
consumption. With cereals, supply chains are even more diverse. In addition, there can be several
different types or varieties of the same product. Dessert apples are a good example, white or brown bread
is another. There is a need to specify the product carefully, bearing in mind the purpose of the data
collection exercise.
Whilst data specific to the UK is usually available from government and trade association sources, there
are a number of well known multilateral agencies providing data at a macro-level at the individual country
EVID4 Evidence Project Final Report (Rev. 06/11) Page 20 of 27
level as well as across groups of nations. This often draws on data available from individual nations.
Some of these sources include Eurostat, FAO, World Bank, IMF, OECD, UNCTAD and WTO.
Figure 7. Schematic representation of the availability of evidence concerning the economic sustainability
of the different eatwell plate food groups (based on the representative products shown in Table1).
Key:
GREEN
ORANGE
RED
GREY
Food group
Bread, rice, potatoes,
pasta and other
starchy foods
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Good evidence available in the form of generic comprehensive reports and refereed papers
Some evidence / limited evidence across sector
Insufficient evidence / no evidence on sector
Not applicable
Economic sustainability indicator
Land (area)
Production
Prices
Value of
output
Employment
Trade
Fruit and Vegetables
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Land (area)
Production
Prices
Value of
output
Employment
Trade
Milk and dairy foods
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Meat, fish, eggs,
beans, etc.
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Foods and drinks high
in fat and/or sugar
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Land (area)
Production
Prices
Value of
output
Employment
Trade
Land (area)
Production
Prices
Value of
output
Employment
Trade
Land (area)
Production
Prices
Value of
output
Employment
Trade
Total eatwell plate
Production
Processing
Transport & Storage
Distribution & Retail
Consumption
Land (area)
Production
Prices
Value of
output
Employment
Trade
EVID4 Evidence Project Final Report (Rev. 06/11) Page 21 of 27
4.0 The healthy diet and interrelationships between attributes from across the three pillars
Sections 3.1, 3.2 and 3.3 reveal that whilst some attributes are seemingly specific to each pillar, there may
be close interrelationships with attributes in other pillars. For example, the “Human toxicity potential” and
“ecotoxicity potential” attributes within the environmental sustainability pillar might well be expected to
have close associations with the “Health and welfare” attribute in the social pillar. It is also the case that
some attributes directly overlap. For example the “land use” attribute in the environmental sustainability
pillar would have a direct relationship with the “land (area)” attribute in the economic sustainability pillar
and landscape which falls within the “societal dynamics” attribute in the social sustainability pillar. Hence,
one would expect that any changes in the food supply chain that affect attributes of sustainability in one
pillar may well have related as well as separate impacts on attributes within the other pillars, either
positively or negatively. It is not surprising, therefore, that changes in diet could have a number of
possible effects on attributes across the pillars.
Recent examples of high level reports and publications that consider food consumption and sustainability
issues include a consideration of the opportunities for reducing GHG in the food chain (Garnett 2011), and
the impact of changing diet on GHG emissions (Audsley et al 2009; MacDiarmid et al, 2011). In all of
these publications, the focus has been predominantly on GHG emissions. Whilst this undoubtedly reflects
the generic interest in GHG emissions from a climate change viewpoint, it also reflects the limitations in
evidence across the other environmental attributes highlighted in Figures 2-4 as well as the lack of precise
evidence from the economic and social directions. The lack of readily available evidence and metrics
relevant to these other attributes means that such studies may highlight certain behaviour change as
environmentally beneficial from the GHG emissions viewpoint, but overlook other very important impacts,
both positive and negative (as acknowledged by Garnett 2011).
For example, the eatwell plate promotes the consumption of more vegetables, fruits and starch-based
products, and a reduction in meat and dairy consumption compared with current UK dietary behaviour. A
significant reduction in the demand for meat and dairy products by UK consumers might be expected to
result in a decline in UK demand for meat production (assuming that the current output could not be
supported by demand from overseas). This, in turn, would be expected to reduce the environmental
impacts of cattle production (Figure 4) and of importing feed. However, increased consumption of
vegetables and fruits might require increased imports to provide supplies out of season. There would also
be accompanying changes in land use which might involve the increased production of food crops, maybe
fruits and vegetables, or perhaps energy crops. From the social sustainability viewpoint, the changes in
landscape may be deleterious if cattle husbandry was to be replaced by monotonous crop monocultures
(indeed, it may be that people would prefer a mixed landscape over any singular usage). From the
economic sustainability viewpoint, such changes would have substantially negative impacts on the meat
and dairy food chains, reducing commercial and employment opportunities in these sectors particularly in
the agronomy, processing, transport and distribution and retail sectors. There would also be an associated
impact on supply and employment related to non-food industries using by-products from livestock (e.g.
adhesives, leather, gelatine etc). Increased unemployment would be expected to have negative social
connotations. From the environmental perspective, significant impacts on biodiversity would also be
expected. Thus, consuming less meat and dairy products may have positive benefits to health and to
some environmental attributes, but may result in negative impacts on other environmental, economic and
social sustainability attributes.
The current study demonstrates that for such an evaluation, there is probably sufficient information to
evaluate the economic impacts, for example, on employment and production and prices. From the social
perspective, there is probably sufficient information to evaluate aspects of health but little on cultural
impacts or working conditions. From the environmental perspective, it is clear that from LCA reports, only
a limited number of attributes are covered as shown in Figure 3. However, there may be sufficient nonLCA information concerning biodiversity, water footprint, and reports on UK food waste to be able to make
a reasonable inference as to the effects of dietary change.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 22 of 27
5.0 Conclusions
1) There is a need to produce a more coherent definition of ‘sustainability’. This is particularly true in the
case of ‘social sustainability’. One way to do this might be to convene an expert workshop to debate
and decide an appropriate definition – for Defra or more widely. Perhaps as part of this process, a
number of case studies could be considered/worked through to crystalize the definitions and how these
might be measured. Existing tools for measuring sustainability in different (e.g. non-food) contexts
might also be worth considering, such as Social Lifecycle Analysis, and frameworks and guidelines
such as the UN’s Sustainable Reporting Guideline and Sustainable Development Indicators (e.g. see
Annex B).
2) The evidence base for evaluating the sustainability implications of changing to a healthy diet
predominantly relates to environmental impacts (with some caveats) and relatively less to economic
and social impacts:
a. Evidence for attributes of environmental sustainability:
i.
ii.
iii.
There is sufficient robust evidence concerning three out of thirteen recognised
environmental sustainability attributes particularly at the farm gate end of the production
chain. These attributes are GWP, EP & AP. There is also limited information on the
following five attributes: POCP, ODP, RE, LU & WU. This conclusion has been readily
drawn by evaluating a large number of high quality LCA studies.
However, the remaining five sustainability attributes receive little or no attention in the LCA
literature for food & drinks (except for selected products). The gaps in knowledge
associated with these attributes include impacts on human toxicity (HTP), ecotoxicity (ETP),
abiotic / biotic resource depletion (ARD, BRD) and biodiversity (BD).
There is a large, knowledge base in the scientific literature which is not integrated with LCA
which provides additional information of relevance to sustainability (e.g. fisheries statistics,
water footprint studies).
b. Evidence for attributes of economic sustainability:
i.
ii.
This is also biased towards production, but is more generic and robustness is patchy.
The literature is of a generic nature, and most is not specific to food food group or to stage
of the supply chain.
c. Evidence for attributes of social sustainability
i.
These are not particularly product specific and at a higher level than the environmental and
economic indicators. That is, indicators are linked mostly to how a food product is produced,
processed, distributed and consumed, rather than the specific type of food. There may thus
be a need to approach social sustainability in a different way to the other sustainability
types.
3) There is a pressing need to evaluate the above information in relation to UK consumption data with
particular reference to the most commonly consumed foods.
4) Additional studies are required which adopt a more targeted and consequential approach as follows:
a. It is necessary to prioritise a list of key sustainability attributes across the three pillars. The
relative importance of these attributes may differ according to, for example, their topicality (e.g.
climate change is much more heavily featured in the press than other environmental attributes
such as water use or resource depletion) or the country of production where working conditions,
employment or local biodiversity (e.g. farmland birds) may be an issue.
b. Using the agreed key sustainability attributes, to examine in detail a change in consumption from
one commonly-consumed product to another. Such a study should begin with a detailed change
in the total quantities of food products consumed in the UK and then:
i.
ii.
iii.
Consider alternatives for how the additional quantities of products consumed are produced
(and where they will come from); together with assumptions about how the reduced
quantities of consumption of other specific food products are now being produced (and
where they come from). This allows incorporation of social sustainability.
Assess the potential impact on the agreed sustainability attributes across all three pillars.
Evaluate how and where changes measured by the key environmental attributes might
impact on the social and economic attributes (and vice versa).Understanding the trade-offs
between the three pillars will be crucial in determining the ultimate sustainability of a healthy
diet. Linked to this, future research could be undertaken into potential metrics that might be
used to assess overlaps and tensions between the three pillars of sustainability and the
EVID4 Evidence Project Final Report (Rev. 06/11) Page 23 of 27
impact these have on overall sustainability. The Scottish Government’s forthcoming Health
and Environmental Sustainability Framework for Food and Drink could be a useful tool to
examine here, for example, and the Swedish and Dutch governments also seem to be
making some headway in this area. The aim of this framework is to help guide decision
makers at all stages of the food and drink supply chain (and right across it) in how to
integrate their thinking about health and environmental sustainability, so that they can
prevent actions taken to address one of these considerations having an adverse effect on
the other. The Framework is focused on decision makers in both the public and private
sector, but does not include consumers. The approach should also include assessing the
possibility of “spillover” where sustainability associated with production and consumption
activity is displaced into other sectors. For example, as pointed out by Garnett (2011), even
if consumers change to eating less meat, any financial savings made may just be used to
carry out other GHG-emitting activities such as recreation and travel. This re-bound effect
has also been identified in energy efficiency behaviour change. However, few studies are
able quantify this satisfactorily due to the complexity of understanding the environmental
impact of people’s behaviour resulting from a relatively small increase in wealth due to
energy savings, (Sorrell 2007). Furthermore, since the food supply is global, impacts of
changing UK dietary patterns may influence food production in other countries.
Indeed, the current Defra project FO0427 ‘Future implications of trends in healthy eating on
existing food production and manufacture’ will model possible future UK diet scenarios and
determine their environmental, social and economic impacts.
c. It may also be appropriate to extend studies such as the Livewell study (Macdiarmid et al, 2011)
to evaluate and as far as possible estimate the impacts of dietary change on a greater range of
sustainability attributes than just greenhouse gas emissions. This could provide much greater
insight into the impacts of change than focusing on numerous low level studies. This approach
may be augmented by using a “checklist” to ensure that change between elements of the diet
does not lead to unintended consequences in other attributes of sustainability e.g. changes in
biodiversity, working conditions or through deforestation. A checklist approach might be
especially apt for considering social sustainability, for example, enabling comparison of the
impacts of different production / processing / distribution / consumption methods on low income
groups, ethnic minorities, the mobility impaired, rural communities and other vulnerable social
groups (for which tools such as the Index of Multiple Deprivation might be useful, as one aspect
considered is ‘access to food shops’ – see, e.g.,
http://www.communities.gov.uk/communities/research/indicesdeprivation/). It could also be used
as a way of documenting social attitudes to how different elements of a healthy diet are
produced / processed / distributed / consumed.
5) Food losses along the food chain should be intrinsically accounted for as part of good life cycle
assessment practice, (where the goals of the assessments demand), particularly in the light of the
recent FAO (Gustavsson et al, 2011) and WRAP (Waste & Resources Action Programme, 2009; 2010
a&b) reports. However, the boundaries of many of the studies found in this review do not consider the
impact of food waste at the end of the food chain for specific foods. Understanding differences in food
waste behaviour for specific foods may add a significant multiplier effect and few data sources were
found that quantified this. More research to understand domestic food waste factors could inform life
cycle assessment of foods where impacts may then be reported per kg actually eaten, which more
precisely defines the function of food.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 24 of 27
Acknowledgement
The Project Team wish to express their thanks and gratitude to the following members of the Project
Steering Group and Expert Panel for their guidance, assistance, constructive criticism & comment
throughout the course of this Project.
Steering Group
Name
Institution
Roland Clift
Helena Crow
Paul Finglas
Lucy Foster
Janice Kerr
Sonia Molnar
Thomas Ohlsson
Christopher Ritson
Bruno Viegas
University of Surrey
Scottish Government
Institute of Food Research
Defra
Defra
Defra
SIK – the Swedish Institute for Food and Biotechnology
University of Newcastle
Defra
Expert Panel
Name
Institution
Eric Audsley
Simon Bell
Dylan Bradley
Mary Brennan
Judy Buttriss
Ashok Chapagain
Peter Dennis
Elizabeth Dowler
Georg Engelhard
Tara Garnett
David Harvey
Mark Holmes
Robert Lillywhite
Tom MacMillan
Stephen Morse
Nigel Mortimer
Richard Murphy
Richard Pywell
David Tickner
Bruce Traill
Kairsty Topp
Angela Tregear
Wim Verbeke
Bill Vorley
Peter Whitehead
Adrian Williams
Jeremy Wiltshire
Cranfield University
Open University
Agra CEAS Consulting
University of Newcastle
British Nutrition Foundation
WWF
University of Aberystwyth
University of Warwick
cefas
Food & Climate Research Network
University of Newcastle
ADAS
University of Warwick
Food Ethics Council
University of Surrey
North Energy Associates
Imperial College
Centre for Ecology & Hydrology
Head of Freshwater Programmes, WWF
University of Reading
Scottish Agricultural College
Edinburgh University
University of Ghent
International Institute for Environment and Development
IGD
Cranfield University
ADAS
EVID4 Evidence Project Final Report (Rev. 06/11) Page 25 of 27
References to published material
9.
This section should be used to record links (hypertext links where possible) or references to other
published material generated by, or relating to this project.
Audsley, E., Chatterton, J., Graves, A., Morris, J., Murphy-Bokern, D., Pearn, K., Sanders, D. and Williams,
A. (2010). Food, land and greenhouse gases: the effects of changes in UK food consumption on land
requirements and greenhouse gas emissions. Report prepared for the United Kingdom’s Government’s
Committee on Climate Change, available online at
http://downloads.theccc.org.uk.s3.amazonaws.com/4th%20Budget/fourthbudget_supporting_research_cran
field_dietsGHGLU_agriculture.pdf, accessed 17 May 2011.
Bassett-Mens, C. & van der Werf, H.M.G. (2005). Scenario-based environmental assessment of farming
systems: the case of pig production in France, Agriculture, Ecosystems and Environment, 105, 127-144.
Bassett-Mens, C., van der Werf, H.M.G., Robin, P., Morvan, Th., Hassouna, M., Paillat, J.-M. & Vertès F.
(2007). Methods and data for the environmental inventory of contrasting pig production systems, Journal of
Cleaner Production, 15, 1395-1405.
Broom, D.M. (2010) Animal Welfare: An Aspect of Care, Sustainability, and Food Quality Required by the
Public, Journal of Veterinary Medical Education, 37(1), 83-88.
Cederberg, C. & Stadig, M. (2003). System expansion and allocation in life cycle assessment of milk and
beef production, International Journal of Life Cycle Assessment, 8 (6), 350-356.
eatwell plate, http://www.dh.gov.uk/en/Publichealth/Nutrition/DH_126493 , accessed 26 July 2011.
Ecoinvent Centre, Swiss Centre for Life Cycle Inventories, Dübendorf, Switzerland, www.ecoinvent.org
FAO, (2011) Background document to E-Forum on Sustainability Assessment of Food and Agriculture
systems (SAFA), see
http://www.fao.org/fileadmin/user_upload/suistainability/docs/Background_Document_01.pdf , last
accessed 22 March 2011.
FSA, 2002, McCance and Widdowson’s The Composition of Foods, Sixth summary edition. Cambridge:
Royal Society of Chemistry
Garnett, T. (2011). Where are the best opportunities for reducing greenhouse gas emissions in the food
system (including the food chain)?. Food Policy, 36, S23-S32.
Gustavsson J., Cederberg C., Sonesson U., van Otterdijk R. & Meybeck A., (2011), Global food losses and
food waste: extent, causes and prevention, Food and Agriculture Organisation of the United Nations,
Rome, Italy.
KRAV, (2010). Climate Smart KRAV Fish: Emissions of greenhouse gases from a 400g pack of cod. A
comparison of KRAV-approved cod and average cod, available online at
http://www.krav.se/Documents/Engelska%20sidor/20100528%20Climate%20Smart%20Fish%20Report%2
0-%20final.pdf, (accessed 15 June 2011).
Lewis K.A., Green A., Tzilivakis J. & Warner D.J. (2010). The contribution of UK farm assurance schemes
towards desirable environmental policy outcomes, International Journal of Agricultural Sustainability, 8, 4,
237-249.
Macdiarmid, J., Kyle, J., Horgan, G., Loe, J., Fyfe, C., Johnstone, A. & McNeill, G. (2011). Livewell: a
balance of healthy and sustainable food choices. WWF-UK.
Olesen, I., Groen, A.F. & Gjerde, B. (2000) Definition of animal breeding goals for sustainable production
systems, Journal of Animal Science, 78 (3), 570-582.
PE International, Leinfelden-Echterdingen, Germany, www.gabi-software.com, (accessed 14 June 2011)
Pré Consultants bv, Amersfoort, The Netherlands, www.pre.nl, (accessed 14 June 2011)
Sorrell, S. (2007) The rebound effect: an assessment of the evidence for economy-wide energy savings
from improved energy efficiency http://www.ukerc.ac.uk/support/tiki-index.php?page=ReboundEffect
Van Calker K.J., Berentsen P.B.M., Romero C., Giesen G.W.J. & Huirne R.B.M., Development and
application of a multi-attribute sustainability function for Dutch dairy farming systems, Ecological
Economics, (2006), 57, 640-658.
WRAP, (2009), Household Food and Drink Waste in the UK, Report prepared by WRAP, Banbury, UK.
WRAP, (2010a), Waste arisings in the supply of food and drink to households in the UK, Report prepared
for WRAP, Banbury, UK.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 26 of 27
WRAP, (2010b), Reducing food waste through the chill chain. Part 1: Insights around the domestic
refrigerator, Report prepared for WRAP, Banbury, UK.
EVID4 Evidence Project Final Report (Rev. 06/11) Page 27 of 27