EU BIOFUEL USE AND AGRICULTURAL COMMODITY PRICES:
A REVIEW OF THE EVIDENCE BASE

Report prepared for Action Aid
June 2012

Authors
Bettina Kretschmer
Catherine Bowyer
Allan Buckwell

Disclaimer: The arguments expressed in this report are solely those of the authors, and do
not reflect the opinion of any other party. Any errors that remain in the report are the
responsibility of the authors.
To be quoted as: Kretschmer, B, Bowyer, C and Buckwell, A (2012) EU Biofuel Use and
Agricultural Commodity Prices: A Review of the Evidence Base. Institute for European
Environmental Policy (IEEP): London.
Funding from ActionAid is gratefully acknowledged.
Credentials for pictures on title page (from top): Martin Vidner via flickr, www.fao.org,
SAN_DRINO via flickr, Benediktv via flickr
We are grateful for helpful comments on the report from David Baldock and Pernille
Schiellerup. We furthermore thank Dearbhla Keegan and Paul van Dam for research
support.

Institute for European Environmental Policy
London Office
15 Queen Anne's Gate
London, SW1H 9BU
Tel: +44 (0) 20 7799 2244
Fax: +44 (0) 20 7799 2600
Brussels Office
Quai au Foin, 55
Hooikaai 55
B- 1000 Brussels
Tel: +32 (0) 2738 7482
Fax: +32 (0) 2732 4004

The Institute for European Environmental Policy (IEEP) is an independent not-for-profit
institute. IEEP undertakes work for external sponsors in a range of policy areas. We also
have our own research programmes and produce the Manual of European Environmental
Policy (http://www.europeanenvironmentalpolicy.eu/). For further information about IEEP,
see our website at http://www.ieep.eu or contact any staff member.

2

Contents

EXECUTIVE SUMMARY.......................................................................................................... 5
1

INTRODUCTION ............................................................................................................. 8

2

EU POLICY DRIVING BIOFUELS DEMAND AND OUTLOOK FOR THIS DEMAND TO 2020 . 9

2.1 What is the current scale of EU demand for biofuels and how is this likely to
develop to 2020? .................................................................................................................. 9
2.2 What kind of agricultural crops are currently used to meet EU biofuel demand and
how is this likely to develop to 2020? ................................................................................ 11
3 THE FOOD VERSUS FUEL DEBATE IN THE EU: A BRIEF REVIEW OF SOME OF THE MAIN
DIMENSIONS ...................................................................................................................... 13
3.1

Why the Concern? ................................................................................................... 13

3.2

Controversy over the fuel and food interaction....................................................... 14

3.3 Determining the extent of biofuel land demands generated by the EU mandate –
exploring yield, intensification and use of by-products...................................................... 16
3.3.1
Turning by-products into usable co-products..................................................... 17
3.3.2
Increasing Yield and the potential for agricultural intensification ...................... 18
3.3.3
In summary ....................................................................................................... 20
4 THE ROLE OF VARIOUS KEY FACTORS IN DETERMINING AGRICULTURAL COMMODITY
PRICES AND THEIR VOLATILITY ........................................................................................... 22
4.1

A little historical context .......................................................................................... 22

4.2

Key factors influencing global agricultural commodity prices ................................. 23

4.3

Factors influencing demand ..................................................................................... 25

4.4

Factors influencing supply ....................................................................................... 26

4.5

Other factors influencing international agricultural commodity prices ................... 28

4.6

In summary .............................................................................................................. 31

5 THE ROLE OF EU BIOFUEL INCENTIVES ON AGRICULTURAL COMMODITIES PRICES TO
2020: A REVIEW OF SELECTED MODELLING STUDIES .......................................................... 32
5.1

The role of models and our approach to reviewing them ........................................ 32

3

5.1.1
use
5.1.2

The role of models in understanding the agricultural market impacts of biofuel
32
Key characteristics of interest ............................................................................ 33

5.2 Results of the review ............................................................................................... 37
5.2.1
Findings from studies focusing on the effects of EU biofuel policy ..................... 41
5.2.2
Findings from studies analysing global biofuel mandates................................... 44
5.3 Implications for the significance of EU biofuel policy in relation to projected
agricultural commodity prices by 2020? ............................................................................. 48
6

IMPLICATIONS FOR POLICY ......................................................................................... 50

7

CONCLUSIONS ............................................................................................................. 54

REFERENCES ....................................................................................................................... 56
ANNEX 1: LAND REQUIREMENTS FOR BIOFUELS ................................................................ 64
ANNEX 2: SUMMARY TABLE OF STUDIES REVIEWED .......................................................... 66

4

EXECUTIVE SUMMARY
Biofuel mandates are being used in many parts of the world to encourage the substitution
of fossil fuels in transport uses by renewable energy. Brazil has long had an extensive
programme of producing bioethanol from cane sugar, the USA has massively expanded
production of ethanol from corn and the EU’s Renewable Energy Directive has led to a scale
up in EU biofuel use, most notably of rapeseed biodiesel.
The European Commission will produce a report reviewing the social impacts of EU biofuel
policy by the end of 2012. To support Action Aid in its participation in the debate about the
future of EU biofuels policy, taking into account the social dimension, this report was
commissioned to provide an independent review of the evidence base on the link between
policy-driven EU biofuels demand and global agricultural prices – in particular through
reviewing a selection of modelling based studies. The next step in the causal chain from
agricultural prices to food price impacts, ie prices to consumers on the ground, whilst highly
important lies outside the scope of the study.
The 10 per cent target for the use of renewable energy in transport set in the EU Renewable
Energy Directive is anticipated to lead to a tripling of biofuel use in the EU in 2020 compared
to 2008 levels, according to the National Renewable Energy Action Plans that all Member
States have been required to produce. The plans confirm that the current dominance of
biodiesel over ethanol in the European biofuel market will continue. The EU’s target
predominantly will be met by first-generation biofuels produced from traditional food and
feed crops, translating into a significant additional demand for these crops. It is clear that
this additional demand, alongside the growing global demand for food and biofuel demand
elsewhere will increase agricultural commodity prices but by how much is the subject of
continued debate. The uncertainty about the extent of price increases arises from a range of
factors, including the prevalent interactions between different crops and livestock markets
and between world regions and from the responses to price signals by consumers and
producers. Yield increases stimulated by higher output prices is often suggested as a factor
with the potential to mitigate price increases; another may be related to the rise of protein
by-products as animal feed.
The complexities of interacting factors can only be captured in reasonably sophisticated
economic models and we have reviewed a range of both partial and general equilibrium
models for this report. Nonetheless, because of the wide variety in the context, scope and
methodology of models used to analyse the effects of biofuel mandates, the results of
analyses of the price effects range substantially. Consequently, it is often extremely difficult
to draw comparisons between models.
In those modelling studies focusing on the impacts of EU (as opposed to global) biofuel
policies, the most significant price increases are projected for oilseeds and vegetable oils,
with increases in world prices by 2020 typically ranging between 8 to 20 and 5 to 36 per
cent, respectively (see table below). Wheat prices are projected to increase by between 1
and 13 per cent and the majority of studies project increases of cereal / maize prices of up
to 8 per cent and of sugar prices of up to 2 per cent. One model (ESIM) projects these
increases to be 22 and 21 per cent, respectively. Many of the drivers of differences in results
5

are those that have been under scrutiny in the indirect land use change debate.
Consequently, the studies at the forefront in that debate can be expected to deliver the
most robust results with regard to agricultural market impacts. This is most notably the
IFPRI study, which projects increases in world rapeseed prices (anticipated to be the most
significant feedstock for EU biofuel use in 2020) of around 11 per cent (Laborde, 2011).
While looking at studies that model global biofuel policies does not allow singling out the EU
policy impact, it is in some way a more complete scenario design, given EU policies in the
real world do not take place in isolation but other countries have in place biofuel policies as
well. At least some of the global studies estimate substantially higher price effects for
ethanol crops such as wheat, other cereals and sugarcane.
Table: Summary of price effects per feedstock
Feedstock (group)

Range of price effects

Commentary

Studies that focus on the effects of EU biofuel policy
Oilseeds
8 – 20%
Vegetable oils
1 – 36%
Oilseeds
9 – 20%
Cereals / maize
1 – 22%
The ESIM model (Blanco Fonseca et al, 2010)
projects an increase in maize prices of 22%.
The remaining studies project increases in
maize or cereal prices of ≤8%
Wheat
1 – 13%
Sugar (cane/beet) 1 – 21%
The ESIM model (Blanco Fonseca et al, 2010)
projects an increase in sugar prices of 21%. The
remaining three models reporting results for
sugar project price increases of ≤2%
Studies that analyse the impacts of global/multi-regional biofuel mandates
Oilseeds
2 – 7%
Vegetable oils
35%
OECD (2008) is the only ‘global’ study providing
a figure for vegetable oils
Cereals / maize
1 – 35%
Wheat
1 – 8%
Sugar (cane/beet) ~10%
Timilsina et al (2010) is the only ‘global’ study
providing a figure for sugar, 9.2 or 11.6%
depending on the scenario
Source and notes: see section 5.3

In terms of options to deal with biofuel induced price increases: a reduction, or even
abolition, of policy induced biofuel demand will reduce or prevent altogether agricultural
market and hence price impacts due to biofuel policy. Depending on the development of oil
and therefore fossil fuel prices on the one hand and feedstock prices on the other, the
market might still drive increased biofuel use whenever it is price competitive with fossil
fuels, rendering policy changes less or altogether ineffective. Unlike in the USA, where
biofuels are promoted by a rigid mandate laying out annual targets for biofuel volumes
consumed, EU biofuel targets are less specific, arising de facto from a wider renewable
energy target. At present there is at least an element of flexibility afforded to Member
6

States in terms of how they choose to go about meeting the target. They are obliged to
deliver a certain proportion of renewable energy in transport rather than to meet specific
target volumes of biofuels. Consequently, policy can be adjusted in different ways – either
by altering the target for renewables, setting more specific targets for different forms of
renewable energy (eg from wastes), increasing the efficiency of transport vehicles etc. For
example, EU policy that would encourage Member States to increase support for advanced
biofuels produced from forestry and agricultural residues as well as renewable electricity
while scaling back support for conventional biofuels produced from food and feed crops,
would be a way of reducing pressure on agricultural markets. These alternative options for
promoting renewable energy and increasing policy flexibility need further detailed
consideration. More clarification could usefully be provided by the Commission regarding
the scope of Member State autonomy to adapt their policies up to 2020.
This review assessed the impacts of biofuel use on agricultural commodity prices. The
European Commission, as part of its reporting requirements in 2012, is required to judge the
impacts of biofuels on food prices. This further step increases the complexity of the analysis
substantially. Food supply chains involve an extremely wide range of plant and animal
products with varying degrees of processing of agricultural commodities and varying
proportions of staple foodstuffs in the diet. The vulnerability of consumers across the world
to food price increases differs markedly between countries and across households,
depending inter alia on income levels, household composition, and on the household status
as net consumers or producers of agricultural and food stuffs. We are not aware of studies
that use multi-household models, which would produce a better understanding of how
commodity price rises induced by biofuels are transmitted to different population groups
and allowing more solid estimates of the welfare impacts of biofuel policy. This gap should
be closed in order to provide decision makers with a more complete evidence base feeding
into the political review processes ongoing in 2012.

7

1

INTRODUCTION

In 2012, the European Commission (EC) will produce a report reviewing the social impacts of
EU biofuel policy. Under Article 17 (5) of the Renewable Energy Directive (RED)1, the EC is
obliged to conduct this review into the impacts on 1) food affordability and availability; 2)
land rights; 3) whether producer countries have implemented various International Labour
Organisation conventions and 4) ‘wider development issues’. To support Action Aid in its
participation in the debate about the future of EU level biofuels policy, it commissioned IEEP
to provide an independent review of the evidence base on the link between policy driven
EU biofuels demand and global agricultural prices in particular through reviewing a selection
of modelling based studies. The next step in the causal chain from agricultural prices to food
price impacts, ie prices to consumers on the ground, whilst highly important lies outside the
scope of the study. In Section 5 we provide a brief discussion of the complexity of the
relationship between global agricultural and food prices.
This report is structured as follows: Section 2 gives a brief overview of EU policy driving
biofuel demand and sets out what is currently known about the likely scale and nature of
this demand to 2020. Section 3 outlines the major concerns and uncertainties in the
discussions on biofuel use and agricultural and food price impacts or, in other words, the
‘food versus fuel’ debate. To separately analyse the effects of biofuel policy we have to
understand the wider context of the high volatility in agricultural commodity prices
experienced in recent years. Section 4 gives an overview of some of the most important
factors shaping agricultural commodity prices. Section 5 reviews the results of a selection of
modelling based studies examining the impacts of EU and global biofuel policies on
agricultural markets and prices. Section 6 spells out some implications that might be drawn
for policy on the basis of the modelling studies. Section 7 provides a brief conclusion.

1

Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on the promotion of
the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC
and 2003/30/EC.
8

2

EU POLICY DRIVING BIOFUELS DEMAND AND OUTLOOK FOR THIS DEMAND TO 2020

In this section we briefly summarise the main dimensions of EU policy driving biofuels
demand and provide an outlook for this demand to 2020 in terms of the overall scale of
demand, as well as the nature of demand (ie what kind of crops will be used).
Biofuel use was first promoted in the EU
by the 2003 ‘biofuel directive’2, stipulating Box 1: First- and second-generation biofuels
an indicative target for EU Member States Biofuels can replace conventional diesel (biodiesel)
of 5.75 per cent biofuel use in road and petrol (bioethanol) as transport fuels. Socalled first-generation biofuels are derived from
transport fuels by 2010. There was limited
food and feed crops. The crushing of different oil
progress towards this target in many crops including rapeseed, soy, palm and jatropha
Member States, leading to the (the latter not used for food or feed purposes)
introduction in 2009 of the new binding yields oils that are transformed into biodiesel
target in the Renewable Energy Directive through transesterification. Bioethanol is
constituted a significant further step. The produced by fermenting the sugar components
Renewable Energy Directive calls for an derived from sugar and starchy crops such as
EU-wide 20 per cent renewable energy sugar cane and beet, wheat, rye and maize.
share in gross final energy consumption to Advanced or second-generation technologies are
be reached by 2020. Apart from this being developed that convert lingo-cellulosic
overall target, the RED mandates a 10 per material (including woody crops and forest and
agricultural residues) into biofuel. These offer the
cent renewable energy share in transport
possibility of utilising biomass which is less directly
to be met in each Member State by 2020. competitive for food and feed, and also capable of
According to the National Renewable yielding a much higher energy return.
Energy Action Plans (NREAPs) which the
Directive requires governments to
prepare around 90 per cent of the transport target will be met by first-generation biofuels in
2020.
2.1

What is the current scale of EU demand for biofuels and how is this likely to develop
to 2020?
Figure 1 illustrates the development of EU biofuel consumption over the last decade
showing the importance of EU policy in triggering uptake in the more recent years.
To give an idea about the breakdown of the aggregate position for the EU in Figure 1: the
biggest biofuel consuming countries in 2010 were (in descending order) Germany, France,
Spain, Italy and the UK. Biodiesel accounted for over three quarters of 2010 EU biofuel use.
Germany, France and Spain lead EU biodiesel production in 2010 (EUROBSERV’ER, 2011). EU
bioethanol production is dominated by France, Germany and Spain (2009 data) 3.

2

Directive 2003/30/EC of the European Parliament and of the Council of 8 May 2003 on the promotion of the
use of biofuels or other renewable fuels for transport.

3

http://www.biofuels-platform.ch/en/infos/eu-bioethanol.php
9

Figure 1. Development of biofuel consumption for transport in the EU 27, 2000 to 2010 (in mtoe)

Source: EUROBSERV’ER, 2011, p74

An outlook for future EU biofuel consumption is contained in the compendium of National
Renewable Energy Action Plans (NREAPs) prepared by Member States4. Bowyer (2011) has
analysed the Member States’ NREAPs yielding the following observations, summarised in
Table 15. In order to meet the 10 per cent target:
 The 27 EU Member States will consume 29.6 Mtoe of biofuels in 2020. This
translates into an increase in biofuel use between 2008 and 2020 of 19.5 Mtoe.
 The majority will be conventional first-generation biofuels, making up about 92 per
cent of total predicted biofuel use or 27.3 Mtoe in 2020, equating to 8.8 per cent of
the total energy in transport. Advanced or second-generation biofuels will not gain
an important market share and are anticipated to account for only 0.7 per cent (2.1
Mtoe) of total energy in transport by 2020.
 Out of first-generation biofuels, 72 per cent are anticipated to be biodiesel and 28
per cent bioethanol; anticipated imports amount to 44 per cent of bioethanol and 36
per cent of biodiesel used in 2020 (see Table 1). Actual imported levels of feedstock
could be even higher as it is unclear whether the figures for anticipated imports

4

Member States subsequently are required to report on progress towards the binding renewable energy
targets on a biannual basis; the first of these reports were due at the end of 2011. These can give an
updated picture of the amount of biofuels Member State anticipate using in 2020. At the time of writing, not
all reports had been submitted yet and it was beyond this study to analyse those already submitted.

5

A detailed analysis of the NREAPs by Beurskens et al (2011) for the Commission yields additional information
on biofuel pathways to 2020 in the different Member States (see tables on projected total bioethanol/bioETBE and biodiesel use in renewable transport over 2005-2020 on p178 and p184, respectively).
10

reported in the NREAPs also include imported feedstock for ‘domestic’ processing
into biofuels or only refer to processed biofuels.
Table 1: The projected composition of EU biofuel consumption in 2020 according to NREAPs
Expected quantities of Total
EU biofuels in 2020
Mtoe
Consumption

29.6

Biodiesel
Mtoe
21.3

Imports

11.3

7.7

Production

18.3

13.7

Bioethanol
share
Mtoe
share
72% of total 8.3
28% of total
biofuels
biofuels
36%
of 3.6
44% of ethanol
biodiesel cons
cons
74.4% of total 4.7
25.6% of total
biofuels
biofuels

Source: Own compilation based on Bowyer (2011)

One question arising in this context is how EU Member States have derived their NREAP
estimates, ie whether these are targets or predicted behaviours given certain signals and
other assumptions and how accurate the estimates are expected to be. It seems reasonable
to expect diverse answers to these questions across the different Member States,
depending on the amount of resources, including modelling techniques, national authorities
put into the elaboration of their NREAPs. NREAP estimates should reflect MS policies in
place, including differentiated support for first-generation versus advanced biofuels versus
renewable electricity in transport. This topic would merit further examination 6.
2.2

What kind of agricultural crops are currently used to meet EU biofuel demand and
how is this likely to develop to 2020?
Tables 2 and 3 below show the shares of feedstocks for both EU biodiesel and bioethanol
consumption in 2008 and 2020, respectively. For 2008, figures compiled under the ‘Biofuel
Baseline 2008’ project by Ecofys et al (2011) have been used. Projections for 2020 are
derived from one of the ILUC modelling studies prepared for the European Commission
(Laborde, 2011).
Currently, rapeseed/oil dominates the market of biodiesel destined for EU consumption,
amounting to over half of the total feedstock used. This is followed by soybean and palm oil
with similar shares and a few feedstocks of minor importance. The resource base is more
diversified for ethanol, with sugar cane and beet and wheat dominating current markets
Moving to projected feedstock shares in 2020 (table 3): rapeseed continues to dominate the
biodiesel market, accounting for 57 per cent of the feedstock in 2020. Palm oil makes up
another fourth of all feedstock used, with the remainder from soy and to a lesser extent
sunflower. The market for ethanol destined for EU consumption is projected to be more
heavily reliant on sugar cane in 2020 as opposed to 2008. Sugar cane is projected to account
for almost half of all feedstock. This is followed by wheat, sugar beet and maize. These
6

Work under the Biomass Futures project with the energy sector model RESolve contributes to this:
www.biomassfutures.eu.

11

projections are derived from the IFPRI study (Laborde, 2011) whose scenarios are in line
with the shares of and absolute volumes of biodiesel and ethanol Member States anticipate
using in 2020. Taken together with accounting for observed usage of feedstocks in 2008 in
the baseline, the 2020 figures should give a fairly good picture of the biofuel market in 2020
and the relative importance of different feedstocks. But of course the figures are not free
from uncertainty and like any other results derived from the model depend on factors such
as assumed (relative) crop yield developments and future crop demand from other sectors.
Table 2: Shares of feedstock for biofuels consumed in the EU in 2008
Biodiesel
% share
Ethanol
% share
Rapeseed/oil
Soybean/oil
Palm oil
Tallow
RVO
Sunflower/oil

55
19
16
5
4
2

Sugar beet
Sugar cane
Wheat
Maize
Wine
Rye
Barley
Triticale
Other

23
23
21
13
8
3
1
0.4
8

Source: own calculation based on Ecofys et al, 2011, Tables 14 and 15
Note: RVO is recycled vegetable oil.

Table 3: Shares of feedstock for biofuels consumed in the EU in 2020
Biodiesel
% share
Ethanol
% share
Palm Oil
Rapeseed
Soy
Sunflower

24
57
15
6

Maize
Sugar Beet
Sugar Cane
Wheat

14
18
46
21

Source: own calculation based on Laborde, 2011, Table 3
Note: Percentages from Laborde’s Table 3 are recalculated taking the setting of biodiesel=100% and
all ethanol=100% (instead of all biofuel=100% as in the original table). Percentages for biodiesel
feedstock yield slightly over 100% when summed, due to rounding in the original table.

12

3

THE FOOD VERSUS FUEL DEBATE IN THE EU: A BRIEF REVIEW OF SOME OF THE MAIN
DIMENSIONS

3.1 Why the Concern?
EU mandates and policy support for biofuels, and for biomass use for heat and electricity,
generate additional demand for feedstocks and therefore for land on which to grow them,
with consequences for land use both in Europe and globally (given the international nature
of commodity markets). This situation gives rise to a series of unintended impacts and
therefore raises doubts about the sustainability of the transport target, in particular
whether it actually contributes to reducing GHG emissions from transport, a key underlying
rationale for the whole policy. In this context, 2010 and 2011 have seen an intense
discussion about the indirect land use change (ILUC) impacts of EU biofuel use 7. While this
discussion continues, the year 2012 increasingly will bring the social impacts of biofuel use
on to the agenda given the Commission’s reporting obligations. Already the impacts of
biofuels on food prices has been fiercely debated in recent years particularly in the light of
the agricultural commodity price spikes in 2007/2008 and again more recently in
2010/2011.
The demands placed on global land and food resources are already anticipated to rise in the
coming decades due to a growing global population, income growth and continuing
expansion in meat consumption. At the same time production of many crops is threatened
by the consequences of climate change (see section 4). The additional pressure applied by a
further sector, ie energy supply attempting to regain access to commodities and land to
meet its needs will inevitably drive up the price of agricultural raw materials. This means
that among others the cost of grains for different users such as the flour millers, brewers
and animal feed compounders is raised. In turn, they will pass on their higher costs to final
consumers of bread and other bakery products, drinks and other products and to livestock
producers which will, in turn, raise meat and milk prices too. The outcome is that retail
prices of food rise compared to where they would have been in the absence of the biofuels
policies. The extent of this depends on the importance of the raw material costs in final
product prices and the competitiveness of the food processing and retailing sectors.
The impact of increases in food prices on individual households depends very much on their
composition and income. As spelled out in more detail in section 5 (Box 4), the share of
income spent on food varies widely between low- and high-income households and is much
higher in the former. Because of this, increases in food prices will have a lower impact on
overall welfare and non-food expenditure in high-income groups. It may lead to reductions
in saving rates or in the consumption of other, non-food goods or may lead to a switch from
premium to discount food brands. In low-income groups, however, food price increases can
lead to reductions in food intake or to a switch towards cheaper and less varied diets, risking
malnutrition and/or undernourishment. Also, expenditure on schooling or health services
may be reduced. The uneven effect of biofuel induced commodity price increases on lowand high-income groups becomes more pronounced due to the fact that the former rely
more on staple foods which embody little processing. Therefore, sensitivity to any increase
7

Some of the prominent studies addressing indirect land use change from EU biofuel use are: Bowyer (2011),
Laborde (2011), Blanco Fonseca et al (2010) and Edwards et al (2010).
13

in commodity prices will be much greater in lower income families, whereas the diet of
higher income groups primarily consists of processed food that has passed through a longer
value-adding chain, so that commodity price increases have much smaller impact on retail
prices. In short, policies which raise agricultural commodity prices, such as biofuel
mandates, have a differentially larger impact on poor consumers.
3.2 Controversy over the fuel and food interaction
Central to the question of how the potential use of feedstocks for biofuels will impact on
food prices is the extent of the growth in biomass production and the additional demand on
land needed to enhance energy supply. Unfortunately, there is little agreement on the
precise scale of additional land use demands anticipated from biofuel production. The most
recent modelling results from IFPRI (Laborde, 2011) suggest global land use change
generated by the increase in EU demand for biofuels associated with the RED would be
between 1.73 and 1.87 Million hectares of additional cropland area 8. This is just one
illustration; other analyses have identified significantly larger and smaller estimates of the
extent of land use change (see Edwards et al, 2010, for a discussion of the most important
drivers of differences in results).
The consequences for agricultural commodity (and ultimately food) prices of this diversity
are discussed in detail within the next sections. The primary sources of the disparities in
estimates of the extent of biofuel impact on land and feedstock resources are summarised
below:
 Differences in assumptions regarding feedstock usage to produce fuels, ie different
crops yield dramatically different volumes of biofuel from one hectare of land. They
range from an estimated 5470 litre/hectare (l/ha) for ethanol from sugarcane in
Brazil and 4700 l/ha for biodiesel from Malaysian oil palm to 550 l/ha for biodiesel
from US soy and 950 l/ha for wheat ethanol (FAO, 2008; see also Annex 1).
 Variable consideration of land demands and competition impacts associated with
second-generation biofuels based on lingo-cellulosic material, combined with
different assumptions regarding their market penetration in the medium to long
term. It was originally assumed that the RED would drive expansion in these supplies
ahead of 2020. However, analysis of Member State plans anticipates only 8 per cent
of total biofuel demand by 2020 being delivered from such fuels, and moreover
there remain uncertainties regarding the impact of wood fuel resources in terms of
land and climate (Zanchi et al, 2010).
 Most of the growth in agricultural production in the twentieth century has come
from increases in yields rather than an increase in agricultural area. It expected that
this will remain true for the 21st Century. So a critical variable in explaining the
impact on prices of a rise in the demand for food crops will be yield assumptions.
These assumptions will differ between analyses. Yield growth has certainly slowed in
the last decade for many crops and regions, however this in turn is highly dependent
on some policy factors, especially agricultural research and development
8

The average of these estimates equates to 1.1 per cent of EU Utilised Agricultural Area or 1.7 per cent of EU
arable area (see Annex 1 for the sources of these total EU areas).

14









expenditures. The recent food price scares have stimulated policy change to
increase publicly funded agricultural R&D in many regions, including the EU (see
Foresight, 2011).
Whether biofuels feedstock suppliers are directly competing for the same crops or
varieties as food stuffs ie wheat varieties that are best suited for distilling, hence
conversion to bioethanol, differ from those recommended for milling and flour
production, or those often grown for animal feed (HGCA, 2010 and Nabim, 2012).
There is, however, a question as to whether this would in reality limit the impact on
wheat prices given that differing varieties of wheat are still digestible and are fit and
used for human consumption in developing countries, it is simply that they are bred
to meet different specialist needs in richer nations.
The extent to which arable land use will expand in order to deliver additional
demand generated by biofuels. Expansion in crop land could be ameliorated by
developments in the use of by-products, particularly for animal feed, and the extent
to which intensification of agricultural activities and yield increases for given crops
are viable (see discussion below).
The ability to deliver efficiency savings throughout the agricultural and food supply
chain ie reducing waste in production and consumption phases, hence freeing up
resources. The EU continues to debate whether to take action on food waste, which
is anticipated to rise (if unchecked) to 40 per cent by 2020 in Europe (European
Commission, 2011).
Differing assumptions regarding the consequences of any expansion in the area of
cropped land. A key question under debate is where would this expansion take
place, in particular to what extent would this make use of ‘degraded’ lands,
abandoned crop lands or what has become known as ‘idle’ land. The extent to which
these categories of land are ‘available’ for use and economically productive has been
questioned. For example, it has been reported that the Indian government has
allocated 400,000 hectares of wasteland for jatropha cultivation to deliver biodiesel.
However, these lands are often classified as Common Property Resources (CPR) and
utilised for food, fuel and materials. Analysis has indicated that CPR contribute up to
a quarter of poor household incomes with many of the poorest households
dependent upon them (Oxfam, 2008). Moreover, while production of some biofuel
feedstocks may be possible on degraded lands, inherent limitations mean such
activities cannot compete economically with feedstocks produced on highly
productive lands (FAO, 2008)

To illustrate the differences in the assumptions made in assessing the potential for biomass
resources for energy, Table 4 offers an overview adapted from a recent study. This analysis
by the UKERC (Slade et al, 2011) brings together key studies of global biomass potential for
energy use illustrating the extensive range anticipated in terms of viable energy outputs
from all forms of biomass. It also illustrates the different types of preconditions applied
which explains this variation, ie the low band of estimates would rely on a combination of
the use of residues, wastes and energy crops, moving from low to mid ranges implies a
dominant role for energy crops and significant assumptions concerning changes to the
agricultural system. The UKERC analysis, however, highlights three factors that are
important in all studies when determining the contribution of biomass to energy and
associated consequences for other sectors. These are: 1) the availability of land; 2) the
15

productivity of the biomass grown on the land, ie the yield based on production conditions
and conversion efficiency to biofuel products; and 3) the extent of competition from
alternate uses of land, biomass and biomass waste materials.
Table 4: Exploring the variation in estimated biomass potential for energy
Global biomass potential (EJ9)
High band – over 600EJ

Essential Pre-conditions




Upper mid- over 300 to 600 EJ






Lower mid – over 100 to 300 EJ





Low – up to 100 EJ






Crop yields outpace demand: >2.5Gha10 land for energy crops (includes
>1.3Gha good agricultural land)
High or very high input farming, limited, and landless, animal
production with dung recovery
Low population (<9bn)Vegetarian diet OR extensive deforestation /
conversion to managed forestry
All residues a (< 100EJ constrained use, not included in all studies)
Crop yields outpace demand: >1.5Gha land for energy crops (includes
>1Gha good agricultural land)
Low population OR vegetarian diet OR extensive deforestation /
conversion to managed forestry
All residues (< 100EJ constrained use, not included in all studies) ie
agricultural, forestry and wastes
Crop yields outpace demand: >1.5Gha land for energy crops (includes
>1Gha good agricultural land)
Low population OR vegetarian diet OR extensive deforestation /
conversion to managed forestry
All residues (< 100EJ constrained use, not included in all studies) ie
agricultural, forestry and wastes
Little or no land for energy crops (<0.4Gha total)
High meat diet OR low input agriculture
Limited expansion of cropland area AND high level of environmental
protection
Agricultural residues (<30EJ, not included in all studies)

Source: adapted from Slade et al for UKERC, 2011, diagram page 64
Note: summary of the literature produced by UKERC surrounding the question of the extent of global
biomass potential and the different pre-conditions that have to assumed in order to deliver different
levels of energy supply. This illustrates that any expansion in bioenergy use implies trade-offs and
that such shifts do not operate in a vacuum.

3.3

Determining the extent of biofuel land demands generated by the EU mandate –
exploring yield, intensification and use of by-products
Fundamentally, the question of increased biofuel production impact on food relates to how
easily current agricultural production can be expanded and in so doing, how efficiently the
outputs can be used to meet multiple demands for food, feed and fuel. Expanding the area
of agricultural land presents challenges in terms of environmental impacts; biofuels are in
most cases promoted to help deliver greenhouse gas savings from transport. Hence it is
necessary to minimise carbon dioxide emissions associated with land use change. Therefore,
9

Ej – Exajoule – 1018 Joules ie a measure of energy available

10

Gha – Global Hectares - The global hectare is a measurement of biocapacity of the entire earth. One global
hectare is a measurement of the average biocapacity of all hectare measurements and biologically
productive areas on the planet earth.
16

much of the debate has focused on the extent to which this can be minimised through the
more effective use of existing resources, ie through increasing yields on existing land,
intensifying production and putting by-products of biofuel production to use as animal
feeds. It is argued that if more effective use can be made of existing resources land use
change and food price consequences associated with expanded use of biofuels could be
limited.
3.3.1 Turning by-products into usable co-products
The extraction of the energy content of wheat and maize and the oil from rapeseed and
other oilseeds leaves an important protein by-product which is utilised in the animal feed
compounding industry. In turn, this new source of animal protein can displace some of the
need to grow protein crops domestically or import the protein from abroad. Different
models of the kind examined in Section 5 will have incorporated this animal protein effect
to different extents.
As of 2007, the EU livestock and poultry industries were consuming over 50 million tonnes
of oilseed meal per annum, approximately 65 per cent of this was soybean meal with the
remainder supplied primarily by rapeseed meal (25 per cent) and sunflower seed meal. Feed
use is anticipated to expand to approximately 55 Million tonnes by 2020 (EC DG AGRI,
2011b). Analysis to date has focused on the ability of co-products to replace imports of soy
based feeds, reducing overall demand for soy associated with animal feed. The main coproducts are from the domestic production of wheat based bioethanol in the form of dried
distillers grain with solubles (known as DDGS) and production of rapeseed in the from of
rapemeal and glycerine 11. DDGS from corn ethanol production is used extensively in animal
feed, particularly in the US, however, the use of DDGS from wheat has been less extensive in
Europe.
The Gallagher review (Renewable Fuels Agency, 2008) estimated that a greater supply and
use of protein rich biofuel by-products could potentially lead to fewer crops being grown
specifically for animal feed. This was considered particularly valuable, as protein crops have
lower yields compared to cereals, meaning they require a relatively large land area in
comparison. Analysis by CE Delft (Kampman, 2008) for the review estimated net land
requirements per tonne of biofuel could be reduced by 60 to 81 per cent by making full use
of this potential feed displacement effect. However, such displacement relies on a number
of preconditions. The price of such by-products must be lower than for traditional feeds in
order to promote a change from the status quo. This is because fundamental questions
remain regarding the substitutability of such by-products, given that the extent and nature
of the protein differs from traditional feeds. By-products from biodiesel in particular ie oil
seed meals and glycerine, also can be potentially recycled to provide energy for the
processing of biofuels, hence reducing the direct emissions from the production of the fuel.
This offers a compelling competing use for such by-products.

11

DDGS is a protein rich by-product of wheat ethanol production with approximately 0.3kg/kg produced.
Rapeseed meal is protein rich and produced at a level of approximately 0.5kg/kg with glycerine produced at
approximately 0.035kg/kg from rapeseed oil esterification.

17

Analysis for the pig (Jagger, 2008) and poultry (Acamovic et al, 2008) sectors in the UK has
shown limitations on the ability to utilise DDGS from wheat, rapemeal and glycerine.
Significant differences in terms of the nutrient content of DDGS have been noted in terms of
the extent, quality and digestibility of the amino acids present. In addition, DDGS from
wheat contains high levels of fibre, which may limit an animal’s ability to take on sufficient
quantities of feed. Only limited proportions of DDGS are, therefore, recommended within
the feed mix. In the case of rapemeal, high levels of glucosinolates can be present, these
compounds are known to have negative health consequences for animals consuming it at
high volumes; the same holds for residues of methanol and salts (sodium or potassium
chloride), likely to be present in biodiesel by-products. In addition, the metabolism of
glycerol requires the activation of a specific enzyme and above certain thresholds of intake
it is simply excreted rather than utilised by the animal.
3.3.2 Increasing Yield and the potential for agricultural intensification
Can we produce more crops on the same agricultural area, hence limiting the need to
expand the area of arable land to deliver biofuel demand? In the past 40-50 years increases
in production have largely been the result of higher yields from crops as a result of
developments in crop breeding, in plant protection, weed and disease control and in
mechanisation, rather than from expanded areas of agricultural land. For example, in the EU
the area of agricultural land has remained relatively stable. However, on average crop yields
have increased by around 29 per cent12. The Gallagher review estimated that high and low
yield improvement scenarios result in approximately ±10% influence on total land demand
for biofuels (Renewable Fuels Agency, 2008).
There is considered to be the potential to enhance crop yields, including within the range of
feedstocks used for biofuel production – see figure 2. However, this potential is highly
variable depending on location and determined by historic investment in the agricultural
potential, ability to support research and development into yield expansion, the willingness
of society and especially opinion influencers to accept new technologies in agricultural
production, and the acceptability of the environmental impacts of increasing agricultural
productivity. Analysis for the World Bank (2011) has combined assessments of potential for
yield increase with an estimate of a country or region’s ability to expand agricultural land.
This provides a typology of the likely impact of expanded demand for agricultural
commodities across the globe, given that EU biofuel demands will generate a footprint
globally. Four different types of global regions are identified based on their projected ability
to increase yield and/or increase the area of agricultural land. Given some uncertainties
such as the precise availability of ‘suitable’ land given other uses and demand, this highlights
some broad patterns, for example that there is little potential for expansion in production
either based on yield or agricultural area in countries such as China, Vietnam or Western
Europe. However, significant potential based on purely yield increase in countries such as
Kenya, Malawi, Ukraine, Central America etc is noted (see Box 2).
The stimulation of yield increases, rather than the expansion of agricultural land, in
response to the additional demands from biofuels should not, however, be assumed.
12

Based on IEEP et al (unpublished), Interim report for the DG Environment Study – Land as an Environmental
Resource, project code ENV.B.1/ETU/2011/0029.

18

Critically, the Gallagher Review (Renewable Fuels Agency, 2008) noted that advances in
yield are dependent on three drivers:
 Public investment in research and infrastructure;
 Supportive legislative and trade agreements; and
 Private investment supported by profitability of production – which depend in turn
on agricultural product and input prices.
The FAO (2008), and others, believe that biofuels could provide a mechanism for
reinvigorating investment in agriculture, in particular research and development. Analysis
has, however, shown that crop productivity is higher in high-income countries than those
with low or middle income. It is also telling that despite a significantly lower baseline in
terms of productivity between 1992 and 2002, yield increases in low and middle-income
countries were only marginally higher than in high-income countries. While biofuels could in
principle stimulate yield increase, in practice this will be contingent on appropriate
investments (UNEP, 2009).
Figure 2. Reviewing the potential for yield increases for four key biofuel feedstocks

Source: UNEP, 2009, p74

19

Box 2: Summarising the potential for agricultural expansion globally and providing a typology for
the nature of this expansion

Type 1: Little Land for Expansion, Low Yield Gap
This group includes Asian countries with high population density, such as China, Vietnam, Malaysia,
the Republic of Korea, and Japan, Western European countries, and some countries in the Middle
East and North Africa with limited land suitable for rainfed production, such as the Arab Republic of
Egypt and Jordan
Type 2: Suitable Land Available, Low Yield Gap
This group includes countries where land with reasonably well-defined property rights and where
infrastructure access is fairly abundant and technology advanced, mainly in Latin America
(Argentina, Uruguay, and central Brazil) and Eastern Europe
Type 3: Little Land Available, High Yield Gap
This group includes the majority of developing countries, including relatively densely populated
areas in highland Ethiopia, Kenya, Malawi, the Philippines, Ukraine, Cambodia, and Central
American countries (such as El Salvador) with limited land availability as well as Middle Eastern and
North African countries where water availability constrains the expansion of agricultural production.
Although there is little land available, large numbers of smallholders may be locked into poverty
because the area cultivated remains far below the yield potential.
Type 4: Suitable Land Available, High Yield Gap
This group includes sparsely populated countries—such as the Democratic Republic of Congo,
Mozambique, Sudan, Tanzania, and Zambia—with large tracts of land suitable for rainfed cultivation
(in areas of sufficient precipitation) but also a large portion of smallholders who only achieve a
fraction of potential productivity (figure 3.5). In some cases, such as Sudan, these areas are located
in areas with political tensions and dispute. Labor supply often constrains expansion by
smallholders, implying that not all potentially suitable land is used for crop production.
Source: World Bank, 2011, page xxxvi

3.3.3 In summary
The divisions within the ‘food-fuel debate’ lie partly in different interpretations of areas of
genuine uncertainty and partly in the interests and positions of the participants. Clearly
20

biofuel policies will add to other factors which have pushed up demand for certain
commodities and there will be subsequent impacts on land use, but this will occur against a
background of changes in market prices and national policies, giving rise to impacts which
will vary considerably between countries and social groups within them. Generic effects at a
global level are only one concern – local impacts are another.

21

4

THE ROLE OF VARIOUS KEY FACTORS IN DETERMINING AGRICULTURAL COMMODITY
PRICES AND THEIR VOLATILITY

The purpose of this section is to give an overview of some of the important factors that
influence agricultural commodity prices and commodity price volatility. This is to prepare
the ground for the analysis in section 5 which examines what we know about the influence
of policy driven EU demand for biofuels on global agricultural commodity prices. The
purpose is thus to enable the reader to understand the ‘normal’ dynamics of agricultural
commodity prices as a background to interpreting the analysis in section 5.
4.1 A little historical context
There is nothing new about using biomass to provide energy for transport. Indeed until one
century ago, practically all transport ‘fuel’, apart from steam trains and ships, was the forage
and grains fed to horses (or other draft animals). In many parts of the world this is still the
case. Huge areas of land were liberated for food production in developed countries during
the 20th century as general transport, as well as agricultural traction, switched from horses
to utilise the internal combustion engine and fossil fuels. This is one reason why real
agricultural commodity prices fell throughout the 20th century despite unprecedented
population and economic growth and diet change. It is still the case that many developing
countries use animals for transport and agricultural traction. Equally, traditional biomass
energy in the form of wood, charcoal and animal manure continues to play an important
role as a cooking and heating fuel in the developing world, making up around 90 per cent of
all bioenergy use globally (for instance WBGU, 2009).
It is the development in more recent years and decades in industrialised and newly
industrialised (most notably Brazil) countries that have started producing ‘modern’ forms of
biofuels and bioenergy to displace fossil fuels in the internal combustion engines of road
vehicles, as well as in electricity and heat generation, that represents a paradigm shift.
These novel (at least in the sense of their scale) uses of agricultural biomass to produce
energy constitute potentially a significant additional demand for agricultural commodities.
This outward shift in demand for agricultural commodities must be expected to have some
impact in raising agricultural commodity prices above where they would have been before
the new energy demand came on stream. The question is therefore not whether there is an
impact on agricultural commodity prices but how big it will be.
These questions lie at the core of the food versus fuel debate, introduced in the previous
section, that became highly publicised in response to the 2007-08 agricultural price spike. As
can be seen from Figure 3, this spike was not the end of the story. In the wake of the global
financial and economic crisis taking off in 2008 and due to the strong supply response to the
2007-08 prices there was a surge in production with above-average harvests in 2008.
Agricultural commodity prices dropped steeply from their 2007-08 highs. But the year 201011 saw a renewed price spike. The literature on agricultural commodity prices and their
determinants has soared in response to these recent experiences. Any discussion on the
role of the key agricultural market drivers is bound to refer to recent experiences, because
the subsequent analyses provide useful explanations assessing the relative importance of

22

different factors. One outcome of the compendium of analyses is that it is by now broadly
established that biofuel demand was one amongst many factors explaining these spikes 13.
Box 3: The importance of price volatility
It is the volatility of prices as much, or arguably even more, as the level of prices that causes
concern. The FAO et al (2011) work for the G20 focusing on agricultural price volatility bears witness
to this fact. The most vulnerable households to high prices are likely to find it even more challenging
to find ways of coping with the unpredictable nature of volatile prices. This is an aspect not
accounted for in the models reviewed in the next section that are either only able to predict price
changes in some target year, eg 2020, or include linear projections over time. In reality, agricultural
markets will hardly follow such smooth paths towards reaching the higher price level, but will rather
experience volatility that is difficult to predict. This unpredictability also makes it more difficult to
adopt mitigating measures to compensate the most vulnerable for higher agricultural and food
prices, for instance by means of emergency food aid. And, very importantly, the high prices create
an environment of uncertainty not favourable for stimulating investment in agricultural
development and hence increasing supply.

4.2 Key factors influencing global agricultural commodity prices
Understanding the dynamics of global agricultural commodity prices is complex as they
involve a number of interlinked processes at different temporal and spatial scales. Here we
simply provide an overview of some of the key factors to help understanding of the studies
which model the specific contribution of biofuels policy. The principal factors in the recent
commodity price spikes were:
 Population growth
 Income growth and associated changes in diets
 Weather conditions
 Changing climate
 Technological advances
 Crude oil and other energy prices
 Exchange rate movements
 Changes in the stock levels of agricultural commodities
 Trade policy
 Speculation and, more generally, financial market activity, and
 Biofuels policies
These will be considered first through their impacts on demand, then on supply, and then
we consider other influences and the importance of price volatility as well as price levels.
First we show two depictions of the dramatic price developments in the last decade.

13

A recent analysis conducted by Ecofys et al for the European Commission adds to this discussion. The
contribution of rising biofuel consumption in the EU to the 2007-08 cereal price increases is called ‘modest’,
while it is shown to be ‘significant in pushing up other food prices, notably prices of oilseeds and vegetable
oils’ due to the bigger share of biodiesel production and use (Ecofys et al, 2011, p143). At the same time, the
authors note that the increase in global biofuel production in combination with harvest shortfalls (two
factors that were found to be ‘mutually reinforcing’), does ‘explain a significant part’ of the 2007-08 price
increases (page iv).

23

Figure 3. Development of food and cereal price indices – nominal and real, 1995 - 2011
300
250
200
150
100
50
00

Food Price Index_nom

Cereals Price Index_nom

Food Price Index_real

Cereals Price Index_real

Source: http://www.fao.org/worldfoodsituation/wfs-home/foodpricesindex/en/. The real indices are
constructed by using the World Bank Manufactures Unit Value Index (MUV).

This FAO chart shows how the index of nominal international cereal and food prices had
plateaued in the late 1990s, but started showing upward movement in 2006 before
increasing more than two and a half-fold by the end of 2007. Following the 2009 slump they
once again more than doubled before settling back. The indices in real terms showed lower,
though almost as dramatic, changes.
Figure 4 below takes a longer perspective comparing the 2007-08 price spikes to the big
price spikes of the 20th century, showing the contrasting story between agricultural
commodity prices and energy (as well as metal) prices in the price spikes. In the 1950 price
spike energy prices were stable. In the mid-1970s, the surge in agricultural prices preceded
that in energy prices. The reverse was true in the recent spike.

24

Figure 4. Commodity Price Indices (Real, MUV‐deflated, 2000=100)

Source: Baffes and Haniotis, 2010, p27
Note: deflated using the World Bank Manufactures Unit Value Index (MUV)

4.3 Factors influencing demand
The principal drivers of growth in food demand are population, incomes, changing diets and
biofuels policies. Population growth leads to a sustained increase in demand for food and
agricultural commodities. Similarly, income growth, especially of those living below or close
to the poverty line, is expected to lead to absolute increases in the demand for food
commodities. Furthermore, income growth itself is generally associated with changes in
dietary patterns, particularly involving an increase in consumption of livestock products,
meat and dairy produce (see Schroeder et al, 2010; Delgado, 2003). Expansion of livestock
consumption, in turn, increases the demand for agricultural staples via increased animal
feed requirements of energy and protein. The responsiveness of the demand for food
commodities and any other goods to changes in disposable income is measured by the
income elasticity and this elasticity is considerably higher in developing countries compared
to high-income countries14. To these demographic and economic drivers of consumption
growth more recently has been added, especially in N America and Europe, the policy driven
shift in demand through biofuel policy providing strong incentives for consumption
regardless of market prices. Note however that although it has been biofuels policy which
has driven their recent expansion, if the trend towards high and rising oil prices persists this
could make biofuels competitive with fossil fuels and encourage their expansion even if
biofuels policies are phased out.
14

Hertel et al (2008) have compiled income elasticities for 113 world regions as part of the Global Trade
Analysis Project (GTAP). Looking at this dataset shows that for most OECD countries the income elasticities
for the product category ‘Grain-Crops’ is well below 0.1 whereas it reaches close to 0.7 in least-developed
countries (Table 14.5).
25

Looking at the past and anticipated importance of these factors, however, it should be
noted that population growth has been slowing in recent decades, from 1.8 per cent over
1970-1990 to 1.3 per cent per annum over 1990-2010 and it is expected to slow further,
with global population stabilising and even falling after the middle of this century (HM
Government, 2010, quoting UN statistics). Also, according to analysis by the European
Commission (EC DG AGRI, 2011a), growth in world demand has slowed down for most
commodities in the 1997-2008 period compared to the three preceding 12-year intervals
analysed. This pattern holds in particular for wheat, rice, total grains and total feed grains,
while there has been (continued) strong growth in palm oil demand of around 9 per cent
per annum over 1997-2008. It is further pointed out that both China and India ‘have a
history of effective self-sufficiency policies in the production of staple grains’ (HM
Government, 2010, p50). This limits the extent to which their markets are connected with
other world regions and can hence exert influence on global market prices.
Other factors are likely to persist: looking at future pressures from increased demand,
Abbott et al (2011) in analysing the 2010-11 price spikes note that the major drivers in 2011
were large and persistent demand shocks from biofuel policies as well as the demand
derived from Chinese soybean imports to build up domestic stocks. These demands are
expected to grow more slowly in the future, but will nevertheless persist to maintain US
ethanol production and Chinese soybean stock levels. According to Abbott et al, this
persistence is different from past demand surges over the last century, which were often
more short-lived. In terms of the impact of past EU biofuel demand, a recent analysis has
been conducted by Ecofys et al (2011) for the European Commission. They note that the
contribution of rising biofuel consumption in the EU on the 2007-08 cereal price increases
was ‘modest’, while it is shown to be ‘significant in pushing up other food prices, notably
prices of oilseeds and vegetable oils’ due to the bigger share of biodiesel production and use
(Ecofys et al, 2011, p143). At the same time, the authors note that the increase in global
biofuel production in combination with harvest shortfalls, two factors that were found to be
‘mutually reinforcing’ does ‘explain a significant part’ of the 2007-08 price increases (page
iv). The anticipated impact of future EU biofuels demand on agricultural markets is
discussed in section 5.
4.4 Factors influencing supply
The classic determinants of agricultural supply are costs of production, technology and then
uncontrollables such as climate, weather, pest and disease.
Although agriculture utilises a large part of the land area of most countries, and especially in
developing countries it employs a significant proportion of the labour force, land rents and
prices and wages are determined by, rather than determinants of, agricultural supply and
prices. However, increasingly as agriculture modernises it becomes more dependent on
variable inputs purchased from the rest of the economy, particularly energy and fertilisers
(which are in turn energy intensive in their production).

26

Petroleum and natural-gas-derived fertilisers15 make up an important share of production
costs (for some crops, eg maize, more than for others, eg soybeans, see for instance HM
Government, 2010, p36) as do other energy inputs for instance for machinery use and
transport. Van der Mensbrugghe et al (2011) provide evidence for the strong response of
many commodity prices to changes in energy prices. In the light of rising energy prices, this
factor is expected to remain crucial. The rising energy and fertiliser prices have therefore
raised agricultural costs, shifted up the agricultural supply curve with significant effect on
agricultural prices. It is particularly this fundamental rise in agricultural costs which has lead
the FAO and OECD to predict in each of their annual outlook publications since 2008 that
agricultural prices will not return in the foreseeable future to their low levels at the
beginning of this century (OECD/FAO, 2010). Likewise, the World Bank expects energy
prices to be the major contributor to post-2015 increases in food prices (cited in Agra
Europe, 2012).
More recently and as a result of increasing biofuel demand, a further link has emerged
between oil prices and agricultural output prices. Rising oil prices raise the economic
viability of biofuels compared to fossil fuels, all other things being equal despite the increase
in their own production costs. Increased biofuel production increases the demand for
biofuel feedstock, consequently raising their price until the competitiveness of biofuels with
fossil fuels equalises. Schmidhuber (2007) has put forward the theoretical argument that
biofuel markets establish this closer link between energy and agricultural markets, once oil
prices has crossed a certain threshold making biofuels competitive. Empirical analyses have
provided evidence for this link between the price levels for energy and different agricultural
commodities as well as their volatility and it seems that this evidence is growing and
becoming more compelling, based on literature collated by Hertel and Beckman (2011)16. At
the same time, Hertel and Beckman concede it is still a recent phenomenon and caveat the
existing evidence from econometric studies as ‘[suffering] from insufficient historical time
series’ (p9).
After production costs the next most important determinant of supply is the state of
technology deployed in any farming system. This in turn is multi-dimensional depending on
the biotechnology, biology, mechanical, chemical and management systems utilised. It
embraces the genetic potential of plants and animals, the plant protection and animal
health measures, the degree and type of mechanisation, and the sophistication and
information used in management. The size and structure of farms also interacts with
technology. These factors are highly variable across the world, and Europe, and are
themselves developing all the time.

15

EC DG AGRI (2008) provides time series of crude oil and fertiliser prices illustrating their co-movement and
parallel spikes over 2007-08 (p20).

16

Rapsomanikis and Hallam (2006) and Balcombe and Rapsomanikis (2008) study the sugar-oil-ethanol links
and find evidence for threshold cointegration between price pairs. Evidence for a link between corn prices
and oil prices is found by Tyner (2009) (as quoted by Hertel and Beckman, 2011). Du et al, 2009 (as quoted in
Hertel and Beckman, 2011) tests for the existence of a link between the volatility of crude oil prices and of
agricultural commodities (corn and wheat). No evidence is found for spill over effects from 1998-2006.
Estimations using data for the more recent years (October 2006 to January 2009), however, ‘indicate
significant volatility spill over from the crude oil market to the corn market’ (Hertel and Beckman, 2011, p7).
27

The simplest indicator of the net effect of technology is crop or animal yields per hectare or
per animal. Yields of agricultural commodities themselves are highly variable from one year
to the next depending on the weather conditions prevailing over the crop cycle. Indeed
harvest shortfalls in particular of wheat in Australia, a major wheat producer and exporter,
are mentioned as one of the more important reasons for the 2007-08 price spikes (Pfuderer
and del Castillo, 2008; HM Government, 2010). Yield developments are expected to be
influenced by a changing climate. However, the extent to which this will harm and benefit
different world regions and the impact on world crop balances are not easily understood or
projected (see for instance Msangi and Rosegrant, 2011, and Fischer, 2011).
The longer-term development of yields also depends on technological advances, realised
over the decades to come. Bruinsma (2011, p262) illustrates the slowing trend in yield
growth for major crops globally, which was 1.7 per cent per annum on average over 19612007 and is projected to be 0.8 per cent per annum for 2005/07-2050. 17 Similarly, European
Commission analysis for 2010-2020 suggests that average yield growth will not be able to
keep up with demand growth, even though the latter is expected to slow down over the
period studied, therefore expecting further upward pressure on prices, which in turn, so the
optimistic expectation, could speed up yield growth (EC DG AGRI, 2011b). However, there is
considerable evidence to show close relationships between agricultural crop yield growth
and expenditures on research and development in agriculture (see for instance Piesse and
Thirtle, 2010). Private sector R&D effort is driven by economic returns, so rising agricultural
prices might be expected to stimulate such activity. Public sector R&D is a political decision,
and high-level global commitments have been made since the agricultural price spikes to
significantly increase R&D on sustainable agriculture18.
4.5 Other factors influencing international agricultural commodity prices
The level of international stocks in agricultural commodities is a major determinant of the
extent to which markets can deal with, for instance weather related, supply shortfalls. In
other words, stock levels affect the availability of supplies and therefore can have a major
impact on prices especially in the short run. Stock levels can be affected by one-off events
like a particularly bad harvest but also by persistent demand growth that exceeds growth of
crop production. Another important factor are strategic decisions by countries to build up or
reduce stocks. A frequently voiced concern is that information on international stock levels
is poor and non-transparent, implying uncertainty about stock that may have repercussions
for price expectations. Indeed falling stock levels have been suggested by many as an
important influence for recent price spikes (FAO et al, 2011). Purchasers are naturally
alarmed if stock levels are seen to be low, and will tend to bid up prices to ensure they can
satisfy their market. Partly as a response to these concerns, the report prepared for the
2011 G20 meeting of agricultural minsters by several international organisations (FAO et al,
17

He caveats these numbers by clarifying that such aggregate figures may hide variations between and within
regions as well as between crops, but claims that the slowdown does reflect a pattern representative for
most major crops.

18

Exemplified by the European Commission in their proposals for the next 2014-2020 Multiannual Financial
Framework to more than double agricultural R&D (June 2011, available at:
http://ec.europa.eu/budget/biblio/documents/fin_fwk1420/fin_fwk1420_en.cfm).
28

2011) calls for an Agricultural Market Information System (AMIS) to improve information on
stocks.
Pfuderer and del Castillo (2008) point at a negative correlation between stocks-to-use ratios
and world prices. This is more closely analysed by the European Commission, revealing that
maize and wheat prices ‘are strongly and inversely linked to changes in stocks-to-use ratios’.
This elasticity of price to movements in stocks is furthermore increasing, ‘more than
doubling since the mid eighties’. This development is particularly pronounced for wheat:
Wheat prices historically changed by around 2 per cent in response to a 10 per cent change
in the stocks-to-use ratio. But since the mid-1980s, the elasticity has increased to 14 per
cent (EC DG AGRI, 2011a, p9). Soybean prices are much less responsive to stock changes
attributed to traditionally low world stocks and the fact that soy is grown both in northern
and southern hemisphere so that weather shocks can be balanced out. The responsiveness
of sugar prices to stock changes has been increasing while no link between rice prices and
stocks-to-use ratios is found (EC DG AGRI, 2011).
It is interesting to see how stocks have actually developed over time. HM Government
(2010) show that the time series of wheat stocks-to-use ratios over the last decades reveals
three low points that correspond to periods of spiking cereal prices in the early 1970s, mid1990s (more modest rise in cereal prices compared to the other two periods) and more
recently in 2007-08. On the other hand, Abbott et al (2011, following Headey and Fan, 2010)
argue that the explanatory power of stock changes is reduced when considering the
development of stocks-to-use ratios for the world excluding China, which has followed a
stock-reducing policy in recent times especially for grains. They show that the argument that
stocks continuously declined over the 2000s and hence running up to the 2007-08 price
spikes holds for wheat, but not for soybeans, rice or corn. In a similar vein, EC DG AGRI
(2011a) analysis suggests that ‘available data (averaged out over 12-year intervals) show
that for most commodities, actual stocks-to-use ratios have remained around long-term
averages and in some cases they have even increased over time (from 1997- 2008)’ (p10).
Finally, it was mentioned above that recently political decisions in China to build up soybean
stocks have triggered a surge in demand that contributed to the 2010-11 price spikes and is
projected to be persistent so as to maintain stock levels (Abbott et al, 2011).
Trade policies and the resultant degree of integration of global agricultural commodity
markets are important for price levels and volatility. Increased trade volumes, and a wider
geographic range of sources for instance, can help smooth out supply shortages in one
region by imports from another region. The prospect of exporting commodities sought on
international markets may act as a stimulant to agricultural development. However, the
recent periods of sharply increasing prices have witnessed reactions by governments in the
form of trade measures, such as export restrictions, in an attempt to stabilise domestic crop
supplies and moderate domestic food price inflation. Pfuderer and del Castillo, 2008 provide
a list of measures imposed in a range of countries. The role of such measures is potentially
powerful in certain markets; their impacts in 2007-08 were found to be most pronounced in
the wheat and rice market, and export bans are cited as having altogether triggered the
spikes in the latter market (HM Government, 2010). Martin and Anderson (2011) have
estimated that trade barriers aimed at insulating domestic markets have contributed 45 and
30 per cent to the increases in rice and wheat prices in 2006-08, respectively. The impacts of
29

export restrictions are likely to go beyond a short-term destabilising affect but may reduce
incentives for production, hence ‘muting the supply response and thereby helping to hold
international prices higher for longer’ (HM Government, 2010, p46). Their future relevance
is hardly predictable, as it largely depends on decisions taken by individual governments and
whether this triggers similar or even retaliatory measures by others.
The rising volume19 of agricultural commodities subject to speculation or more generally
financial market activity is frequently referred to as one of the culprits for higher and more
volatile prices (a summary of such activities explaining both their function and effects is
provided by van der Mensbrugghe et al, 2011, p196). However, empirical evidence for the
influence of speculation is largely inconclusive. As the Agricultural Outlook summarises,
‘[most] researchers agree that high levels of speculative activity in futures markets may
amplify price movements in the short term although there is no conclusive evidence of
longer term systemic effects on volatility’ (OECD/FAO, 2011, p16; Abbott et al, 2011, also
summarise inconclusive evidence). A potentially important channel is the way in which
speculative activity builds up expectations on future prices, which can increase inflationary
pressures or volatility. But it is also pointed out in the context of the 2007-08 price spikes
that the chain of causality has been the other way around in the sense that ‘speculative
flows into agricultural futures markets followed, rather than caused the price increases’, and
the same holds for price volatility (HM Government, 2010).
Finally agricultural markets can be disturbed by macroeconomic developments for example
in exchange rate movements. Many agricultural commodity prices are denoted and traded
in US Dollars. If the US Dollar depreciates against currencies of importing countries then
these commodities look cheaper and the quantity demanded increases, adding to the other
demand pressures. Of course if an importing country’s currency depreciates against other
major currencies faster than the Dollar then their imports will look more expensive. The US
Dollar did indeed depreciate at the time of the 2007-08 financial crisis (see Figure 5
depicting the movement of the US Dollar exchange rate against the Euro). The future
outlook is highly uncertain and depends on a range of macroeconomic conditions, many of
which are well outside the sphere of influence of the US Federal Reserve, such as the future
developments in China and in the Eurozone and hence the strength of the Euro. Both oil
prices and exchange rates tend to be volatile. Therefore, because of their influence on
agricultural commodity prices, they can be expected to contribute to price volatility in
agricultural markets with potentially severe important implications for food security.

19

More information on increasing volumes can be found in Torero, M ‘Understanding the causes of food price
volatility and mitigating its consequences’, presentation given at Brussels Development Briefings on Food
Price Volatility, 30 Nov 2011; available at: http://brusselsbriefings.net/past-briefings/no-25-food-pricevolatility/.

30

Figure 5. Euro/Dollar exchange rate

Source: http://www.oanda.com/currency/historical-rates/

4.6 In summary
Agricultural commodity prices are highly complex and influenced by a wide variety of factors
which shift demand, supply or both, or directly impact on prices. This multitude of effects is
neatly summarised in Figure 6 below. Biofuels policy is undoubtedly a significant factor, but
not the only one of many causes of shifts in agricultural commodity markets. Because food
is a fundamental human need, when markets exhibit volatility, especially in periods of
shortage, markets can overshoot with dramatic, but thankfully usually short-lived doubling
or trebling of prices in periods of weeks.
Figure 6. Factors contributing to food price inflation and volatility

Source: Committee on Climate Change, 2011, p41

31

5

THE ROLE OF EU BIOFUEL INCENTIVES ON AGRICULTURAL COMMODITIES PRICES TO
2020: A REVIEW OF SELECTED MODELLING STUDIES

The purpose of this section is to provide a review of the existing evidence base on the link
between policy driven EU demand for biofuels on the one hand and agricultural commodity
prices on the other. This is done through a review of selected modelling based studies. As
noted earlier we will not be looking at the impact of biofuel prices on retail food prices as
experienced at the national level. Essentially, hardly any models yield these results. Working
out separately the consequences of increased agricultural commodity prices for food prices
and the resulting welfare impacts for different groups of consumers is outside the scope of
what can be achieved here. Box 4 spells out some of the factors involved determining the
results of this pass-through and the associated welfare implications.
For the purpose of this report we have looked at a range of modelling studies that have
been conducted in recent years using economic modelling tools, both partial and general
equilibrium models. The studies reviewed include work based on models by international
organisations such as the OECD and FAO, models in the hands of universities including the
GTAP model (Global Trade Analysis Project, coordinated by Purdue University), an EU
agricultural sector model CAPRI (developed mainly by University of Bonn) as well as work by
the European Commission’s Joint Research Centres (JRC). A lot of work on the agricultural
market effects on biofuel use has been conducted in the USA looking at the implications of
the US Energy Independence and Security Act (EISA), which was adopted in 2007. Given
their regional focus not all of these studies are relevant in the present context and are
therefore not included in the summary table in Annex 220.
In this section we first consider the role of models in understanding the impacts on
agricultural markets of biofuels policy. Second, we set out our approach to reviewing the
models and explain the main dimensions compared. Third, we set out the results of the
review by considering each of the characteristics in turn. Fourth, and finally, we reflect on
what the studies taken together indicate about the significance of EU biofuel policy on
agricultural commodity prices.
5.1

The role of models and our approach to reviewing them

5.1.1 The role of models in understanding the agricultural market impacts of biofuel use
Because biofuel demand adds to an existing demand for agricultural commodities, many of
which are traded on global markets, understanding the impacts of biofuel policies requires a
methodology that can incorporate these interactions through international trade.
Sophisticated economic models are needed in order to take a range of important economywide feedback effects into account, akin to the analysis of the land use change impacts from
biofuel use (see for example Kretschmer, 2011). These include: impacts felt across
agricultural sectors because of related markets (expressed by ‘cross-price elasticities’ that
determine the extent to which an increase in the price of, for instance, wheat influences the
20

Examples of studies focusing exclusively on the impact of US biofuel policy and therefore not considered
include: Elobeid and Hart (2007), Hayes et al (2009), Rajagopal et al (2009), and Hertel and Beckman (2011).

32

demand for, for instance, corn); impacts on other economic sectors, most notably the
energy and, in particular, the oil sector in the case of biofuels; international market impacts
due to linkages established by trade flows; consumption response in the food and feed
sectors.
This last point warrants some additional explanation: Reduced demand for food is likely to
be the results of drops in real incomes due to the higher retail prices for food. This response
is both real and an artefact of models. The fact that one of the modelled responses to
increased biofuel demand is a reduction in food consumption has gained increasing
attention in the ILUC debate. Using the GTAP model to study the land use change and
associated emissions effects from US biofuel mandates, Hertel et al run an alternative
scenario holding food consumption constant, so as to derive a ‘food neutral ILUC factor’,
they find that emissions from ILUC ‘per MJ of increased annual production capacity’ increase
by 41 per cent (2010a, p230)21. On this topic, Marelli (in Fritsche et al, 2012, p49) reports
that substantial increases in crop production would occur when food consumption is not
reduced, in other words pointing at an ‘ILUC credit’ due to reduced food consumption. As is
stressed in several places in this report, the impacts in the real world will be much more
complex than suggested by modelling results: a, for example, two per cent decrease in food
consumption globally or on a country-level estimated by an economic model hides the fact
that different population groups will be impacted to different extents.
Many of the models employed for this purpose have their origins in (agricultural) trade
analysis and are either partial (PE) or general equilibrium models (the latter also denoted as
CGE, computable general equilibrium models). The role of models is to capture the
multitude of economic interactions of the kind summarised above and they do this by
adopting a range of simplifying assumptions, for instance about the behaviour of economic
actors such as producers and consumers. Most of the models used for analysing biofuel
policies aim to run alternative scenarios and analyse the resulting impacts. They are usually
not put forward as straight forecasting devices; a notable exception is the AGLINK-COSIMO
model that is employed for providing projections for the OECD-FAO Agricultural Outlook. In
order to make the models workable, economic sectors as well as countries are aggregated in
to clusters of sectors and regions. The precise aggregation depends on the question at hand:
for the analysis of biofuel policies, one can expect a more detailed disaggregation of the
agricultural sector as well as a singling out of important biofuel producing and consuming
countries. Making use of aggregation, together with typically linear dynamics over time,
makes it practically impossible to derive reliable predictions such as the impact of EU biofuel
use on wheat prices in country X at time T.
5.1.2 Key characteristics of interest
We have reviewed the models focussing on a selection of the key characteristics most
relevant to the objectives of this study:

21

The supporting online material disaggregates the differences in price effects between the two scenarios,
which accounts for up to 0.8 percentage points for global export prices (this refers to their figure for coarse
grains, changing from a 7.22 per cent to a 8.04 per cent increase in the scenario keeping food production
constant, see Table S4 of the supporting online materials).
33










Type of model, ie partial or general equilibrium models;
Geographic and sectoral scope of the model;
Whether and in which way crop land expansion is modelled;
Whether and in which way biofuel by-products are modelled;
Whether and in which way second-generation biofuels are modelled and what their
assumed contribution to biofuel targets is;
Which biofuel policy scenarios are modelled;
Modelled (absolute) biofuel use in 2020 in the EU and globally;
In which way agricultural price changes are reported.

The importance of some of these aspects in the food versus fuel debate has been
introduced in section 3. The following paragraphs address the remaining characteristics and
explain how all of them are treated in models, discussing whether these are significant
consequences for the results obtained from running the models concerned. The list of
characteristics can be found again in the summary review of all the models, presented in
Table 5 where it forms the column headings. Annex 2 provides further information about
the studies reviewed including more details on the modelling of biofuel technologies and
land use (change).
Partial or general equilibrium models
An advantage of partial equilibrium models is that they entail a detailed representation of a
specific economic sector, in the biofuel context this would be the agricultural sector,
including the detailed representation of agricultural policies, of a broad range of agricultural
processes and of different land uses. This detailed representation comes at the expense of a
lack of representation of the remaining economy - that is often lumped together into a
single entity. General equilibrium models, on the other hand, have the advantage of
accounting for all economic sectors and world regions, thereby fully accounting for the
economy-wide feedback effects mentioned before. Sectors as well as countries are grouped
into aggregates in order to make the computation manageable. The disadvantage of this
comprehensive representation of economies is typically a less detailed representation of
individual sectors than found in partial equilibrium models (see also Kretschmer and
Peterson, 2010, and Kretschmer, 2011).
The type of model used to examine a question will have implications for estimates of price
effects. Kretschmer and Peterson (2010) point at the higher, and potentially unrealistically
high, implicit supply elasticities in GE compared to PE models (as mentioned in Britz and
Hertel, 201122); higher elasticities imply more responsive supply in the face of a demand
shock and hence lower price increases. Also, an element of discussion in the ILUC debate is
that GE models tend to assume a more elastic food market, ie a more pronounced drop in
22

Britz and Hertel (2011) (whose analysis is not included in the review tables because of their focus on a
stylized scenario of a biodiesel-only quota) look at the relative performance of GTAP-only model runs and
results from a couple model system including GTAP and CAPRI, in other words compare the results from a
CGE model (GTAP) with a combined CGE-PE model system. They report an increase in EU oilseed prices of 48
per cent in the year 2015, while price changes for other crops are negligible due to scenario design.
According to Britz and Hertel, the driver of higher price effects is the fact that production constraints and
current policies in the EU are better represented through CAPRI; therefore the integrated system displays
‘less domestic supply response and more imports of oilseeds and oils’ than the CGE-only model run.
34

food consumption for a given increase in prices, see the discussion above. Gerber et al
(2008) expect price effects to be higher for partial than general equilibrium models. This is
in line with findings from a review of modelling studies by Timilsina et al (2010).
Out of the models reviewed, five are partial and five are general equilibrium models and
one study employs a spread-sheet based approach for calculating price effects.
Geographic and sectoral scope of the model
Models differ according to their geographical and sectoral detail. The corresponding column
in the table gives an idea about the number of regions and sectors the world is divided into
in the models considered. As mentioned before the partial equilibrium models reviewed
here tend to focus on the agricultural sectors, in which case the number of commodities is
mentioned in the table.
Modelling of crop land expansion
The recent debate on Indirect Land Use Change (ILUC) has highlighted the considerable
challenges and at the same time has done a good deal to advance the modelling of land use.
The way in which production factor ‘land’ is modelled as well as the treatment of byproducts (next point) are among the crucial assumptions in the present context, as has been
argued in more detail in section 3. This implies that important insights in terms of
agricultural price effects should be gained from the models that have been at the forefront
of the ILUC debate. Allowing for land expansion and intensification of land use (for instance
through increased fertiliser application but also through longer term yield response to
higher prices) potentially reduces upward pressure on prices as a result of surging crop
demand, a point that is inter alia made in Keeney and Hertel (2009). The way that different
models deal with the issue of land expansion and yield response is addressed extensively in
Edwards et al (2010; see also ‘Box 2. Overview of the source of uncertainties’ in Laborde,
2011, p24).
Table 5 below reports whether land model design is such that the area of cultivated land is
‘allowed’ to expand (or contract) in response to price levels or other market conditions. Five
of the eleven studies explicitly mention and account for land expansion. The remaining
studies do not provide any details, so that it cannot be determined if and in what way the
expansion of cultivation is modelled. Annex 2 provides some further details about the way
land is modelled.
Modelling of biofuel by-products
The importance of by-products when it comes to assessing the agricultural market response
to biofuels and hence price impacts has been discussed in section 3. Evidence from
modelling studies is provided by Taheripour et al (2010), who have found that in their
scenarios of the implications of simultaneous EU and US biofuel policies, neglecting byproducts may overstate price impacts by 30 to 40 per cent for some outcome variables.
In six of the eleven studies by-products are included in the modelling. One study explicitly
mentioned that by-products were not considered, while it is not clear from the remaining
ones.

35

Contribution from second-generation biofuels
The use of second-generation biofuels is often put forward as a way of reducing the
pressures on land use and agricultural markets (see section 3) relative to supplies from food
crops. However, EU Member States anticipate using only limited amounts of secondgeneration biofuels up to 2020 (see section 2) and therefore model scenarios that rely on a
substantive share of second-generation biofuels should be gauged with caution.
Out of the studies reviewed, five do and another five do not include second-generation
biofuels, with two studies not specifying any distinction between the two. Where secondgeneration biofuels are included, these are modelled in diverging ways; in some studies it is
assumed that the land take requirement of supplying these biofuels is nil, while others
assume proportionately lower feedstock and/or land use requirements than for firstgeneration biofuels. Annex 2 provides some further details on this.
Biofuel policy scenarios & Biofuel use in 2020
The models reviewed differ tremendously in the way biofuel (policy) scenarios are designed.
As far as this information is available, we make clear in Table 5 what the assumptions are
about biofuel use or its shares of the transport fuel market in the reference scenario and in
the alternative policy scenarios. Key questions addressed are: 1) do the scenarios consider
only EU biofuel use and policies or are global use and mandates included as well? 2) Where
biofuel mandates are modelled as percentage shares of total fuel use, what absolute
magnitudes, and therefore feedstock demands, are implied and how do these magnitudes
change from reference to policy scenario? 3) What do different scenarios assume about the
market penetration of second-generation biofuels?
Given the focus of this study on the effect of EU biofuel policy in 2020, there is a further key
question about whether the model’s assumptions about the total amount of biofuel use in
2020 and the increase between the reference use and the 2020 policy scenario use is in line
with the volumes that the NREAPs predict. The comparison between the 2010 and 2011
IFPRI studies (Al-Riffai et al, 2010 and Laborde, 2011) on this point and others shows the
importance of the amount of mandated biofuel use modelled in relation to the predicted
results on indirect land use change.
Only five of the eleven studies actually give figures for the assumptions made on absolute
biofuel quantities and these show a diverse picture of the scale of biofuel quantities
modelled and especially in terms of the differences in biofuel use between reference and
policy scenarios.
Reporting of price effects
Studies differ in the way they report price effects. A straightforward way of reporting the
biofuel induced price impacts adopted in many studies is via a so-called ‘comparative-static’
analysis where prices in, for example, a 2020 biofuel policy scenario are compared to prices
in a 2020 reference scenario. This is indeed the approach followed by the majority of the
studies reviewed. Two of the studies in Table 5, however, report price changes over time,
and these two different sets of results cannot be compared. One study (Davies, 2012) only
reports price changes in a scenario where biofuel support is removed compared to a

36

baseline of existing support, which cannot be compared to the other studies based on the
information provided.
5.2 Results of the review
Given the range of different modelling characteristics that are important drivers in shaping
the commodity price impacts that the models report as arising from greater biofuels
production, the following two sub sections summarise the results with regard to the
agricultural commodity price impacts of 1) EU biofuel policy (next section) and 2) global or
multiregional policies (subsequent section) 23. The results from all studies are summarised
and tabulated below (Table 5) and in Annex 2. Table 5 is divided into two parts, with the first
part reporting on studies that focus on the impacts on EU biofuel policies while studies in
the second part model global, or at least multiregional targets.
Looking at the compendium of studies is useful in order to understand the magnitude of
likely agricultural market impacts and how the models looking at these impacts have
developed. In order to trace this development, we summarise the studies in chronological
order. However, it should also be noted that insights from some of the earlier studies are
not generally such as to address the question of how far the RED leads to increases in
agricultural prices as, in the absence of the NREAPs, there was still considerable uncertainty
at the time about the quantity of biofuels that would actually be used in the EU by 2020.
The compendium of studies indicates fairly wide ranges of results. Nonetheless, it is clear at
the same time that the impacts of EU biofuel policies are more pronounced on the markets
for biodiesel compared to ethanol feedstocks, while global impacts produce a more
balanced or even reversed picture. Section 5.3 provides a summary of the impacts of EU
biofuel policy on the prices of different feedstocks.

23

Many of the earlier studies in Table 5 (and Annex 2) are also reviewed in Blanco and Fonseca (2010), to
which those interested are referred to for further information. Further earlier reviews are provided by
Gerber et al (2008) and Timilsina and Shrestha (2010). Gerber et al (2008) have conducted a review of
biofuel induced food price effects as predicted by partial and general equilibrium models. The listed results
show a highly varying picture difficult to interpret which is mainly due to the differences in assumptions,
scenario and model designs and different ways in which results are reported.
37

Geographical
and sectoral
scope

Biofuel policy scenarios

Absolute biofuel
use in 2020 policy
/
scenario
increase EU (in
Mtoe)
Absolute biofuel
use 2020 policy
/
scenario
increase world (in
Mtoe)line
with
In
2020
NREAP
demand

Model
type

Contribution from
nd
2 -generation
biofuels?

Study and
model

land
Crop
expansion
possible?
Byproducts
included?

Table 5: Summary of key model characteristics and results

Studies that focus on the effects of EU biofuel policy (differ in the extent to which global biofuel policies and use are included in the reference)
EC DG AGRI PE
Global
?
Yes
Yes
EU biofuel policy only
34.6
/ ?
>/<
(2007),
EU+Turkey,
Baseline with 6.9% biofuel share in 2020
10.8
ESIM
USA & RoW,
Policy with 10% mandate for 2020 (2nd
ag sectors
generation contributing 30%)

Banse et al
(2008),
LEITAP

Edwards et
al (2008)

Blanco
Fonseca et
al (2010),

CGE

Global
32 regions
23 sectors

Yes

No

No

Baseline w/o blending obligation
Policy with 10% mandate for 2020
EU 10% 1st generation mandate for 2020

?

?

?

?

?

?

No
model
Ad hoc
calcs
PE

EU27

-

-

No

Global
(see
Annex 2 for
details on all
models)

?

Yes

Yes

EU policy focus but global use accounted for
(existing and announced policies modelled in
all scenarios)
Counterfactual no policy or support
Baseline 10% mandate in 2020 (2nd
generation, modelled as non land using,
st
contribute 30%, implying a 7% 1 -generation
share due to double counting)

~33.0 /
28.1
(Increase
2008 to
2020
baseline:
22.4)

~124.3
23

/

≈/≈

CGE

Global
11 regions
43 sectors

Yes

Yes

No

EU policy focus but global use accounted for
Baseline with constant 2008 blending of 3.3%
Policy with 8.4% 1st-generation mandate

27.2
15.5

111.2
/
7.4 (=production)

=/≈

AGLINKCOSIMO,
ESIM,
CAPRI

Laborde
(2011),
MIRAGE-

EU policy focus but global use accounted for
(volumes not specified)

/

Resulting per cent impact on
world commodity prices, 2020 policy compared
to 2020 reference,
unless otherwise stated

EU prices (apparently) over 2006-2020:
Cereals
3 to 6%
Sunflower seeds 15%
Rapeseed
8 to 10%
nd
‘significantly higher’ effects without 2 generation biofuels
Note these price rises are against an underlying
assumed decrease in real prices
Oilseeds
8.5%
Sugar
2.0%
Cereals
6.0%
Context: declining real ag prices
Compared to no biofuel use, world
cereal price
4%
vegetable oils 24%
AGLINK-COSIMO:
veg oil ~15%, small for ethanol feedstocks
ESIM:
Wheat 8, maize 22, sugar 21%
Soybeans 0.5, rapeseeds 10, sunflower seeds
11%
Veg oils palm 1, soy 17, rape 35, sunflower 36%
CAPRI:
Wheat 12.5, maize 8.0, sugar beet 2.0%
Oilseeds 19.5, Vegetable oil 27.1%
Selected price changes for policy scenario:
Wheat
1.01%
Maize
0.74%

Biofuel policy scenarios

Absolute biofuel
use in 2020 policy
/
scenario
increase EU (in
Mtoe)
Absolute biofuel
use 2020 policy
/
scenario
increase world (in
Mtoe)line
with
In
2020
NREAP
demand

Geographical
and sectoral
scope

Contribution from
nd
2 -generation
biofuels?

Model
type

land
Crop
expansion
possible?
Byproducts
included?

Study and
model

Biof

Davies
(2012)
AGLINKCOSIMO

PE

Global
All
major
commodities

?

?

?

EU policy but global use accounted for (see
below for combined EU+US scenario)
Existing biofuel support Baseline including
biofuel support policies (2011 Agricultural
Outlook projection)
Removing EU support by abolishing import
tariffs and tax incentives from 2011 onwards
Studies that analyse the impacts of global (or at least multi-regional) biofuel mandates
Rosegrant
PE
Global
?
?
Yes
Global policies
et al (2006),
115 regions,
1. Aggressive growth and baseline yields
IMPACT
30
2. Cellulosic biofuels as of 2015
commodities
3. Aggressive growth + cellulosic + increased
yield growth
Use of 2nd-generation modelled by reducing
demand for 1st-generation biofuel feedstock
OECD
PE
Global
?
Yes
Yes
Global policies
(2008),
All
major
No biofuel policy support
AGLINKcommodities
Existing biofuel support – world
COSIMO
Existing plus new policies incl 50% met by 2nd
generation
nd
2 -generation: on avg 50% of biomass from
cropland, rest from crop residues (see note
15 in OECD, 2008, p91)

39

?

?

?

?

?

?

25.4
19.7
(2013-17
avg)

/

96.0
38.7
(2013-17
avg)

/

</≈

Resulting per cent impact on
world commodity prices, 2020 policy compared
to 2020 reference,
unless otherwise stated

Sugarcane/beet
0.88%
Rapeseed 11.34%
Rape oil 9.20%
Palm oil 4.5%
Avg % change in 2011-2020 prices:
Removing EU support vs existing biofuel support:
EU prices: Wheat -7, veg oil -12, oilseeds -4%
Global prices: Wheat -3, coarse grains -2, oilseeds
-2, veg oil -5%

For scenarios 1, 2, 3:
Cassava 135 89 54%
Maize
41 29 23%
Oilseeds
76 45 43%
Sugar beet 25 14 10%
Sugar cane 66 49 43%
Wheat
30 21 16%
Avg % change in 2013-2017 world prices
compared:
No biofuel support vs existing support:
Oilseed -3, Wheat -5, Coarse grain -7, Veg oil 16%
Existing plus new policies vs no support:
Oilseed 7, Wheat 8, Coarse grain 13, Veg oil 35%

Geographical
and sectoral
scope

land
Crop
expansion
possible?
Byproducts
included?

Fischer et al
(2009a, b),
IIASA
model
system

AGE

Global
34 regions
10 sectors (9
ag, 1 rest)

Yes

Taheripour
et al (2010),
GTAP-BIO

CGE

Global

Yes

Yes

No

Timilsina et
al (2010)

CGE

Global
25 regions
26 sectors

Yes

?

No

Depends on scenario

Yes

Absolute biofuel
use in 2020 policy
/
scenario
increase EU (in
Mtoe)
Absolute biofuel
use 2020 policy
/
scenario
increase world (in
Mtoe)line
with
In
2020
NREAP
demand

Model
type

Biofuel policy scenarios

Global policies
Reference biofuel use constant after 2008
Policy scenarios following World Energy
Outlook, WEO, announced 2020 targets met
Ambitious biofuels target scenarios, TAR,
mandatory, voluntary and indicative targets
are met
2nd-generation from ‘non-food’ land (see
Fischer et al, 2009b, p168-76 for details)
Global (focus EU and US policies, Brazil incl)
US 15 billion gallon and EU 6.25% mandate in
2015
Separate scenarios with and without byproducts

?

Global policies
Baseline with already 5.4% biofuel share
Announced global targets scenario (9% share)
Double global targets scenario (announced
targets doubled while the timing of their
implementation is unchanged)

?

Contribution from
nd
2 -generation
biofuels?

Study and
model

Resulting per cent impact on
world commodity prices, 2020 policy compared
to 2020 reference,
unless otherwise stated

~25.9
~20.4

</≈

WEO scenario no 2nd-generation: cereals ~ 12,
other crops ~9%
TAR scenario: cereals ~ 35, other crops ~27%
TAR scenario rapid 2nd-generation: cereals ~ 15,
other crops ~19%

?

-

?

?

EU prices changes over 2006-2015, w/o and w/
by-products:
Cereal grains 11.0 and 5.6%
Other grains 11.9 and 8.7%
Oilseeds 31.6 and 26.9%
Sugarcane 10.7 and 7.8%
Prices changes announced, double targets:
Sugarcane/-beet 9.2, 11.6%
Corn 1.1, 3.7%
Oilseeds 1.5, 3.1%
Wheat 1.1 2.4%
Other cereal grains 1.0, 2.3%
Paddy rice 0.8, 1.6%
Vegetables, fruit 0.7, 1.5%
Livestock 0.4, 1.1%

/

(WEO
policy
scenario)

94 / ~69.6
(WEO
policy
scenario)

Context of significant crop price increases over
the baseline

Source: Own compilation. Notes: Explanation of abbreviations: PE = partial equilibrium model; CGE = computable general equilibrium model; AGE = applied
general equilibrium; ? = not specified. Column ‘In line with NREAP 2020 demand’: reports 1) whether biofuel use in 2020 is in line with NREAP projections for
2020, ie 27.2 Mtoe first-generation, 29.6 Mtoe all biofuels (first symbols) and 2) whether the increase in EU biofuel use is on line with the increase from the
‘reference’, taken to be biofuel use in 2008 prior to the adoption of the RED, to the NREAP 2020 biofuel use, ie roughly 19.5 Mtoe (Bowyer, 2011, based on
NREAP and EurObserv’ER data).
40

5.2.1 Findings from studies focusing on the effects of EU biofuel policy
The European Commission analysed potential price increases as part of the impact
assessment accompanying the climate-energy package of 200824. The impact assessment
acknowledged potential negative impacts but very much stressed the fact that increased
prices would benefit food producers and could revitalise rural areas globally. It is claimed
that ‘[the] market will have a full opportunity to adapt to the EU's target. The food security
impacts of the EU policy, both positive or negative, are therefore likely to be relatively small’
(p131). In concrete terms, the analysis suggests that ‘the commodity price impact of the
EU's own 10% biofuel target will be on a rather small scale, compared to the impacts of
other policies at the global level. Implementation of the policy will cause cereal prices to rise
by 3% to 6% compared to the 2006 level. The price of sunflower seed will increase by 15%
and of rape seed by 8% to 10%, while prices of animal feed will fall due to the increased
availability of biofuel co-products’ (pp130-131). This quote refers to analysis by EC DG AGRI
(2007), see Annex 2. The percentage changes reported refer to price increases over the
projection period 2006-2020 accounting for biofuel policy in place; they do not refer to a
comparison in the price level between baseline and policy scenario in the end-of-projection
year (eg 2020). It should be noted in this context that more recent analysis by the European
Commission recognises future EU biofuel use as increasingly driving future increases in EU
cereal and oilseed use as well as area expansion for sugar beet cultivation (EC DG AGRI,
2011b).
An academic paper by Banse et al (2008) using a general equilibrium model has found price
impacts that are rather at the lower end relative to the compendium of studies and results.
The low price effects are surprising given that their model does not account for by-products
accruing from biofuel production and replacing animal feed. One explanation for the
relatively low price increases put forward by the authors is the fact that their modelling only
includes EU biofuel policies instead of accounting for global mandates, which is an
important aspect in the scenario design. Furthermore, the incorporation of the land supply
curve allows for global land expansion in response to increased demand for crops and
therefore land is a less limiting production factor. It should be noted that unlike other
studies, it does not become clear from the paper what the absolute increase in biofuel use is
between the policy and reference scenarios and the fact that only EU biofuel mandates are
modelled is in line with the other studies reviewed in this section.
Analysis by the European Commission’s Joint Research Centre (JRC), Edwards et al (2008),
does not rely on modelling but instead uses a more ad-hoc approach to calculating price
effects. The study confirms that higher price increases can be expected for oilseeds and
vegetable crops, a conclusion drawn by all of the studies focusing on the impacts of EU
policy reviewed here. Given the stronger reliance of biodiesel use in the EU, this finding is
not surprising. (Also OECD, 2008, modelling simultaneous EU and US policies, project the
most pronounced increase for vegetable oil prices.)
24

Commission Staff Working Document SEC(2008) 85, VOL. II of 27.2.2008 ‘Annex of the Impact Assessment:
Document accompanying the Package of Implementation measures for the EU's objectives on climate
change and renewable energy for 2020’, http://ec.europa.eu/energy/renewables/doc/sec_2008_852_ia_annex.pdf.

Blanco Fonseca et al (2010) present JRC analysis based on three partial equilibrium models,
AGLINK-COSIMO, ESIM and CAPRI. The scenario results largely follow the predictions in
terms of biofuel use as derived from the NREAPs, given this study was prepared for the
European Commission to analyse and prepare the ground for potential legislation on ILUC. It
should be noted, however, that the increase in biofuel demand from counterfactual to
baseline scenario is considerably higher than the increase from 2008 biofuel consumption to
2020 NREAP demand for all biofuels, including second-generation. On the other hand, and
arguably more important for explaining differences in global price effects, the increase in
global biofuel consumption is considerably lower than, for instance, in Fischer et al (2009), a
study that explicitly addressed global biofuel markets and policies, as discussed in the next
section. The reason for the lower global consumption in Blanco Fonseca et al (2010)
presumably is precisely because the scenario design focuses on the EU and (apart from EU
policy), global biofuel policies are kept constant in the different scenarios. Despite this lower
increase in supply in Blanco Fonseca et al (2010) compared to Fischer et al (2009), price
increases in the former are higher. The extent to which the models in Blanco Fonseca et al
(2010) allow for crop land expansion is not entirely specified, but it is mentioned that for
instance AGLINK-COSIMO models land use change for selected countries only. Possibly, this
may result in a less elastic cropland area response, explaining the higher price effects. Again,
given the higher proportion of biodiesel in meeting EU biofuel demand, prices for biodiesel
feedstocks, in particular vegetable oil increase more substantially than prices for ethanol
feedstocks.
Further analysis based on the AGLINK-COSIMO model is provided by Davies (2012), an
economist at the UK Department for Environment, Food and Rural Affairs (Defra). The
results from this study are not readily comparable to other studies because Davies
compares scenarios conversely by only reporting price changes in a scenario where biofuel
support is removed compared to a baseline of existing support. In qualitative terms, his
results are in line with other studies and show the highest impact on vegetable oil prices.
This holds for both changes in European and global prices, both of which are reported 25.
Finally, we have obtained data on price effects from the author of the often quoted ‘IFPRI
study’, Laborde (2011), the major study used by the European Commission in the
preparation of its impact assessment on ILUC and which has been improved in the course of
this process (the preceding study is by Al-Riffai et al, 2010). Using a general equilibrium
model, Laborde finds again higher impacts on the markets for biodiesel compared to
ethanol feedstocks. That said, as in Blanco Fonseca et al (2010), increased global biofuel use
is accounted for but is kept constant across scenarios. In terms of the absolute magnitudes
of global biofuel use in 2020, Laborde (2011) lies between the studies of Fischer et al (2009)
and Blanco Fonseca et al (2010). Comparing the absolute increase in biofuel use in the EU
from the 2020 reference situation to the 2020 policy situation, which potentially is of crucial
importance for the extent of price increases, this increase is almost twice as high in Blanco
Fonseca et al (2010) compared to Laborde (2011), ie 28.1 Mtoe compared to 15.5 Mtoe.
This is because in the counterfactual scenario of Blanco Fonseca et al, biofuel use in the EU
decreases from 2008 to 2020 because the scenario formulation assumes that biofuel
25

Annex 2 adds results from a scenario that removes US policy support.
42

support measures are abolished altogether, whereas Laborde (2011) fixes 2008 biofuel
consumption unchanged over the projection period in his reference scenario (this also
implies that with regard to the increase of biofuel use from 2008 to the 2020 policy
scenario, the two models diverge much less). This, together with the fact that the models in
Blanco Fonseca et al are partial and not general equilibrium models, and an assumingly less
sophisticated way of modelling land use change could explain that price increases in Blanco
Fonseca et al (2010) tend to be higher than in Laborde (2011).
Table 6 shows differentiated EU and world average price impacts of enhanced biofuel use
for selected feedstocks. These are percentage changes between the price levels in 2020 in a
policy scenario compared to the price levels in 2020 in the reference scenario. In most
cases, EU price impacts are larger in response to EU biofuel policy, except for
sugarcane/beet and palm oil, both majorly or entirely imported feedstocks. Comparing the
first and second sets of columns shows the scale of biofuel demand in influencing
commodity prices. Two scenarios are illustrated. In the first, total EU biofuel demand in
2020 is derived from the estimates put forward by Member States in their NREAPs. This
results in an 8.4 per cent share of biofuels in the total transport fuel market. The second
scenario assumes higher use of advanced biofuels and renewable electricity and a
correspondingly lower (5.7 per cent) share of conventional biofuels in the EU transport fuel
market. It can be seen that the price impacts significantly increase between the two (note
that the 2010 IFPRI study modelled a central scenario with a 5.7 per cent blending share).
Table 6: Commodity price effects from Laborde (2011) for selected sectors and world regions
(comparison between 2020 reference and 2020 biofuel policy scenario)
8.4% EU biofuel 5.7% EU biofuel
share in 2020
share in 2020
EU27
Wheat
1.48%
Maize
1.09%
Sugarcane/beet 0.01%
Soybeans
2.64%
Sunflower
7.06%
Rapeseed
14.08%
Palm oil
4.43%
Rape oil
16.40%
Soy oil
9.79%

World
1.01%
0.74%
0.88%
2.48%
4.82%
11.34%
4.53%
9.20%
7.30%

EU27
0.73%
0.52%
0.01%
0.94%
2.56%
5.50%
1.53%
6.52%
3.71%

World
0.48%
0.33%
0.41%
0.88%
1.77%
4.27%
1.55%
3.56%
2.69%

Acknowledgement: We thank David Laborde for sharing the data underlying this table, which is not
reported in Laborde (2011). Please note that neither David Laborde nor IFPRI bear responsibility for
the accuracy of the way in which the data are presented in this report.
Note: All results are for the business as usual version of the scenarios (as compared to a set of trade
liberalisation scenarios also modelled). Price changes reported are for consumer prices.

Table 7 is meant to put the ‘comparative-static’ changes between the policy and the
reference scenario in 2020 (Table 6) into perspective with the price changes observed over
the projection period 2008 to 2020 in the policy scenario. This gives a better idea about the
importance of biofuel induced price changes. We focus here on the 8.4 per cent ‘NREAP’
43

scenario only. The highest price changes over time are observed for the biodiesel feedstocks
(with the exception of palm oil). Looking at world prices, biofuel policy contributes around
40 per cent to the observed increase in rapeseed prices of close to 30 per cent, for example.
The contribution is more pronounced when looking at EU27 prices, rising to 46 per cent. The
contribution of biofuels to the (limited) price increases in ethanol crops is small. The threedigit contribution to sugarcane/beet price increases can be disregarded given the extremely
small increases over 2008 to 2020 barely exceeding zero per cent, which inflates the
numbers for sugarcane/beet in the last two columns of Table 7.
Table 7: Price changes in the 8.4% policy scenario of Laborde (2011) and the contribution of
biofuels to these price changes over time

Wheat
Maize
Sugarcane/beet
Soybeans
Sunflower
Rapeseed
Palm oil
Rape oil
Soy oil

2008-2020
change
policy
scenario
EU27
World

Contribution
of
biofuels to 20082020 changes
EU27
World

8.64%
5.31%
0.01%
11.94%
15.22%
30.48%
7.57%
34.15%
31.03%

17.12%
20.54%
100.00%
22.11%
46.39%
46.19%
58.49%
48.02%
31.55%

11.60%
5.60%
0.14%
11.35%
11.69%
28.62%
7.91%
25.77%
25.71%

8.71%
13.22%
612.92%
21.84%
41.25%
39.62%
57.30%
35.71%
28.39%

Acknowledgement as for Table 6. Note: The contribution is calculated by dividing the comparativestatic price change in 2020 for the 8.4 per cent policy scenario (from Table 6) by the change over time
as in the first two columns of the present table.

5.2.2 Findings from studies analysing global biofuel mandates
Even though the scope of this study is on the impacts of EU biofuel policies on agricultural
commodity prices, we deem it relevant to consider examples of studies that model multiregional or even global mandates. While this does not allow singling out the EU policy
impact, it is in some sense a more complete scenario design, given EU policies in the real
world do not take place in isolation. In this context, we highlight the study by Hertel et al
(2010b), analysing the effects of simultaneous EU and US biofuel expansion over 2006 to
2015 and comparing the effects on agricultural outputs and land use (but not on prices) to a
situation where only one region’s policy is modelled. The authors note the large impacts on
global production and land use, especially because of expanding biodiesel use in the EU and
less so due to ethanol expansion in the US. Importantly, they conclude that the combined
effects are greater than the sum of the individual policies’ effects, highlighting the
pronounced land use change to arable cropping taking place in different regions.
Rosegrant et al (2006) is an example of an early study modelling global biofuel targets. The
price effects found are particularly high. It should be noted, however, that this is compared

44

to a baseline with 1997 as its base year and biofuel use is frozen at base year levels, when
they were relatively small.
OECD (2008) looks at the joint impacts of the US and EU biofuel policies, based on which the
authors conclude that there could be substantial agricultural market impacts. The most
pronounced increases are found for vegetable oil prices.
The study by Fischer et al (2009a) investigates two sets of scenarios, both including
projections for biofuel use and policies: 1) a set of scenarios based on World Energy Outlook
2008 (WEO) projections that assume that both China and the EU meet their targets for the
year 2020 with a few years’ delay while the US will have only attained 40 per cent of the
targets laid out in the Energy Independence and Security Act by 203026; 2) a set of biofuel
target (TAR) scenarios that assume that all mandatory, voluntary and indicative targets for
biofuel use in major developed and developing countries are met. Global biofuel
consumption is about twice as high in the TAR scenarios, 189 Mtoe compared to 94 Mtoe in
2020 and thus much higher compared to modelling studies that only include the EU target.
While the price effects found in the ambitious TAR scenarios are substantial, including
increases in the cereal price index in 2020 by 35 per cent compared to the reference
scenario, this change drops to 12 per cent in the less ambitious WEO scenario (with no
availability of second-generation fuels). It is not clear how much oilseed or vegetable oil
prices increase as these are not singled out but are included, so we assume, in the ‘Other
crops’ aggregate, whose price increases by 9 per cent in the WEO scenario. Fischer et al
(2009a) provide interesting additional information about the origin of the increased cereal
supply to the biofuel sector. On average across the scenarios, they conclude that 66 per cent
is derived from increased production, 24 per cent from reduced feed use (with the more
significant reductions taking place in developed countries) and 10 per cent from reduced
food use (with 75 per cent of the absolute cereal food consumption occurring in developing
countries).
The results from Taheripour et al (2010) cannot be easily compared to other studies
because the authors report changes over time (2006-2015) rather than comparisons
between baseline and policy scenario at the end of the projection period. Their study
modelling the simultaneous impacts of EU and US biofuel policy is, however, useful in
showing the importance of taking by-products into account in the sense that price effects
are, partly considerably, smaller in the scenario including by-products.
Results from Timilsina et al (2010) are summarised as follows. Their baseline includes a
global biofuel share of 5.4 per cent, which increases to 9 per cent in the ‘announced targets’
scenario. In an alternative, more ambitious policy scenario, currently announced global
targets are doubled. Absolute amounts associated with these shares are not specified,
which makes a comparison to other studies more difficult. Nevertheless, the shares
modelled are considerable while the price effects found are rather modest given the
scenarios modelled. Also, while price effects double for some crops in the ‘doubled targets’
26

The WEO does not spell out 2020 biofuel use in the EU but 2030 use, when the EU is assumed to have met a
10% biofuels target, is predicted to be 25.9 Mtoe. This number falls short of the 29.6 Mtoe biofuel use (out
of which 27.3 Mtoe first-generation biofuels, see Bowyer, 2011) predicted by the NREAPs for 2020.
45

scenario, this happens only in those cases where the effects were small in the first place.
The only more pronounced price increase is found for sugarcane/-beet, with a 9.2 per cent
increase, rising to 11.6 per cent in the ‘doubled targets’ scenario. One explanation could be
that the general equilibrium model used allows for flexible adjustment to the increased
demand for biofuel feedstocks across the economies, which makes the comparison with
Fischer et al (2009a), employing a PE model, difficult. However, comparing the results to
Laborde (2011), another general equilibrium model, it is rather striking that the increases
from the ‘doubled targets’ scenario in Timilsina et al (2010) tend to be hardly higher. Having
said this, sugar prices do increase more substantially than in Laborde (2011) but at the same
time increases in oilseeds are considerably reduced.
Box 4: From global to local – from commodity to food prices
The ‘food versus fuel’ debate often revolves around the idea that the impact of increased demand
for biofuel feedstocks directly affects consumer food prices. The link between increasing demand for
biofuels and the final price of food to the consumer is, however, not the focus of the modelling
studies examined here. Instead, the modelling studies reviewed for this paper analyse the impact of
biofuel demand on price levels for agricultural commodities. The pass-through from agricultural
commodity prices to consumer food prices in practice is much more complex and varied between
markets and over time and any estimates thereon will be crude as they depend on a range of
factors, including:
 Different use of staples in developed versus developing countries: A good example is corn, which
is used for direct human consumption in a range of developing countries but is primarily a feed
grain in developed countries;
 The scale of contribution of agricultural raw material costs to food production costs and hence
prices. In many countries there is now a large gap between farm-gate and consumer retail food
prices and an enormous multiplicity of food products. The extent of processing, storage and
distribution varies considerably.
European Commission analysis exemplifies these points by commenting on the pass-through of
commodity price increases to consumer prices in the EU in 2007-08. While headline and food
inflation did increase, the contribution of agricultural commodity price increases to consumer food
prices was moderate due to ‘(i) the appreciation of the euro; (ii) the declining share of agricultural
raw materials in food production costs compared to energy and labour costs (mainly due to
increased processing) and (iii) the low share of food in the total household expenditure (today an
average EU-27 household spends around 14% of its total income on food)’ 27. The third point
mentioned by the Commission is additional to the pass-through of commodity to retail prices and is
crucial in determining the welfare impacts of rising commodity and food prices. The share of food
expenditure out of total household expenditure varies between countries depending on income, age
and composition of families. As a general rule, people with lower income levels and poorer countries
tend to spend a higher proportion of their income on food, whereas this proportion decreases
substantially in higher income groups (OECD/FAO, 2010; Ivanic et al, 2011)28. To exemplify this
point: Wiggins et al (2008; adopted from Trostle, 2008) calculate the pass-through of a 50 per cent
27

From p6 of Communication COM(2008) 321 final of 20.5.2008 from the Commission to the European
Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on
‘Tackling the challenge of rising food prices - Directions for EU action’, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2008:0321:FIN:EN:PDF.

28

In concrete numbers, Economic Research Service (US Department for Agriculture) analysis of 51 countries
found that in high-income countries, on average 16 per cent of income was spent on food, whereas in lowincome countries this average was 55 per cent (Meade and Rosen, 1997).
46

increase in the price of staples on illustrative food budgets in high- and low-income countries. They
start from the assumption that a typical high-income-country consumer spends 10 per cent of her
income on food, while this figure is 50 per cent in a low-income-food-deficit country. Based on
assumptions on the share of staples in total food consumption as well as on the domestic passthrough of staples prices29, the 50 per cent higher prices increase food expenditure as a share of
total income to 10.6 per cent and to 60.5 per cent for the high- and low-income-country consumers,
respectively. This underlines the greater impact of increases in staple commodity prices on
households in developing countries.
Clearly the transmission of price changes from global to domestic markets differs between countries
and regions30, explained among others factors by:
 A country’s food import dependency;
 Domestic policy measures aimed at delaying pass-through to domestic end-user prices, eg food
subsidies. While these may ease the impact on food consumers, it may have unfavourable
repercussions for producers in the sense of taking away the incentive to increase production in
response to higher prices (Wiggins et al, 2008).
 Movements of the local currency exchange rate against the US Dollar;
 The distance between major centres of production and consumption, and the coast; and
transport costs (HM Government, 2010).
These considerations are illustrated in a simplified way in Figure 7 below.
Figure 7: From agricultural commodity to food prices and welfare implications

Source: Own compilation

29

A partial pass-through on staples of 60 per cent is assumed; the assumptions on staples as a percentage of
total food spending are 20 and 70 per cent for the low- and high-income countries, respectively.

30

Wiggins et al (2008, p29) review the degree of transmission of international prices for world regions with
transmission found to be highest in Asia and lowest in Africa with Latin America showing a more diverse
picture.
47

5.3

Implications for the significance of EU biofuel policy in relation to projected
agricultural commodity prices by 2020?

Having looked at the compendium of studies, we provide an alternative representation of
projected price impacts based on feedstock or feedstock groups in Table 8. This allows for a
quick overview of the ranges found for the feedstocks that will be of particular importance
in meeting the EU 2020 targets (see section 2.2).
Table 8: Summary of price effects per feedstock
Feedstock (group)

Range of price effects

Commentary

Studies that focus on the effects of EU biofuel policy
Oilseeds
8 – 20%
Vegetable oils
1 – 36%
Oilseeds
9 – 20%
Cereals / maize
1 – 22%
The ESIM model (Blanco Fonseca et al, 2010)
projects an increase in maize prices of 22%.
The remaining studies project increases in
maize or cereal prices of ≤8%
Wheat
1 – 13%
Sugar (cane/beet) 1 – 21%
The ESIM model (Blanco Fonseca et al, 2010)
projects an increase in sugar prices of 21%. The
remaining three models reporting results for
sugar project price increases of ≤2%
Studies that analyse the impacts of global/multi-regional biofuel mandates
Oilseeds
2 – 7%
Vegetable oils
35%
OECD (2008) is the only ‘global’ study providing
a figure for vegetable oils
Cereals / maize
1 – 35%
Wheat
1 – 8%
Sugar (cane/beet) ~10%
Timilsina et al (2010) is the only ‘global’ study
providing a figure for sugar, 9.2 or 11.6%
depending on the scenario
Source: Own compilation based on the preceding analysis
Note: Figures are rounded to full percentages. The ranges in this table do not include results from
Davies (2012) because of a different reporting of price effects that cannot be compared to the other
studies; Taheripour et al (2010) for similar reasons (see Table 5); and Rosegrant et al (2006) given
their results are considered outdated and therefore it does not seem sensible to compare them to the
remaining studies.

We conclude the following as to the likely impacts of EU biofuel policy on agricultural
commodity prices in 2020 based on the evidence from models:
 Table 5 summarised the results of different studies and, just as importantly, the
key modelling characteristics. It is clear that the compendium of studies provides
a very heterogeneous picture from which it is very difficult to draw definite
conclusions;











The price changes projected into the future found in the studies reviewed here
are all positive, but not massive, especially in comparison to the recently
experienced global commodity price spikes. But they are not negligible either.
Another rather robust result is that there is an important difference in impacts
between commodity groups. Prices of oil crops, ie oilseeds or vegetable oils,
depending on the study, are more affected than cereals and sugar prices as a
result of EU policy. This is not a great surprise given that almost three quarters of
EU biofuel use in 2020 is anticipated to be biodiesel. This picture is summarised
in Table 8 above. This situation would be expected to change if in future the EU
adopts a policy to mitigate ILUC, as is now being discussed. That would shift the
EU biofuel demand in favour of ethanol rather than biodiesel and so potentially
increase the impacts on the prices of the principal ethanol crops, such as sugar
cane and maize.
Probably one of the more relevant studies to look at is the recent study by
Laborde (2011) that has been substantially improved in the course of preparing
the evidence base for the European Commission’s impact assessment on indirect
land use change. Laborde shows pronounced increases for, among others,
rapeseed oil, sunflower seeds and rapeseed. Focusing on the latter, the price
increases in 2020 in the policy scenario compared to the reference scenario are
11.3 and 14.1 per cent for global and EU average prices, respectively.
While these are clearly significant impacts, they do not translate to a large extent
to other crop markets and neither does the range of models project substantial
price effects for ethanol feedstocks given the relatively lower level of EU ethanol
use in 2020. Price increases for the ethanol feedstocks wheat, maize and
sugarcane/beet do not exceed 1.5 per cent in Laborde (2011). At the other end
of the range, the partial equilibrium models ESIM and CAPRI (Blanco Fonseca et
al, 2010) project increases of 8 and 12.5 per cent for wheat prices and 22 and 8.0
per cent for maize prices, respectively.
Finally, we note that EU biofuel policies do not take place in isolation in reality
and simultaneous biofuel support policies in several world regions multiply the
pressure on agricultural market impacts including commodity prices.

49

6

IMPLICATIONS FOR POLICY

The modelling results summarised in section 5.3 indicate that estimates of the effect of the
EU biofuels mandate on agricultural prices show a fairly wide range of values. In the studies
focusing on the impacts of EU biofuel policy, the most significant increases are projected for
oilseeds and vegetable oils, with increases in world prices typically ranging between 8 to 20
and 5 to 36 per cent, respectively. Increases in world rapeseed prices, which according to
projections, are expected to be the most significant feedstock for EU biofuel use in 2020, are
around 11 and 13 per cent in Laborde (2011), depending on the scenario. The ranges found
leave a considerable degree of uncertainty. The welfare effects of higher international
agricultural commodity prices (see the bottom part of Figure 7) are an even more complex
subject, a detailed analysis of which is not attempted here. However, in order to understand
the policy options for addressing the price changes identified in the models it is useful to
highlight the following points.
From the perspective of (urban) food consumers, the higher agricultural raw material prices
will in turn be transmitted down the food chain into higher retail food prices, which will
diminish living standards (all other things being equal). The size of this impact on food price
inflation is beyond any of the studies examined. Qualitatively it is expected that poorer
countries and poorer households will bear a larger burden proportionately than richer
households and countries.
It should, however, be noted that higher agricultural prices can be welcomed by rural
households in the developing world which are agricultural producers. With higher prices –
assuming such changes are transmitted back to rural areas and primary producers – their
production will become more remunerative. The surplus dumping activities of developed
country agricultural producers (like the USA and EU) have been criticised for reducing world
prices and harming the interests of poor farming households in developing countries.
Conversely, the same logic would presume that policies raising world prices would help the
interests of these groups. It is, however, commonly asserted that while increasing prices hit
net consumers of agricultural produce immediately, the supply response triggered by higher
prices (that would produce positive welfare effects in agricultural communities) takes longer
to materialise given crop growing cycles and investment costs. It may even be hampered
altogether in the presence of high price volatility that makes investments more risky while
at the same time increasing the vulnerability of the poorest households (Box 3).
The impacts of renewable energy policy on food prices are of course in addition to the direct
higher price effects of the transport fuel itself, which will result from the biofuels mandate.
Again as poorer families will tend to spend a higher proportion of their incomes on fuel as
well as food, the differential social effect is compounded. None of the studies reviewed here
is able to analyse these multiple effects on the household level. To do this, models need to
account for different household structures distinguishing between characteristics such as
urban, rural, land-owning. Such analyses of the micro level impacts of global price changes
triggered by biofuel policy are badly needed in order to truly understand the costs and
benefits of biofuel induced agricultural market impacts.

50

Ultimately the test of the policy is its efficacy and cost effectiveness in delivering its primary
objective which, in Europe at least, is to displace fossil fuels, leading to lower emissions of
greenhouse gases and thus mitigating harmful climate change. At the same time, this should
not compromise other policy objectives, such as reducing poverty and hunger in the
developing world. This study was framed to consider just one element of such an
assessment, namely the effect on agricultural commodity prices. We have not explored the
climate impact of EU biofuels policy, which requires taking account both of direct and
indirect effects through indirect land use change, although this was a main focus of some of
the studies reviewed here. It is clearly important to know if the commodity price raising
effect of biofuels policy is a cost-effective route for mitigating climate change and if such
mitigation is being achieved.
Finally, we offer some discussion of a selection of options for future EU biofuels policy that
are ‘on the policy makers’ table’ as put forward by different interest groups:
 Removal or reduction of the EU’s 10 per cent target for renewable energy in
transport;
 Enforcing stricter sustainability standards, including the removal of subsidies for
unsustainable biofuels and/or the promotion of more sustainable biofuels;
 Making biofuel mandates more flexible.
Several prominent reports have called for a removal of biofuel mandates and subsidies. A
paper prepared by ten international organisations to feed into the G20 discussions on
agricultural market volatility under the French presidency (2011), FAO, IFAD et al (2011),
recommended that ‘G20 governments remove provisions of current national policies that
subsidize (or mandate) biofuels production or consumption’ (p27). A report by the High
Level Panel of Experts on Food Security and Nutrition of the Committee on World Food
Security (HLPE, 2011), calls for the CFS (Committee on World Food Security) to ‘demand of
governments the abolition of targets on biofuels and the removal of subsidies and tariffs on
biofuel production and processing’ (p13). IFPRI suggests as one of seven steps to prevent
recurring food crises that ‘public policies, particularly in the United States and the European
Union, should aim to curtail and reform existing biofuel policies and subsidies to maximize
environmental benefits while minimizing biofuel demand’s possible contribution to the
volatility of international and domestic food markets’ (Fan et al, 2011, p3). Changing policy
to reduce or abolish policy-driven biofuel demand could reduce or altogether eliminate
agricultural commodity market impacts. However, the scale of the effect would be lessened
in a market scenario with high oil prices (and a favourable ratio of biofuel output prices and
feedstock prices), in particular given the production capacity already installed.
There are several reasons for removing biofuel support policies and agricultural market
impacts would be a factor in the equation. The Commission has to report on the impacts on
food prices and clearly there is currently a lack of understanding as to what extent the
agricultural commodity price increases, already uncertain in themselves, would translate
into retail price increases in different locations in the world and the effects this would then
have on households, some of which are agricultural producers as well as consumers. Within
the developed world which is imposing these biofuel policies, there are quite different
effects for different consumers. Having said this, it is valid at this stage to make the link to
the wider discussions on the sustainability of biofuels. The ongoing ILUC debate has
51

questioned the ability of biofuels to reduce GHG emissions. Given the uncertainty about the
ability of biofuels to contribute to one of their main underlying rationales, reducing GHG
emissions in the transport sector, any further potentially harmful impacts will compound
concern regarding their social acceptability, undermining the case for supporting policies.
The idea of removing subsidies for unsustainable biofuels and/or promoting more
sustainable biofuels takes us into the more general debate on the environmental and social
implications of biofuels. Whatever would be identified as unsustainable biofuels, removing
support for those would reduce the pressure on the markets for the respective feedstocks.
As examples of more sustainable biofuels, advanced biofuels produced from agricultural,
forestry or other wastes are often cited. Once their environmental and social implications
are better established31, promoting the use of biofuels from such feedstocks could be an
option to increase the GHG saving potential from biofuel use while at the same time
reducing their impacts on agricultural markets and commodity prices. Volumes of these
biofuels would be considerably lower than those from agricultural crops, at least in the
shorter term.
With regard to making biofuel mandates more flexible, discussion is focused on the US
situation. Unlike in the EU, the US Energy Independence and Security Act (EISA) mandates
the use of quantities of biofuels (in gallons) on a yearly basis up to 2022. Several authors
have pointed at the harmful effect of biofuel mandates in the sense that they add an
inelastic demand to agricultural markets, which are generally considered already to be
inelastic (because people do not generally change their consumption of food much in
relation to its price) (Babcock, 2011; Hertel and Beckman, 2011; Meyer and Thompson,
2010; HM Government, 2010). With such inelastic demand small changes in production can
have a disproportionate impact on prices, as seen in the price volatility of recent years. It
could be argued that EU policy is from its outset more flexible. First of all, Member States
are free to promote more heavily other forms of renewable energy in the transport sector,
hence reducing the reliance on (first-generation and crop-using) biofuels. Also, the RED does
not prescribe a stepwise roadmap on how to meet the targets by 2020. On the latter point,
Member States are tasked with defining their own trajectory to 10 per cent in the form of
the NREAPs and biannual progress reports. This could be seen as a ‘softer’, potentially more
flexible, approach compared to the strict year-on-year targets contained in the US EISA.
Nonetheless, changes to the timetable and to permit reductions in vehicle fuels
consumption to be pursued in place of a specific renewable energy target for transport fuels
would be means of adding further flexibility.
Finally, looking ahead, policy makers may have less influence over the amount of biofuels
that will be used in future in the light of rising fossil fuel prices. If the competiveness of
biofuels increases in the market with high oil prices (and favourable feedstock prices), the
role of policy, be it in the form of subsidies, mandates or tax exemptions, diminishes. Hertel
and Beckman (2011) cite Hertel et al (2010b), according to whom ‘higher oil prices
accounted for about two-thirds of the growth in US ethanol output over the 2001-2006
period. The remainder of this growth is estimated to have been driven by the replacement
31

See eg Kretschmer et al (2012) for a summary of the environmental considerations surrounding the use of
cereal straw for cellulosic ethanol production.
52

of the banned gasoline additive, MTBE, with ethanol in petroleum refining. In the EU, those
authors estimate that biodiesel growth over the same period was more heavily influenced
by subsidies. Nonetheless, they estimate that oil price increases accounted for about twofifths of the expansion in EU biofuel production over the 2001-2006 period’ (p6). Abbott et
al (2011) make the same point, highlighting the 14 billion gallon production capacity already
installed in the US, arguing that production would not fall with oil prices remaining high and
producers being able to cover variable costs, even if changes were made to the RFS
mandate. Likewise, the EU biofuel market is characterised by a structural overcapacity,
meaning that the existing capacity provides much of that needed to meet the 2020 NREAP
targets (Ecofys, 2012). Future policy scenarios need to take account of changing market
conditions.

53

7

CONCLUSIONS

This report addresses one of the principal concerns in the current biofuel debate by
providing a review of the evidence base on the impact of EU biofuel use on agricultural
commodity prices up to 2020.
The 10 per cent target for the use of renewable energy in transport set in the EU Renewable
Energy Directive is anticipated to lead to a tripling of biofuel use in the EU in 2020 compared
to 2008 levels. These are the figures derived from Member States’ National Renewable
Energy Action Plans, which also suggest that the dominance of biodiesel over ethanol in the
European biofuel market will continue. With almost exclusive reliance on first-generation
biofuels produced from traditional food and feed crops, this increase in biofuel demand
translates into a significant additional demand for these crops. In line with standard
economic theory, increasing demand while holding all other things, most notably food and
feed consumption, equal increases market prices. In the real word, ‘all other things’ are not
kept constant but rather change according to complex interactions between markets for
different crops, livestock and between world regions and according to responses to price
signals by consumers and producers. In the food versus fuels debate, the use of by-products
as animal feed and yield increases stimulated by higher output prices are two factors with a
potential to mitigate price increases. Nonetheless, they are expected to increase over the
period to 2020.
Partial and general equilibrium models are the primary tools of economists to study the
impacts of biofuel use on agricultural markets. However, they differ in their assumptions
about biofuels and in the way in which feedback mechanisms are taken into account. This,
combined with differences in biofuel scenario designs, make the comparison of different
model results a challenging task. The analysis in section 5 summarises the key characteristics
from the relevant modelling studies reviewed here; we further classified models according
to their focus on EU biofuel policy versus studies looking at the impacts of simultaneous
targets in several world regions or globally.
From looking at the EU studies, it is clear that the impacts on biodiesel feedstock prices are
more pronounced than those on ethanol feedstock. We suggest that those studies that have
been refined to contribute to the debate on ILUC and whose results with respect to ILUC are
used by a multitude of stakeholders can be expected to be more reliable than others in
terms of assessing the impacts of EU policy on agricultural commodity prices. This is most
notably the ‘IFPRI study’ by Laborde (2011), which projects oilseed price increases in 2020 in
the policy scenario compared to the reference scenario of up to 11.3 and 14.1 per cent for
global and EU average prices, respectively. Future EU legislation on ILUC may lead to a shift
from biodiesel to more ethanol use with a corresponding shift of the agricultural market
impacts from biodiesel to ethanol feedstocks.
In terms of policy options to deal with biofuel induced price increases, it is straightforward
that a reduction or even abolition of biofuel demand would reduce or prevent altogether
agricultural commodity market and hence consumer food price impacts due to biofuel
policy. However, depending on the development of oil and therefore fossil fuel prices on the
one hand and feedstock prices on the other, the market might still drive biofuel use
54

whenever these are price competitive with fossil fuels, rendering policy changes less
effective. Unlike in the USA, where biofuels are promoted by a rigid mandate laying out
annual targets for biofuel volumes consumed, EU policies relating to biofuels are slightly
more flexible and could be made more so. At present there is at least an element of
flexibility afforded to Member States in terms of how they choose to go about meeting the
target. They are obliged to deliver a certain proportion of renewable energy in transport
rather than to meet specific target volumes of biofuels. Consequently, policy can be
adjusted in different ways – either by altering the target for renewables, setting more
specific targets for different forms of renewable energy (eg advanced biofuels), increasing
the efficiency of transport vehicles etc. For example, EU policy that would encourage
Member States to increase support for advanced biofuels produced from forestry and
agricultural residues as well as renewable electricity while scaling back support for
conventional biofuels produced from food and feed crops, would be a way of reducing
pressure on agricultural markets. Alternative options for promoting renewable energy and
increasing policy flexibility need further detailed consideration and clarification could
usefully be provided by the Commission regarding the scope that Member States have to
adapt their policies up to 2020.
Finally, it should be noted that this review assessed the impacts of biofuel use on
agricultural commodity prices. The European Commission, as part of its reporting
requirements in 2012, is required to judge the impacts of biofuels on food prices. This
further step increases the complexity of the analysis substantially. Food supply chains
involve an extremely wide range of plant and animal products with varying degrees of
processing of agricultural commodities and varying proportions of staple foodstuffs in the
diet. The vulnerability of consumers across the world to food price increases differs
markedly between countries and across households, depending inter alia on income levels,
household composition, and on the household status as net consumers or producers of
agricultural and food stuffs. We are not aware of studies that use multi-household models,
which would produce a better understanding of the impacts of enhanced biofuel use on
different population groups and allowing more solid estimates of the welfare impacts of
biofuel policy. This gap should be closed in order to provide decision makers with a more
complete evidence base feeding into the political review processes ongoing in 2012.

55

REFERENCES
Abbott, P C, Hurt, C and Tyner, W E (2008) What’s Driving Food Prices? Purdue University,
Farm Foundation.
Abbott, P C, Hurt, C and Tyner, W E (2011) What’s Driving Food Prices in 2011? Purdue
University, Farm Foundation.
Acamovic, T, and Pritchard, S (2008) The nutritional value of Biofuel co-products for poultry,
British Society of Animal Science Annual Meeting, March 31 to April 2, 2008 on behalf of the
Scottish Agricultural College
Agra Europe (2012) World Bank expects drop in food prices in 2012. January 24, 2012, No
2498, p7, www.agraeurope.com.
Al-Riffai, P, Dimaranan, B and Laborde, D (2010) Global Trade and Environmental Impact
Study of the EU Biofuels Mandate. Final report March 2010, International Food Policy
Institute
(IFPRI):
Washington
DC.
Available
at:
http://trade.ec.europa.eu/doclib/docs/2010/march/tradoc_145954.pdf.
Babcock, B A (2011) The Impact of US Biofuel Policies on Agricultural Price Levels and
Volatility. ICTSD Programme on Agricultural Trade and Sustainable Development; Issue
Paper No 35; ICTSD International Centre for Trade and Sustainable Development, Geneva,
Switzerland, www.ictsd.org.
Baffes, J and Haniotis, T (2010) Placing the 2006/08 Commodity Price Boom into Perspective.
Policy Research Working Paper 5371, The World Bank: Washington DC.
Balcombe, K G and Rapsomanikis, G (2008) Bayesian Estimation and Selection of Nonlinear
Vector Error Correction Models: The Case of the Sugar-Ethanol-Oil Nexus in Brazil. American
Journal of Agricultural Economics 90(3), 658–668.
Banse, M, van Meijl, H, Tabeau, A and Woltjer, G (2008) Will EU Biofuel Policies Affect
Global Agricultural Markets? European Review of Agricultural Economics 35 (2), 117–141.
Beurskens, L W M, Hekkenberg, M and Vethman, P (2011) Renewable Energy Projections as
Published in the National Renewable Energy Action Plans of the European Member States Covering all 27 EU Member States with updates for 20 Member States. ECN-E--10-069,
Energy Research Centre of the Netherlands: Petten.
Blanco Fonseca, M, Burrell, A, Gay, H, Henseler, M, Kavallari, A, M'Barek, R, Pérez
Domínguez, I and Tonini, A (2010) Impacts of the EU biofuel target on agricultural markets
and land use: a comparative modelling assessment. JRC Scientific and Technical Reports EUR
24449 EN, European Commission Joint Research Centre, Institute for Prospective
Technological Studies: Seville.

56

Bowyer, C (2011) Anticipated Indirect Land Use Change Associated with Expanded Use of
Biofuels and Bioliquids in the EU – An Analysis of the National Renewable Energy Action
Plans (Update March 2011). Institute for European Environmental Policy (IEEP): London.
Britz, W and Hertel, T W (2011) Impacts of EU biofuels directives on global markets and EU
environmental quality: An integrated PE, global CGE analysis. Agriculture, Ecosystems &
Environment 142(1–2), 102–109.
Bruinsma, J (2011) The resources outlook: by how much do land, water and crop yields need
to increase by 2050? In: Conforti, P (Ed) (2011) Looking ahead in world food and agriculture:
Perspectives to 2050. Food and Agriculture Organization of the United Nations: Rome.
Committee on Climate Change (2011) Bioenergy review. Available at:
http://downloads.theccc.org.uk.s3.amazonaws.com/Bioenergy/1463%20CCC_Bioenergy%2
0review_bookmarked_1.pdf.
Davies, G (2012) Removing Biofuel Support Policies: An Assessment of Projected Impacts on
Global Agricultural Markets using the AGLINK-COSIMO model. Analytical piece of work by
Defra
Economics,
URN:
12D/084.
Available
at:
http://www.decc.gov.uk/assets/decc/11/meeting-energy-demand/bio-energy/5134removing-biofuel-support-policies-an-assessment-o.pdf.
Delgado, C L (2003) Rising Consumption of Meat and Milk in Developing Countries Has
Created a New Food Revolution. The Journal of Nutrition 133, 3907S–3910S.
Du, X, Yu, C L and Hayes, D J (2009) Speculation and Volatility Spillover in the Crude Oil and
Agricultural Commodity Markets: A Bayesian Analysis. Paper presented at the Agricultural
and Applied Economics Association July 26-28, 2009, Milwaukee, Wisconsin.
Ecofys (2012) Assessing grandfathering options under an EU ILUC policy [by Peters, D, van
den Bos, A and van de Staaij, J]. Study commissioned by Transport & Environment. Available
at:
http://www.transportenvironment.org/sites/default/files/media/ecofys_2012_grandfatheri
ng_iluc.pdf.
Ecofys, Agra CEAS, Chalmers University, IIASA and Winrock (2011) Biofuels Baseline 2008.
Study for the European Commission DG Energy under Tender No. TREN/D1/458/2009, by
Hamelinck, C, Koper, M, Berndes, G, Englund, O, Diaz-Chavez, R, Kunen, E, Walden, D et al,
available at:
http://ec.europa.eu/energy/renewables/studies/doc/biofuels/2011_biofuels_baseline_200
8.pdf.
Edwards, R, Szekeres, S, Neuwahl, F and Mahieu, V (2008) Biofuels in the European Context:
Facts and Uncertainties. European Commission Joint Research Centre (JRC): Petten and
Ispra. Available at: http://ec.europa.eu/dgs/jrc/downloads/jrc_biofuels_report.pdf.

57

Edwards, R, Mulligan, D and Marelli, L (2010) Indirect Land Use Change from increased
biofuels demand: Comparison of models and results for marginal biofuels production from
different feedstock. European Commission Joint Research Centre: Ispra.
http://ec.europa.eu/energy/renewables/studies/doc/land_use_change/study_4_iluc_model
ling_comparison.pdf
Elobeid, A and Hart, C (2007) Ethanol Expansion in the Food versus Fuel Debate: How Will
Developing Countries Fare? Journal of Agricultural & Food Industrial Organization 5(2) Art.6.
EUROBSERV’ER (2011) Biofuels Barometer July 2011, Systèmes Solaires le journal des
énergies
renouvelables
N°
204.
Available
at :
http://www.eurobserver.org/pdf/biofuels_2011.pdf.
European Commission (2011) Questions and Answers on Food Waste Minimisation and
Food
Packaging
Optimisation,
MEMO/11/598,
14/09/2011
http://europa.eu/rapid/pressReleasesAction.do?reference=MEMO/11/598.
European Commission DG Agriculture and Rural Development [EC DG AGRI] (2007) Note-tofile AGRI G-2/WM D(2007) of 30 April 2007, Subject: The impact of a minimum 10%
obligation for biofuel use in the EU-27 in 2020 on agricultural markets, Ref.: Impact
assessment of the Renewable Energy Roadmap - March 2007. Available at:
http://ec.europa.eu/agriculture/analysis/markets/biofuel/impact042007/text_en.pdf.
European Commission DG Agriculture and Rural Development [EC DG AGRI] (2008) High
prices on agricultural commodity markets: situation and prospects – A review of causes of
high prices and outlook for world agricultural markets. July.
European Commission DG Agriculture and Rural Development [EC DG AGRI] (2011a) High
commodity prices and volatility …what lies behind the roller coaster ride? Agricultural
Markets Brief, Brief no 1, June 2011.
European Commission DG Agriculture and Rural Development [EC DG AGRI] (2011b)
Prospects for Agricultural Markets and Income in the EU 2011 – 2020, December 2011.
Fan, S, Torero, M and Headey, D (2011) Urgent Actions Needed to Prevent Recurring Food
Crises. IFPRI Policy Brief 16, March, International Food Policy Institute (IFPRI): Washington
DC. Available at: http://www.ifpri.org/sites/default/files/publications/bp016.pdf.
FAO (2008) The State of Food and Agriculture, Biofuels: Prospects, Risks and Opportunities,
Food And Agriculture Organization of the United Nations Rome, 2008
FAO (2010) Bioenergy and Food Security: The BEFS Analytical Framework. Food and
Agriculture
Organization
of
the
United
Nations:
Rome.
http://www.fao.org/docrep/013/i1968e/i1968e.pdf.
FAO, IFAD, IMF,OECD, UNCTAD, WFP, the World Bank, the WTO, IFPRI and the UN HLTF
(2011) Price Volatility in Food and Agricultural Markets: Policy Responses. Policy Report
prepared for the 20), available at: http://www.oecd.org/dataoecd/40/34/48152638.pdf.
58

Fischer, G (2009) World Food and Agriculture to 2030/50: How do climate change and
bioenergy alter the long-term outlook for food, agriculture and resource availability? Expert
Meeting on ‘How to Feed the World in 2050’, Food and Agriculture Organization of the
United Nations, Economic and Social Development Department.
Fischer, G (2011) How can climate change and the development of bioenergy alter the longterm outlook for food and agriculture? In: Conforti, P (Ed) (2011) Looking ahead in world
food and agriculture: Perspectives to 2050. Food and Agriculture Organization of the United
Nations: Rome.
Fischer, G, Hizsnyik, E, Prieler, S, Shah, M and van Velthuizen, H (2009a) Biofuels and Food
Security, Implications of an accelerated biofuels production. Summary of the OFID study
prepared
by
IIASA.
Available
at:
http://www.ofid.org/publications/PDF/pamphlet/ofid_pam38_Biofuels.pdf.
Fischer, G, Hizsnyik, E, Prieler, S, Shah, M and van Velthuizen, H (2009b) Biofuels and Food
Security. OFID study prepared by IIASA. OPEC Fund for International Development (OFID):
Vienna and International Institute for Applied Systems Analysis (IIASA): Laxenburg.
Foresight (2011) The Future of Food and Farming. Final Project Report, The Government
Office for Science: London.
Fritsche, U R, van Steen, H, Petersen, J-E, Marelli, L, Andersson, K, Berndes, G and Urbancic,
N (2012) Proceedings of the Workshop on Biofuels and Indirect Land Use Change, Brussels,
25 January 2012. Report for the European Parliament’s Directorate General for Internal
Policies, Policy Department A: Economic and Scientific Policy, IP/A/ENVI/WS/2011-18,
available
at:
http://www.europarl.europa.eu/document/activities/cont/201202/20120229ATT39558/20
120229ATT39558EN.pdf.
Gerber, N, van Eckert, M and Breuer, T (2008) The Impact of Biofuel Production on Food
Prices: a review. ZEF – Discussion Papers on Development Policy No. 127, Center for
Development Research: Bonn.
Hayes, D, Babcock, B, Fabiosa, J, Tokgoz, S, Elobeid, A, Yu, T, Dong, F, Hart, C, Chavez, E, Pan,
S, Carriquiry, M and Dumortier, J (2009) Biofuels: Potential Production Capacity, Effects on
Grain and Livestock Sectors, and Implications for Food Prices and Consumers. Journal of
Agricultural and Applied Economics 41 (2) 465–491.
Headey, D and Fan, S (2010) Reflections on the Global Food Crisis. International Food Policy
Institute (IFPRI): Washington, DC.
Hertel, T W, McDougall, R A, Narayanan G, B and Aguiar, A H (2008). Chapter 14: Behavioral
Parameters of the GTAP 7 Data Base Documentation. Center for Global Trade Analysis,
Department of Agricultural Economics, Purdue University, West Lafayette, USA.

59

Hertel, T W, Golub, A A, Jones, A D, O’Hare, M, PLevin, R J and Kammen, D M (2010a) Effects
of US Maize Ethanol on Global Land Use and Greenhouse Gas Emissions: Estimating Marketmediated Responses. BioScience 60(3), 223–231. Supporting online material available at:
https://www.gtap.agecon.purdue.edu/resources/download/4606.pdf.
Hertel, T W, Tyner, W E and Birur, D K (2010b) Global Impacts of Biofuels. Energy Journal
31(1), 75–100.
Hertel, T and Beckman, J (2011) Commodity Price Volatility in the Biofuel Era: An
Examination of the Linkage between Energy and Agricultural Markets. Paper (revised,
February 12, 2011) prepared for the NBER Agricultural Economics Conference, March 4-5,
2010, Cambridge, Mass. http://www.nber.org/chapters/c12113.pdf.
HGCA (2010) Growing wheat for alcohol/ bioethanol production, Information sheet 11
http://publications.hgca.com/publications/documents/Biofuels.pdf
HLPE (2011) Price volatility and food security. A report by the High Level Panel of Experts on
Food Security and Nutrition of the Committee on World Food Security: Rome.
HM Government (2010) The 2007-08 Agricultural Price Spikes: Causes and Policy
Implications. Department for Environment Food and Rural Affairs: London. Available at:
http://archive.defra.gov.uk/foodfarm/food/pdf/ag-price100105.pdf.
Ivanic, M, Martin, W, Zaman, H (2011) Estimating the Short-Run Poverty Impacts of the
2010–11 Surge in Food Prices. Policy Research Working Paper 5633, The World Bank:
Washington DC.
Jagger, S (2008) Use of biofuel co-products in pigs - economics and nutritional limitations,
British Society of Animal Science Annual Meeting, March 31 to April 2, 2008 on behalf of the
Association of British Agriculture
Kampman, B, Brouwer, F, Schepers, B (2008) Agricultural land availability and demand in
2020 A global analysis of drivers and demand for feedstock, and agricultural land
availability, Delft, CE.
Keeney, R and Hertel, T W (2009) The Indirect Land Use Impacts of U.S. Biofuel Policies:
The Importance of Acreage, Yield, and Bilateral Trade Responses. American Journal of
Agricultural Economics 91(4), 895–909.
Kretschmer, B (2011) The Land-Use Implications of EU Bioenergy Policy: Going Beyond ILUC.
Institute for European Environmental Policy (IEEP): London.
Kretschmer, B, Allan, B and Hart, K (2012) Mobilising Cereal Straw in the EU to Feed
Advanced Biofuel Production. Institute for European Environmental Policy (IEEP): London.
Kretschmer, B and Peterson, S (2011) Integrating bioenergy into computable general
equilibrium models — A survey. Energy Economics 32, 673–686.
60

Laborde, D (2011) Assessing the Land Use Change Consequences of European Biofuel
Policies. Final report October 2011, International Food Policy Institute (IFPRI): Washington
DC. http://www.ifpri.org/sites/default/files/publications/biofuelsreportec2011.pdf.
Lee H, Hertel T, Sohngen B, Ramankutty N (2005) Towards and integrated land use data
base for assessing the potential for greenhouse gas mitigation. GTAP Technical Paper No 25,
Center for Global Trade Analysis, Purdue University: West Lafayette. Available at:
https://www.gtap.agecon.purdue.edu/resources/download/2375.pdf.
Martin, W and Anderson, K (2011) Export Restrictions and Price Insulation During
Commodity Price Booms. Policy Research Working Paper 5645, The World Bank: Washington
DC.
Meade, B and Rosen, S (1997) The Influence of Income on Global Food Spending. Agricultural
Outlook USDA/ERS.
Msangi, S and Rosegrant, M (2011) World agriculture in a dynamically changing
environment: IFPRI’s long-term outlook for food and agriculture. In: Conforti, P (Ed) (2011)
Looking ahead in world food and agriculture: Perspectives to 2050. Food and Agriculture
Organization of the United Nations: Rome.
Nabim
(2012)
Wheat
guide,
http://www.hgca.com/document.aspx?fn=load&media_id=7611&publicationId=5880
OECD (2008) Biofuel support policies — an economic assessment. OECD: Paris.
OECD/FAO (2010) Agricultural Outlook 2010-2019. OECD: Paris.
OECD/FAO (2011) Agricultural Outlook 2011-2020. OECD: Paris.
Oxfam (2008) Another Inconvenient Truth, How biofuel policies are deepening poverty and
accelerating climate change, Oxfam Briefing 114
Pfuderer, S and del Castillo, M (2008) The Impact of Biofuels on Commodity Prices. UK
Department for Environment, Food and Rural Affairs (Defra), London.
http://archive.defra.gov.uk/evidence/series/documents/impact-biofuels-commodities.pdf.
Piesse, J and Thirtle, C (2010) Agricultural R&D, technology and productivity. Phil. Trans. R.
Soc. B 365, 3035-3047.
Rajagopal, D, Sexton, S, Hochman, G, Roland-Holst, D and Zilberman, D (2009) Model
estimates food-versus-biofuel trade-off. California Agriculture 63(4), 199-201.
Rapsomanikis, G and Hallam, D (2006) Threshold cointegration in the sugar-ethanol-oil price
system in Brazil: evidence from nonlinear vector error correction models. FAO Commodity

61

and Trade Policy Research Working Paper 22, Food and Agriculture Organization of the
United Nations: Rome.
Renewable Fuels Agency (2008) The Gallagher Review of the indirect effects of biofuels
production
Ros, J, van den Born, G, J, and Notenboom, J (2010) The contribution of by-products to the
sustainability of biofuels, Netherlands Environmental Assessment Agency (PBL).
Rosegrant, M W, Msangi, S, Sulser, T and Valmonte-Santos, R (2006) Biofuels and the Global
Food Balance. International Food Policy Research Institute (IFPRI): Washington, DC.
Available at: http://www.ifpri.org/sites/default/files/publications/focus14_03.pdf.
Schmidhuber (2007) Biofuels: An emerging threat to Europe’s Food Security? Impact of an
Increased biomass use on agricultural markets, prices and food security: A longer-term
perspective. Notre Europe Policy Paper 27, May. Available at: http://www.notreeurope.eu/uploads/tx_publication/Policypaper-Schmidhuber-EN.pdf.
Schroeder, T C, Barkley, A P and Schroeder, K C (2010) Income Growth and International
Meat Consumption. Journal of International Food and Agribusiness Marketing 7 (3) 15–30.
Trostle, R (2008) Global Agricultural Supply and Demand: Factors Contributing to the
Recent Increase in Food Commodity Prices. WRS-0801 Economic Research Service,
USDA, Washington. http://www.ers.usda.gov/Publications/WRS0801/WRS0801.pdf
Slade, R, Saunders, R, Gross, R and Bauen, A (2011) Energy from biomass: the size of the
global resource, An assessment of the evidence that biomass can make a major contribution
to future global energy supply, A report produced by the Imperial College Centre for Energy
Policy and Technology for the Technology and Policy Assessment Function of the UK Energy
Research Centre
Taheripour F, Birur D, Hertel T, Tyner W (2008) Introducing liquid biofuel into the GTAP
database. GTAP Research Memorandum No. 11 2007 (revised May 1, 2008), Center for
Global Trade Analysis, Purdue University: West Lafayette. Available at:
https://www.gtap.agecon.purdue.edu/resources/download/3939.pdf.
Taheripour, F, Hertel, T W, Tyner, W E, Beckman, J F and Birur, D K (2010) Biofuels and their
by-products: Global economic and environmental implications. Biomass and Bioenergy 34,
278–289.
Timilsina, G, Beghin, J, van der Mensbrugghe, D and Mevel, S (2010) The Impacts of Biofuel
Targets on Land-Use Change and Food Supply, A Global CGE Assessment. Policy Research
Working Paper No. 5513, The World Bank: Washington DC.
Timilsina, G R and Shrestha, A (2010) Biofuels Markets, Targets and Impacts. Policy Research
Working Paper No. 5364, The World Bank: Washington DC.

62

Tyner, W (2009) The integration of energy and agricultural markets. Presented at the 27th
International Association of Agricultural Economists, Beijing, China, August 16-22, 2009.
van der Mensbrugghe, Osorio-Rodarte, I, Burns, A and Baffes, J (2011) Macroeconomic
environment and commodity markets: a longer-term outlook. In: Conforti, P (Ed) (2011)
Looking ahead in world food and agriculture: Perspectives to 2050. Food and Agriculture
Organization of the United Nations: Rome.
WBGU (2009) Welt im Wandel: Zukunftsfähige Bioenergie und nachhaltige Landnutzung.
Wissenschaftlicher Beirat der Bundesregierung Globale Umweltveränderungen (WBGU):
Berlin.
Wiggins, S, Fioretti, E, Keane, J, Khwaja, Y, McDonald, S, Levy, S and Srinivasan, C S (2008)
Review of the indirect effects of biofuels: Economic benefits and food insecurity. Report to
the Renewable Fuels Agency by the Overseas Development Institute (ODI): London.
https://files.pbworks.com/download/YOk0fu7xGm/npnet/12639257/ODI%20%282008%29%20Review%20indirect%20effects%20biofuels%20Economic%20benefits_and_food_insecurity.pdf
Zanchi, G, Pena, N, Bird, N (2010) The upfront carbon debt of bioenergy, Research by the
Joanneum Research for Birdlife International.

63

ANNEX 1: LAND REQUIREMENTS FOR BIOFUELS
In order to get an under understanding about the magnitudes of land used for biofuel
production this annex presents 1) biofuel yields per hectare, 2) estimates of land required
for EU biofuel consumption as well as 3) feedstock used for biofuel production in relation to
global crop production. Table A1 summarises biofuel yields per hectare in different
measures as well as their inverse, hectare requirements per Mtoe biofuel.
Table A1: Examples of biofuel yields per hectare32

Sugarcane ethanol (Brazil)
Wheat ethanol (global)
Palmoil (Malaysia) biodiesel
Soy biodiesel (USA)
Rapeseed biodiesel (EU)
EU biofuels average*

l/ha

toe/ha

5470
950
4700
550
777-1249

2.75
0.48
3.73
0.44
0.62-0.99

ha/toe
(=million ha/Mtoe)
0.364
2.095
0.268
2.293
1.009-1.623
0.590

Source: FAO (2008), Blanco Fonseca et al (2010); own conversion in tonnes of oil equivalent (toe)
based on lower heating values (LHV) of 21.1 MJ/l for ethanol and 33.3 MJ/l for biodiesel.
Notes: One litre of ethanol replaces 0.66 litre of gasoline; one litre of biodiesel replaces 0.91 litre of
biodiesel. * Ecofys et al (2011), based on amount of biofuels consumed in the EU in 2008 and the
total land estimated to be used for their production in 2008, see Table A2.

Ecofys et al (2011) provide biofuel land use estimates for 2008 as summarised in Table A2.
Using ecological-economic modelling tools, Ecofys et al (2011) show that ‘the increase in
biofuel production in the EU between 2000 and 2008 has led to an increased global
agricultural land use of 1.3 Mha’ (million hectare), falling short of the 3.6 million hectares
estimated to be used for EU biofuel production in 2008. This shows that ‘part of the land
used for biofuels feedstock production became available through yield improvements of
other crops, or at the cost of decreasing production of other crops’ (p135).
With regard to future land needs: A simple calculation based on the average hectare
requirement of 0.59 per tonne of oil equivalent biofuel together with an anticipated EU
biofuel consumption of 29.6 Mtoe in 2020 yields a land requirement for EU biofuel use in
2020 of 17.5 million hectares globally. Note this represents a very crude estimate, as it
neglects any yield improvements (which will reduce the land requirement) and does not
account for potential shifts in biofuel pathways (could either reduce or increase the land
requirement). To put the admittedly crude estimate in perspective: Current EU arable land
amounts to 104 million ha, while the utilised agricultural area (UAA; includes arable land,

32

Note this table serves as an illustration and the figures are crucially dependent on assumptions on crop
yields, and conversion efficiencies. This becomes apparent in particular with regard to the figure above on
wheat ethanol. While the FAO figure is based on global yield assumptions, a figure derived from a German
source gives a much higher yield for wheat ethanol of 2760 litres per hectare (http://www.biokraftstoffe.info/kraftstoffe/bioethanol/rohstoffe/). This converts to a requirements of 0.721 hectares per
tonne of oil equivalent.
64

permanent crops and permanent grassland) amounts to 170 million ha 33. Extrapolated
global biofuel land needs in 2020 for EU biofuel consumption are around 10 per cent of
current EU utilised agricultural area or 17 per cent of 2007 EU arable area.
Table A2: Land use in EU 2008

Gross land use
of which in the EU
of which in third countries
Net land use (accounting for co-products)
Source: Ecofys et al (2011)

Million hectare for EU
biofuel consumption in 2008
7
3.6
3.3
3.6

Table A3 compiles data on the use of feedstock for biofuel production. The figures refer to
the quantity of a given feedstock used to produce biofuels in the EU. These are compared to
global production of a given crop. The resulting shares are all relatively small, with
noticeably differences between them, however. While vegetable oil use for biofuel
production in the EU reaches 5.3 per cent, the share for wheat is negligible with 0.5 per
cent.
Table A3: Use of feedstock commodities for biofuel production in the EU in 2007

Wheat
Sugar beet
Vegetable oils

Use for biofuel Global feedstock Share of global
production in EU production
(in production
(in 1000 tonnes) 1000 tonnes)
0.5%
2,851
591,833
3.8%
9,281
244,237
5.3%
5,698
106,517

Source: Based on OECD (2008, p78).
Note: The quantities of feedstock used for biofuel production in the EU (first column) may include
imports and domestic feedstocks and are used to produce biofuels that may either be domestically
consumed or exported.

33

These are figures for 2007 and 2010, respectively, taken from: Eurostat statistical database
(http://epp.eurostat.ec.europa.eu/portal/page/portal/agriculture/data/database); Eurostat (2011) EUAgricultural census 2010 - first results,
http://europa.eu/rapid/pressReleasesAction.do?reference=STAT/11/147&format=HTML&aged=0&language
=EN&guiLanguage=en.
65

ANNEX 2: SUMMARY TABLE OF STUDIES REVIEWED

Study and model used
Rosegrant et al (2006)
Short model description:
- Partial equilibrium. Uses IFPRI’s IMPACT model34
(International Model for Policy Analysis of Agricultural
Commodities and Trade)
- Biofuels modelled implicitly: ‘The model contains three
categories of commodity demand: food, feed, and other
use demand. This study manipulates the “other use”
demand category in order to reflect the use of
commodities as biofuel feedstocks…’
- Second-generation biofuels: cellulosic feedstock sources
(crop residues and switchgrass) not represented, so use
of second-generation biofuels modelled by reducing the
demand for first-generation feedstock
- No trade in biofuels
- Biodiesel use only in the EU
 Regions and sectors
30 commodities accounting for ‘virtually all of world food
production and consumption’; 115 model regions
 Modelling land
Not specified
 Modelling by-products
Not specified

34

Scope and scenarios (geo – time – research questions)
Main results
Scenario description:
 Reporting of EU versus global price impacts?
- Baseline scenario: Biofuel volumes frozen at 1997 (base No, only world prices
year) levels
 Differentiated reporting of commodity versus retail
price impacts?
- Policy scenarios:
No, commodity level only
1: Aggressive biofuel growth holding projected  Comparator for price effects
productivity increases for yields at baseline projection Policy 2020 compared to baseline 2020 (not entirely clear
levels: biofuel shares (energy equivalent) in 2020 are: from report)
RoW 2%, USA 4%, China 8%, EU 10%, India 11%, Brazil
58%
Price increases in 2020 compared to the baseline in
2: Cellulosic biofuel as of 2015: crudely modelled by scenarios 1, 2 and 3, respectively, for six crops reported
holding demand for conventional biofuel crops constant are:
after 2015
3: Aggressive biofuels growth with productivity change - Cassava: 135, 89, 54%
and availability of cellulosic biofuels
- Maize: 41, 29, 23%
- Oilseeds: 76, 45, 43%
- Sugar beet: 25, 14, 10%
 Base year – model horizon
- Sugarcane: 66, 49, 43%
1997 to 2020
- Wheat: 30, 21, 16%
 Absolute magnitudes of biofuel use (EU & global)
Not specified
 Increase in magnitudes in scenarios
Not specified
 NREAP demand considered?
No, not available at the time, but 10% biofuel share
modelled (corresponding biofuel volume not specified)

Model description available at: http://www.ifpri.org/sites/default/files/publications/impactwater.pdf

Study and model used
EC DG AGRI (2007)35
Short model description:
- ESIM (European Simulation Model), partial equilibrium
model for the agricultural sector36
- biodiesel and ethanol demand up to 2020 based on
energy projections of DG TREN’s PRIMES model
 Regions and sectors
Covers agricultural products; all EU Member States +
Turkey and USA modelled as individual countries;
remaining countries are aggregated as 'rest of the world'.
 Modelling land
Not specified
 Modelling by-products
Taken into account

Scope and scenarios (geo – time – research questions)
Scenario description:
- Policy baseline: 2020 situation under 2003 biofuels
directive, expected to lead to a 6.9% biofuel share or
23.8 Mtoe biofuel consumption
- New policy scenario: 2020 situation under a 10% biofuel
target, leading to 34.6 Mtoe, ie an additional 10.8 Mtoe
of biofuel use in 2020
- Penetration of second-generation biofuels: 30% of
domestic biofuel production from second-generation,
sensitivity runs lowering contribution; secondgeneration biofuels seem to be from (land-using) crops,
but assumptions not specified apart from ‘significantly
higher energy per hectare’ yields mentioned as a general
remark
- Agricultural policies as in 2007 and kept constant
thereafter
 Base year – model horizon
20004/05 to 2020
 Absolute magnitudes of biofuel use (EU & global)
‘Baseline’ versus ‘new policy’ in 2020: 23.8 Mtoe vs 34.6
Mtoe; global use not specified
 Increase in magnitudes in scenarios
10.8 Mtoe; global not specified
 NREAP demand considered?

Main results
 Reporting of EU versus global price impacts?
No, reported prices appear to be EU27 prices
 Differentiated reporting of commodity versus retail
price impacts?
No, commodity level only
 Comparator for price effects
Comparison over time 2006-2020
Price increases over 2006-202037:
- Cereals: 3 to 6%
- Sunflower seeds: 15%
- Rapeseed: 8 to 10%
- Soy bean oil : ‘significant’ increases
In the policy baseline (2003 directive) ‘prices [in 2020] for
agricultural raw materials would be similarly firm as
under the 10% scenario [in 2020], however with slightly
lower increases’ (p7)
No 2nd-gen biofuel production: 50% import share and
‘agricultural prices would be significantly higher’ (p9)
Share of arable land used for biofuel production in 2020
with 10% target: 15%

35

Results summarised in EC DG AGRI (2007) are cited in the annex of the impact assessment conducted by the European Commission in preparation for the 2008 climateenergy package: http://ec.europa.eu/energy/renewables/doc/sec_2008_85-2_ia_annex.pdf.

36

According to EC DG AGRI (2007) the model is recursive-dynamic, while a short model description available at http://agrilife.jrc.ec.europa.eu/ESIM.htm states it is
comparative static.

37

Edwards et al comment on results for price developments over time in EC DG AGRI (2007): ‘In 2007 OECD projected significant falls in inflation-adjusted world crop
prices by 2020, whereas FAPRI projected a much gentler fall. [EC DG-AGRI 2007] bases its projection on OECD, so the rise in prices due to biofuels is partly masked by the
overall fall in real prices in the background projection’ (2008, p17).
67

Study and model used

Scope and scenarios (geo – time – research questions)
No, not available at the time, but magnitudes modelled
here are higher than biofuel consumption of close to 30
Mtoe (27.3 Mtoe first-generation) in 2020 according to
the NREAPS
Banse et al (2008)
Scenario description:
- 10% EU biofuel blending target by 2020 compared to a
Short model description:
reference (‘Global economy’) scenario with no biofuel
- Computable general equilibrium model LEITAP, which is blending obligations
a modified version of the GTAP-E model
- Declining real agricultural prices in the reference
- Biofuels not modelled as separate sectors but rather as scenario as a result of the considerable degree of trade
intermediate inputs to the petroleum industry; and agricultural market liberalization assumed as well as
therefore, trade in biofuels cannot be modelled
the inherent assumptions on productivity growth rates
- Second-generation biofuels not included
 Base year – model horizon
 Regions and sectors
2001 (GTAP 6) to 2020
37 regions and 23 sectors, with focus on the detailed  Absolute magnitudes of biofuel use (EU & global)
representation of EU countries and other important Not specified
agricultural market actors and on agricultural and energy  Increase in magnitudes in scenarios
sectors (see Banse et al, 2008, p122 for details)
Not specified
 Modelling land
 NREAP demand considered?
Land availability determined by land supply curves No, not available at the time. 10% target modelled but
representing the relationship between land supply and associated quantities are not specified
rental rates allowing for conversion of non-agricultural
land into cropland;
Allocation of land across different land covers and uses
according to a nested CET function (see Fig 4, p124)
 Modelling by-products
Not included
Edwards et al (2008)
Scenario description:
st
- 10% 1 generation ethanol and biodiesel use in EU in
Short model description:
2020, compared to no EU biofuel use in 2020
- No modelling; acknowledge that modelling needed to
take cross-feedstock effects into account
 Base year – model horizon
- Calculations based on DG TREN fuel demand scenario Looking at 2020
calculating the amount of biodiesel and bioethanol  Absolute magnitudes of biofuel use (EU & global)
68

Main results

 Reporting of EU versus global price impacts?
No, only world prices
 Differentiated reporting of commodity versus retail
price impacts?
No, only commodity level
 Comparator for price effects
Policy 2020 compared to reference 2020
World price increases between reference and policy
scenario in 2020:
- Oilseeds: 8.5%
- Sugar: 2%
- Cereals: 6%
To note the development of prices over time: even in the
policy scenario, sugar and cereal prices decrease from
2011-2020 (both by around 6-7%), while only oilseed
prices increase (1-2%)

 Reporting of EU versus global price impacts?
No, only world prices
 Differentiated reporting of commodity versus retail
price impacts?
No, only commodity level
 Comparator for price effects
2020 policy compared to 2020 no biofuel use

Study and model used
needed for 10% shares biodiesel and 10% ethanol
shares; calculating the amount of vegetable oil / cereals
needed to meet this and the share of this demand out of
total global supply in 2020 (based on FAPRI); combined
with assumptions on long-term area response supply
flexibilities to calculate price increases

Scope and scenarios (geo – time – research questions)
Only partly specified, see below
 Increase in magnitudes in scenarios
See above (reference assumes zero biofuel use)
 NREAP demand considered?
No, not available at the time; mention that 10% of 2020
diesel demand is ~19.2 Mtoe (p9; no corresponding figure
for ethanol mentioned)

 Regions and sectors
Not applicable (EU focus)
 Modelling land
Not applicable
 Modelling by-products
No
OECD (2008)

Scenario description:
- Counterfactual with no biofuel policy support
Short model description:
- Baseline scenario incl existing biofuel support measures
‘[projecting] a substantial further growth in the
AGLINK-COSIMO, OECD-FAO partial equilibrium modelling production and use of both ethanol and biodiesel’. No
system underlying the Agricultural Outlook.
second-generation biofuels
- Policy scenario modelling the separate and combined
effects of new policies (US EISA and EU RED) assuming a
 Regions and sectors
contribution to the target is met by second-generation
Includes major temperate-zone agricultural commodities; biofuels (on average 50% of biomass for secondextended to include sugar and sweeteners;
generation assumed to come from land otherwise used
Large developing country coverage through link to Cosimo
for food and feed production, rest from crop residues,
 Modelling land
see note 15 in OECD, 2008, p91)
Not specified
 Modelling by-products
 Base year – model horizon
Included (OECD, 2008, p125 for details)
2008-2017

69

Main results
Prices compared to no biofuel use in 2020:
- 10% first-generation ethanol blend in EU gasoline would
use, ~2.5% of world 2020 cereals, causing a world
cereals price increase of at least 4%.
- 10% first-generation biodiesel blend in EU diesel would
use ~19% of world 2020 vegetable oils, causing a world
price increase of at least 24%.

 Reporting of EU versus global price impacts?
No, only world market prices
 Differentiated reporting of commodity versus retail
price impacts?
No, only commodity level
 Comparator for price effects
Comparison of average 2013-2017 prices in different
scenarios
Average 2013-2017 world market prices in counterfactual
with no support compared to baseline:
- Oilseeds: -3%
- Wheat: -5%
- Coarse grains: -7%
- Vegetable oils: -16%

Study and model used

Scope and scenarios (geo – time – research questions)
 Absolute magnitudes of biofuel use (EU & global)
In policy scenario, 25.4 Mtoe in EU and 96.0 globally
(2013-2017 average)38
 Increase in magnitudes in scenarios
From counterfactual (no support) to policy scenario: 19.7
Mtoe in EU and 38.7 globally (2013-2017 average)39

Short model description:
- Ecological-economic modelling approach, incorporating
the FAO/IIASA Agro-ecological Zone (AEZ) model and the
IIASA recursive-dynamic, applied general equilibrium
‘global food system’ model (focusing on agriculture but
representing other sectors as well)
- Results from the food system model are ‘downscaled’ to
the spatial grid … for quantification of land cover
changes’
- Feedstock for second-generation biofuels (contribution
depends on scenario) assumed to only use ‘non-food’
land (see Fischer et al, 2009b, p168-76 for details)

 NREAP demand considered?
No, not available at the time; 10% EU biofuel target in
policy scenario (including second-generation biofuels)
translating into absolute volumes and changes as above.
Absolut magnitude lower then according to NREAP (29.6
Mtoe) but absolute increase very similar (19.5 Mtoe).
Scenario description:
- Reference scenario keeping biofuel use constant after
2008
- Several alternative scenarios including biofuel use as
predicted by the World Energy Outlook (WEO) assuming
that EU and China meet announced 2020 targets, while
USA lag in meeting EISA mandates
- More ambitious biofuels target scenarios (TAR) assuming
that mandatory, voluntary and indicative targets for
biofuel use in major developed and developing countries
are met, with a resulting biofuel use about double and
reaching a ~7% share in final consumption of total
transport fuels globally (see Fischer et al, 2009, p15 for
detailed summary of scenarios);

 Regions and sectors

 Base year – model horizon

Fischer et al (2009a, b)

Main results
Average 2013-2017 prices of existing and new policies
compared to counterfactual with no biofuel support:
- Oilseeds: 7%
- Wheat: 8%
- Coarse grains: 13%
- Vegetable oils: 35%

 Reporting of EU versus global price impacts?
No, only world prices
 Differentiated reporting of commodity versus retail
price impacts?
Seem to be on commodity level effects
 Comparator for price effects
Policy 2020 (2030) compared to reference 2020 (2030)
Price increases compared to reference scenario
(according to Fig3, p18):
- WEO scenario without second-generation biofuels
(WEO-V2): increase in cereal prices in 2020 ~ 12%, other
crops ~9%
- Biofuels target scenario (TAR-V1): increase in cereal
prices in 2020 ~ 35%, other crops ~27%
- Scenario with rapid cellulosic ethanol deployment (TAR-

38

Approximate figures based on information read from Figures 2.1 and 2.2 in OECD (2008, p65), reporting changes in biofuel use in the counterfactual scenario compared
to the baseline as well as Figures 2.7 and 2.8 (p71) , reporting changes in biofuel use in the policy scenario compared to the baseline. Conversion from litres to Mtoe
based on lower heating values (LHV) of 21.1 Megajoule per litre (MJ/l) for ethanol and 33.3 MJ/l for biodiesel.

39

As before.
70

Study and model used
34 regions and 10 sectors (nine main agricultural sectors,
one aggregated sector that represents all remaining
activities) (for details see Fischer et al, 2009b, p134)
 Modelling land
AEZ model represents land resources availability taking
into account suitability and productivity parameters. Land
conversion derived from land demand from world food
system model together with constraints and criteria as
contained in AEZ model (p136)
 Modelling by-products
Included

Blanco Fonseca et al (2010)
Short model description:
integrated Agro-economic Modelling Platform (iMAP), a
platform for the three partial-equilibrium, agro-economic
models AGLINK-COSIMO, ESIM and CAPRI

Scope and scenarios (geo – time – research questions)
2000; to 2020 and 2030
 Absolute magnitudes of biofuel use (EU & global)
EU biofuel use not specified, but underlying WEO
projection is 25.9 Mtoe in 2020, thus falling short of
NREAP prediction; global use in 2020: 94 = 63 + 31 Mtoe
(developed world + developing world)
 Increase in magnitudes in scenarios
Not specified; estimated to be 20.4 Mtoe for the EU and
69.6 globally40
 MS NREAP considered?
10% target part of biofuels target scenario, corresponding
biofuel volume see above, falls short of NREAP projection,
but absolute increase in line.
Scenario description:
- Baseline scenario: meeting 10% biofuels target in 2020,
with both first- and second-generation biofuels in the
ratio 70:30
- Counterfactual scenario: no mandatory biofuel targets
nor tax exemptions or other fiscal stimuli
- Both scenarios assume that ‘all countries outside the EU
continue with their biofuel policies as already either
implemented or announced at the start of 2009’ (p.vii)
- ESIM: 7% target in 2020 met by first-generation

AGLINK-COSIMO: recursive-dynamic partial equilibrium
model; includes first and second-generation biofuels, the
latter exogenously included and assuming their
production has no impacts on agricultural markets and
land use
 Base year – model horizon
AGLINK-COSIMO model version 2009, ESIM: base year
ESIM: comparative static partial equilibrium model 2005; projection to 2020
including first-generation biofuels
 Absolute magnitudes of biofuel use (EU & global)41

Main results
V3): increase in cereal prices in 2020 ~ 15%, other crops
~19%
Price increases for coarse grains are largest due to maize
ethanol production (Fig 4, p19). Large price drops for
protein feed due to by-products.

 Reporting of EU versus global price impacts?
Only world prices reported
 Differentiated reporting of commodity versus retail
price impacts?
No, only commodity level reported
 Comparator for price effects
2020 baseline compared to 2020 counterfactual
AGLINK-COSIMO:
Based on Fig 3.2 and 3.3: very limited impacts on prices
for ethanol feedstocks, more substantial for vegetable
oils (no % change given, roughly in the order of +15% in
the baseline compared to counterfactual based on
inspecting Fig 3.3, p33)

40

The reference assumes that biofuel use is frozen at 2008 levels. The authors cite the WEO 2008 to inform their scenario projections. The increases reported here are
changes from biofuel use in 2006 to 2020, both according to WEO 2008. Given that biofuel use in 2008 exceeded use in 2006, these figures are overestimates (in the case
of the EU, use in 2008 was 9.5 Mtoe according to http://www.eurobserv-er.org/pdf/biofuels_2011.pdf).

41

The figures on absolute magnitudes are valid for AGLINK-COSIMO and CAPRI but some deviation with regard to biofuel use in the counterfactual scenario in ESIM, where
biofuel use does not decrease from 2009 to 2020 but slightly increases (see Blanco Fonseca, 2010, p68).
71

Study and model used
CAPRI: comparative-static, spatial, partial equilibrium
model; biofuels not modelled as separate sectors; instead
exogenous quantities of biofuels (from AGLINK) that are
linked to corresponding cereals and vegetable oil
feedstock
requirements;
CAPRI
then
models
consequences for by-products and agricultural markets
 Regions and sectors
AGLINK-COSIMO: 39 individual primary and processed
products; 52 countries and regions
ESIM: 43 individual primary and processed products; EU27
represented in detail, all other countries apart from USA
and Turkey aggregated to ‘rest of the World’ region
CAPRI: 47 individual primary and
processed products; 28 regions
 Modelling land
It is mentioned that land use change is modelled for
‘selected countries only’ in AGLINK-COSIMO and modelled
at ‘Member State level’ in ESIM; in CAPRI land use change
is modelled at NUTS 2 level within the EU but not outside
of EU
 Modelling by-products
Included in all (but based on exogenous biofuel demand in
CAPRI)
Taheripour et al (2010)

Scope and scenarios (geo – time – research questions)
In baseline scenario, EU use in 2020 is around 33.0 Mtoe,
global use around 124.3 Mtoe (see below)
 Increase in magnitudes in scenarios
Increase from counterfactual to baseline 2020 use: 28.1
Mtoe in the EU and 23 Mtoe globally
Increase from baseline 2008-2020 use in the EU is 22.4
Mtoe, so close to NREAPs
 MS NREAP considered?
AGLINK-COSIMO: 8.5% ethanol & biodiesel out of
conventional petrol/diesel in 2020 (out of which 7% firstand 1.5% second-generation biofuels), yielding a 2020
biofuel demand of around 33 Mtoe, hence close to
NREAPs42

Main results
ESIM:
Increases in wholesale prices in 2020 baseline compared
to 2020 counterfactual (no biofuel support) (Table 4.1,
p71):
- Wheat, maize, sugar: around +8, 22, 21%, respectively
- Soybeans, rapeseeds, sunflower seeds: around +0.5, 10,
11%, respectively
- Vegetable oils (palm, soy, rape, sunflower): around +1,
17, 35, 36%, respectively
CAPRI:
Price increases of selected crops in 2020 baseline
compared to 2020 counterfactual (Table 5.5, p83):
- Wheat, maize, sugar beet: +12.5, 8.0, 2.0%
- Oilseeds: +19.5%
- Vegetable oil: +27.1%

Scenario description:
 Reporting of EU versus global price impacts?
- Simultaneous 2015 EU and US biofuel mandates, ie 15 US, EU and Brazil price effects reported
Short model description:
billion gallons of ethanol use in the USA by 2015 and a  Differentiated reporting of commodity versus retail
CGE model based on GTAP-E with agro-ecological zones 6.25% biofuels share in transport in the EU by 2015 (as in
price impacts?

42

Based on total demand of 49,435 million litres (from Table 3.5, p34) and lower heating values (LHV) of 21.1 MJ/l for ethanol and 33.3 MJ/l for biodiesel. The resulting
split is 10.7 Mtoe ethanol and 22.4 Mtoe biodiesel. Global data derived from Tables 3.12 and 3.15 (p39 and p44, respectively).

72

Study and model used
Scope and scenarios (geo – time – research questions)
(AEZ) module but here in the GTAP-BIO (Taheripour et al, Hertel et al, 2010b).
2008) version including biofuel sectors and by-products
- Separated scenarios with and without by-products.
- Historical data validation for 2001-2006 so that the base
 Regions and sectors
year becomes 2006.
‘28 sectors/industries, 30 commodities, and 18 regions
comprising the major biofuel producers (including US, EU,  Base year – model horizon
and Brazil) as well as non-biofuel producers’ (p279)
2006 (updated from 2001 from GTAP 6 database) to 2015
 Modelling land
 Absolute magnitudes of biofuel use (EU & global)
Competition for land between sectors modelled by Not specified
including GTAP-AEZ land use module (Lee et al, 2005) that  Increase in magnitudes in scenarios
divides land use into 18 agro-ecological zones (AEZ) that Not specified
with similar climate, precipitation and moisture conditions  NREAP demand considered?
 Modelling by-products
No, not available at the time and model horizon only up to
Yes; it is the main contribution of the paper to augment 2015
the GTAP-BIO database to include by-products from
biofuel production (DDGS and soil and oilseed meals)
Timilsina et al (2010)
Scenario description:
- Announced targets (AT) scenario: implementation of
Short model description:
biofuel use targets globally consistent with what
- Global dynamic computable general equilibrium model countries already have announced
augmented with an explicit land allocation module
- ET: a doubling of the announced targets while the timing
- Separate biofuel production sectors (three types of of their implementation is unchanged (eg 20% biofuel
ethanol, one for biodiesel)
share in the EU by 2020)
- Authors highlight higher Armington substitution
elasticities between imported and domestic crops than  Base year – model horizon
in Keeney and Hertel (2008)
2009 (updated based on historical data from GTAP 7
- No second-generation biofuels included
database with 2004 data) to 2020
- Highlight US biofuel trade policy in line with existing  Absolute magnitudes of biofuel use (EU & global)
tariff regime modelled
Not specified
 Regions and sectors
 Increase in magnitudes in scenarios
25 regions and 26 sectors (not specified but as derived Not specified
from figures and tables)
 NREAP demand considered?
 Modelling land
No, but 10% target modelled for the EU in the AT scenario
Land allocation module including AEZs and nested CET (corresponding volume of biofuel use not specified)
function or land allocation across land covers and uses
73

Main results
Results for processed food and feed sectors reported
(small changes within and across scenarios)
 Comparator for price effects
Price changes are reported for 2006-2015
Some examples of changes in EU prices over 2006-2015
(see table 4, p283, for details), in the model run without
and with by-products, respectively:
Cereal grains: 11.0 and 5.6%
Other grains: 11.9 and 8.7%
Oilseeds: 31.6 and 26.9%
Sugarcane: 10.7 and 7.8%
Price increases in USA and to a lesser extent in Brazil also
significantly affected by inclusion of by-products.
 Reporting of EU versus global price impacts?
No, only world prices reported
 Differentiated reporting of commodity versus retail
price impacts?
Commodity prices but also effects reported for ‘Processed
food sector’
 Comparator for price effects
Policy 2020 compared to baseline 2020
Important when assessing price effects: already a
considerable biofuel penetration in the baseline, 5.4%
share globally rising to 9% in the AT scenario. This
together with high oil prices leads to significant increases
in crop prices over the baseline.
Changes in world crop prices (first figure for AT, second
figure for ET scenario):
- Sugarcane/-beet: 9.2%, 11.6%
- Corn: 1.1%, 3.7%

Study and model used
following earlier GTAP based studies (see Timilsina et al,
2010, Appendix B for details)
 Modelling by-products
Not specified whether included

Laborde (2011)
Short model description:
Updated version of the global CGE model MIRAGE-Biof
(Valin et al, 2010), including ethanol and biodiesel
production factors based on a range of feedstocks singled
out from the GTAP database (more detailed model
description in Al-Riffai et al, 2010)

Scope and scenarios (geo – time – research questions)

Scenario description:
- Baseline scenario keeps biofuel blending constant at the
2008 rate of 3.3% over the projection period to 2020
(also implying a constant ratio of 83/17 biodiesel/ethanol
share, leading up to 11.7 Mtoe EU biofuel consumption
in 2020 up from 10.2 Mtoe in 2020
- Policy scenario including an EU first-generation biofuel
consumption of 27.2 Mtoe in 2020, in line with the
NREAPs, taking into account the NREAPs predicted
biodiesel/ethanol ratio of 72/28 (translating into a share
of 8.6%).
- Further distinction between a ‘trade policy status quo’
scenario (BAU) and a free trade in biofuels scenario (FT)

 Regions and sectors
43 sectors and 11 regions (database was compiled on a
more detailed level of 57 sectors and 35 regions, including
23 sectors newly disaggregated from GTAP database) (see
Al-Riffai et al, 2010, p39 for details)
 Modelling land
Previous Al-Riffai et al (2010) study modelled a 5.6% EU
Land resources are differentiated according to agro- biofuel share in their central scenario.
environmental zones (AEZ). Land allocation between
different covers and uses via nested CET function
 Base year – model horizon
 Modelling by-products
Reports development over 2008-2020 but uses GTAP 7
Included
database with 2004 as base year
 Absolute magnitudes of biofuel use (EU & global)
27.2 Mtoe first generation land-using biofuels in 2020
policy scenario; world biofuel production is 111.2 Mtoe in
Al-Riffai et al (2010)
 Increase in magnitudes in scenarios
Additional EU consumption of 15.5 Mtoe (ie change from
baseline 2020 to policy 2020; 17 Mtoe increase from
baseline 2008 to policy 2020); increase in world
production of 7.4 Mtoe from reference to policy
74

Main results
- Oilseeds: 1.5%, 3.1%
- Wheat: 1.1% 2.4%
- Other cereal grains: 1.0%, 2.3%
- Paddy rice: 0.8%, 1.6%
- Vegetables, fruit 0.7%, 1.5%
Livestock: 0.4%, 1.1%
 Reporting of EU versus global price impacts?
Results on price effects received from the authors split
according to regions
 Differentiated reporting of commodity versus retail
price impacts?
‘Other food’ sector included in data received from
authors)
 Comparator for price effects
Policy 2020 compared to baseline 2020
Study for the European Commission focusing in ILUC
impacts, but results for price effects received from the
authors, see Tables 6 and 7.
Previous study by Al-Riffai et al (2010) on price effects (no
figures or tables in the report):
‘The model simulations show that the effect of EU biofuels
policies on food prices will remain very limited, with a
maximum price change on the food bundle of +0.5% in
Brazil and +0.14% in Europe’ (p12)

Study and model used

Scope and scenarios (geo – time – research questions)
scenario43
 NREAP demand considered?
Yes, this is one of the major changes to the 2010 study
Davies (2012)
Scenario description:
- Baseline scenario incl existing biofuel support measures,
Short model description:
this is the 2011 Agricultural Outlook projection
- Two separate scenarios: 1) Removing EU biofuel support
Der analysis based on the OECD’s AGLINK-COSIMO model, by abolishing import tariffs and tax incentives from 2011
partial equilibrium modelling system underlying the onwards and 2) Removing US biofuel support by
Agricultural Outlook. Baseline is the 2011 Agricultural abolishing import tariffs and tax credits for bioethanol
Outlook projections.
from 2011 onwards and phasing out the Renewable Fuel
Standard’s quantitative mandate for bioethanol
 Regions and sectors
Covers the global agricultural sector
 Modelling land
Not specified
 Modelling by-products
Not specified

 Base year – model horizon
2011-2020
 Absolute magnitudes of biofuel use (EU & global)
Not specified
 Increase in magnitudes in scenarios
Not specified
 NREAP demand considered?
Cannot be determine given the information on absolute
volumes is not available

Main results

 Reporting of EU versus global price impacts?
Yes, for the scenario that removes EU biofuel support
 Differentiated reporting of commodity versus retail
price impacts?
No, only commodity level
 Comparator for price effects
Comparison of average 2011-2020 prices in different
scenarios
Average 2011-2020 prices in removing EU support
compared to baseline with existing support (numbers
refer to EU and world price impacts, respectively):
- Oilseeds: -4, -2%
- Wheat: -7, -3%
- Vegetable oils: -12, -5%
- Coarse grains: NA, -2%
Average 2011-2020 world prices in removing US support
compared to baseline with existing support:
- Oilseeds: -6%
- Wheat: -5%
- Coarse grains: -12 to -14%
- Soy: -7%

Source: Own compilation based on studies as referenced.

43

Note that Laborde (2011) does not report world production or consumption of biofuels. Therefore, figures from the earlier study by Al-Riffai et al (2010) are reported
here to get an idea about the magnitudes. Note, however, that Al-Riffai et al (2010) modelled a 5.6% EU biofuel share instead of the 8.6% share in Laborde (2011).
75