Schedule of Annexes and Appendix - ClientEarth

Schedule of Annexes 

and Appendix 

The contested measure is the statutory negative reply on 20 July 2010 under Article 8(3) of 

Regulation (EC) No 1049/2001, referenced in Annex A.7, titled “19 July 2010 Letter” 

Annex 

Document 

and Accompanying Information 

Annex 

Pages 

Pages Found 

in Text 

A.1 IFPRI Terms of Reference 

Author: DG TRADE 

Addressee: IFPRI 

Total Pages: 4 (reference to pages 4 and 5) 

A.2 IFPRI Study 

Author: IFPRI 

Addressee: 

Total Pages: 123 (reference to page 13) 

A.3 2 April 2010 Application 

Author: Nusa Urbancic and others 

Addressee: Bertin Martens 

Total Pages: 2 

A.4 27 April 2010 Extension Email 

Author: Christine Moumal 

Addressee: Tim Grabiel 


A.5 8 June 2010 Confirmatory Application 

Author: Catherine Day 

Addressee: Tim Grabiel and others 


A.6 29 June 2010 Extension Letter 

Author: Marc Maes 



A.7 19 July 2010 Letter 

Author: Maria Isabel Alvarez Cuartero 

3-6 of 209 8 


130-131 of 209 9, 11, 16, 28, 

29, 32, 34 

132 of 209 9, 27 

133-146 of 209 10, 15, 28, 34 

147 of 209 10, 15, 28 

148 of 209 10



A.8 15 October 2009 Application 

Author: Nusa Urbancic 

Addressee: Hikka Summa 


A.9 3 November 2009 Extension Email 

Author: Vassiliki Anagnostou 

Addressee: Nusa Urbancic 


A.10 27 November 2009 Refusal Letter 

Author: Mario Milouchev 


Total Pages: 26 (reference to page 154) 

A.11 17 December 2009 Confirmatory Application 

Author: Nusa Urbancic and others 

Addressee: Catherine Day 


A.12 19 January 2010 Extension Letter 

Author: Marc Maes 



A.13 9 February 2010 Refusal Letter 

Author: Mac Maes 



A.14 22 February 2010 Partial Release Letter 




A.15 2 July 2010 Letter 





150-151 of 209 13, 15, 27 

152-177 of 209 13, 14 

178-192 of 209 14, 15 




201-209 of 209 16

ANNEX A.1

Page 3 of 209




ANNEX A.2

Page 7 of 209 

Global Trade and 

Environmental Impact Study of the 

EU Biofuels Mandate 

Final Draft Report 

March 2010 

This report has been prepared by: 

Perrihan Al-Riffai (IFPRI) 

Betina Dimaranan (IFPRI) 

David Laborde (IFPRI) 

ATLASS Consortium 

Specific Contract No SI2.537.787 

implementing Framework Contract No TRADE/07/A2 

1


DISCLAIMER 

This study was carried out by the International Food Policy Institute 

(IFPRI) for the Directorate General for Trade of the European 

Commission and has not been peer-reviewed. The views expressed in 

this document are the authors' and do not necessarily reflect those of 

the European Commission or IFPRI. 

2


Table of Contents 

Executive Summary ................................................................................................................................. 9 

1 Introduction ................................................................................................................................... 13 

2 Review of Recent Studies .............................................................................................................. 16 

2.1 Impact on Production, Prices, Trade ..................................................................................... 16 

2.2 Modeling Bioenergy .............................................................................................................. 17 

2.3 Land Use Modeling ................................................................................................................ 19 

3 Data and Methodology .................................................................................................................. 26 

3.1 Global Data Base ................................................................................................................... 26 

3.2 Global Model ......................................................................................................................... 28 

3.2.1 Standard MIRAGE Model ............................................................................................... 29 

3.2.2 Energy Modeling ............................................................................................................ 31 

3.2.3 Fertilizer modeling ......................................................................................................... 32 

3.2.4 Modeling of Co-products and Livestock Sectors ........................................................... 33 

3.3 Land Use Module ................................................................................................................... 35 

3.4 GHG Emissions and Marginal ILUC Measurement ................................................................ 36 

4 Baseline, Trade Policy Scenarios, and Sensitivity Analysis ............................................................ 38 

4.1 Sectoral and Regional Nomenclature .................................................................................... 38 

4.2 Baseline Scenario ................................................................................................................... 39 

4.2.1 Macroeconomic Trends ................................................................................................. 39 

4.2.2 Technology .................................................................................................................... 40 

4.2.3 Trade Policy Assumptions .............................................................................................. 40 

4.2.4 Agricultural and Agri-Energy Policies ............................................................................ 41 

4.2.5 Other Baseline Evolutions ............................................................................................. 43 

4.3 Central and Alternative Trade Policy Scenarios .................................................................... 44 

4.4 Sensitivity Analysis Design ..................................................................................................... 45 

4.4.1 Mandate Policy Targets ................................................................................................. 45 

4.4.2 Parameter Uncertainties ............................................................................................... 46 

5 Results and Discussion .................................................................................................................. 48 

5.1 Production and Trade Impact of Trade Scenarios ................................................................. 48 

5.1.1 Biofuel Production and Imports .................................................................................... 49 

5.1.2 Agricultural Production ................................................................................................. 51 

5.1.3 Fuel and/or Feed? ......................................................................................................... 55 

5.2 Land Use Effects .................................................................................................................... 57 

5.2.1 Land use ......................................................................................................................... 57 

3


5.2.2 Emissions ....................................................................................................................... 60 

5.2.3 Crop specific ILUC .......................................................................................................... 63 

5.3 Sensitivity Analysis ................................................................................................................ 65 

5.3.1 Alternative Mandate Targets ........................................................................................ 65 

5.3.2 Land substitution ........................................................................................................... 68 

5.3.3 Land extension .............................................................................................................. 68 

6 Concluding Remarks ...................................................................................................................... 70 

6.1 Lessons Learned .................................................................................................................... 70 

6.2 Suggestions for Further Research ......................................................................................... 71 

7 ANNEXES ........................................................................................................................................ 73 

8 REFERENCES ................................................................................................................................ 119 

LIST of ANNEXES 

Annex I. Construction of the Global Biofuels Database ........................................................................ 73 

Annex II. Modeling Energy and Agricultural Processes of Production .................................................. 78 

Annex III. Final Consumer Energy Demand ........................................................................................... 84 

Annex IV. Fertilizer Modeling ................................................................................................................ 86 

Annex V. Modeling of Co-Products of Ethanol and Biodiesel ............................................................... 88 

Annex VI. Modeling Land Use Expansion .............................................................................................. 90 

Annex VII. Measurement of Marginal Indirect Land Use Change ....................................................... 102 

Annex VIII: The Role of Technology Pathway ...................................................................................... 105 

Annex IX: The Role of Land Extension Coefficients ............................................................................. 107 

Annex X. Biofuels Policies .................................................................................................................... 108 

LIST of TABLES 

Table 1 Regional Aggregation ................................................................................................................ 38 

Table 2. Sectoral Aggregation ............................................................................................................... 39 

Table 3 Level and variation of biofuels production (Mio toe and %) .................................................... 49 

Table 4. Level and Variation of EU Biofuel Imports, by Origin (Mio toe and %) by 2020 ...................... 50 

Table 5. Main Changes in Crop Production (non EU27) in 2020, 1000t ................................................ 52 

Table 6. Real Income Impact of European Biofuel Policies, 2020 (Variation / Baseline) ...................... 55 

Table 7. Variation of Total Land Used (thousands of km²) .................................................................... 59 

Table 8 Decomposition of production increase .................................................................................... 60 

Table 9. Indirect land use emissions related to biofuels in 2020 .......................................................... 61 

Table 10 Emissions balance. Annualized figures. CO2 Mto2 eq. ........................................................... 62 

Table 11. Carbon balance sheet ............................................................................................................ 63 

Table 12 Marginal Indirect Land Use emissions, gCO2/MJ per annum. 20 years life cycle. ................. 64 

Table 13 Marginal Net Emissions by Feedstock. gCO2/Mj. 20 years life cycle. .................................... 65 

Table 14 Protein Content of Oil Cakes used for the Modeling .............................................................. 88 

4


Table 15 Energy Content of Feed for Livestock - Metabolizable Energy ............................................... 88 

Table 16 Share of Land Available for Rainfed Crop Cultivation Computed for the MIRAGE Model (km²) 

............................................................................................................................................................... 96 

Table 17 Number of cattle head (bovine eq) per square kilometers for main regions....................... 100 

Table 18 Reduction of CO2 associated with different feedstock – Values used in calculations ......... 106 

Table 19 Land Extension Coefficients .................................................................................................. 107 

Table 20 Biofuel Use and Mandates in the European countries (% of energy content) ..................... 112 

Table 21. Current official targets on share of biofuel in total road-fuel consumption ....................... 113 

Table 22. Diesel and Biodiesel excise taxes in the European Union ($/liter)...................................... 114 

Table 23. Gasoline and Ethanol excise taxes in the European Union ($/liter).................................... 114 

LIST of FIGURES 

Figure 1 Biofuel Feedstock Schematic ................................................................................................... 34 

Figure 2 EU biodiesel imports by source, Mtoe, in the baseline ........................................................... 41 

Figure 3 Structure of EU Biofuels Production by Feedstock (2020) ...................................................... 51 

Figure 4 Variation of EU Crop Production - 2020 - (volume and percentage) ...................................... 53 

Figure 5 Variation of agricultural value-added in 2020 (%)................................................................... 54 

Figure 6 Variation of value-added in livestock sectors in 2020 (%) – MEU_BAU scenario ................... 56 

Figure 7 Cropland Extension by Region, 2020, Km2 .............................................................................. 58 

Figure 8 Source of Cropland Extension by Type of Land ....................................................................... 58 

Figure 9 Indirect land use emissions and direct savings for different mandate levels, No change in 

trade policy ............................................................................................................................................ 66 

Figure 10 Indirect land use emissions and direct savings for different mandate levels, Free trade 

scenario ................................................................................................................................................. 66 

Figure 11. Structure of production in the GTAP-E model ..................................................................... 79 

Figure 12. Structure of the Capital & Energy Composite in the GTAP-E model .................................... 81 

Figure 13. Structure of the Production Process in Agricultural Sectors in the Revised MIRAGE Model82 

Figure 14. Demand Structure Adapted for Final Energy Consumption ................................................. 84 

Figure 15. Possible concave yield functional forms (ymax = 5) ............................................................. 87 

Figure 16 Land Available for Rainfed Cultivation in Unmanaged Land Area (in km²) ........................... 97 

Figure 17. Example of productivity distribution profile for the USA. .................................................... 98 

Figure 18 Modeling of a Marginal ILUC Shock..................................................................................... 103 

5


LIST of ACRONYMS and ABBREVIATIONS 

AEZ 

Btu 

CAP 

CARB 

CEPII 

CES 

CET 

CGE 

CIS 

CO2 

DDA 

DDGS 

DG 

EC 

EEA 

EPA 

EPA 

EU 

FAPRI 

FAO 

FQD 

GHG 

GJ 

GTAP 

HHV 

Agro-Ecological Zone 

British Thermal Unit 

Common Agricultural Policy 

California Air Resource Board 

Centre d’Etudes Prospectives et d’Informations Internationales 

Constant Elasticity of Substitution 

Constant Elasticity of Transformation 

Computable General Equilibrium 

Commonwealth of Independent States 

Carbon Dioxide 

Doha Development Agenda 

Distillers Dried Grains with Solubles 

Directorate General 

European Commission 

European Environment Agency 

Economic Partnership Agreement 

(US) Environmental Protection Agency 

European Union 

Food and Agricultural Policy Research Institute 

Food and Agriculture Organisation of the United Nations 

Fuel Quality Directive 

Greenhouse Gas 

Gigajoule 

Global Trade Analysis Project 

High Heating Value 

6


IEA 

IFPRI 

ILUC 

IMPACT 

IPCC 

JRC 

LCA 

LES 

LHV 

MFN 

MIRAGE 

MJ 

MToe 

N2O 

OECD 

PE 

RED 

RER 

RFS 

SAM 

USA 

International Energy Agency 

International Food Policy Research Institute 

Indirect Land Use Change 

International Model for Policy Analysis of Agricultural Commodities and Trade 

Intergovernmental Panel on Climate Change 

Joint Research Centre 

Life Cycle Analysis 

Linear Expenditure System 

Low Heating Value 

Most Favored Nation 

Modeling International Relationships in Applied General Equilibrium 

Megajoule 

Million Tons of Oil Equivalent 

Nitrous Oxide 

Organisation for Economic Co-operation and Development 

Partial Equilibrium 

Renewable Energy Directive 

Renewable Energy Roadmap 

Renewable Fuels Standard 

Social Accounting Matrix 

United States of America 

7


UNIT CONVERSION SYSTEM 

Ethanol 

1 US gallon = 3.78541178 liter 

Corn: 1 bushel = .0254 metric ton 

Gasoline: US gallon = 115,000 Btu = 121 MJ = 32 MJ/liter (LHV). HHV = 125,000 Btu/gallon = 132 

MJ/gallon = 35 MJ/liter 

Metric tonne gasoline = 8.53 barrels = 1356 liter = 43.5 GJ/t (LHV); 47.3 GJ/t (HHV) 

Metric tonne ethanol = 7.94 petroleum barrels = 1262 liters 

Ethanol energy content (LHV) = 11,500 Btu/lb = 75,700 Btu/gallon = 26.7 GJ/t = 21.1 MJ/liter. 

Ethanol density (average) = 0.79 g/ml ( = metric tonnes/m3) 

Biodiesel 

1 m3 de biodiesel = 0,78 tep 

Metric tonne biodiesel = 37.8 GJ (33.3 - 35.7 MJ/liter) 

Petro-diesel = 130,500 Btu/gallon (34.5 MJ/liter or 42.8 GJ/t) 

Petro-diesel density (average) = 0.84 g/ml ( = metric 

tonnes/m3) 

Vegetable oil density = 0.89 kg/l 

8


Executive Summary 

Global demand for biofuels has risen sharply over the last decade, driven initially by oil price hikes 

and the need for greater energy security. Support measures were established in many countries in 

recognition of the potential of biofuel development in reducing dependence on fossil fuels, 

increasing farm revenues, and generating less environmental damage through lower greenhouse gas 

(GHG) emissions compared to non-renewable fuel sources. Over the last three years, however, 

scepticism about the positive impact of biofuels has escalated as the trade-offs between food, feed, 

and fuels and their impact on global agricultural markets became more evident, eventually leading to 

the debate over the extent of the role of biofuels in the 2007-08 food price crisis. Furthermore, 

several studies have raised serious concerns about the negative environmental impact of the 

unintended consequences of biofuel production, particularly the indirect land use change (ILUC) 

impact of releasing more carbon emissions as forests and pristine lands are converted to cropland 

due to biofuel expansion. This has led to the current debate over whether, and how, the ILUC effects 

should be accounted for, along with the direct land use change effects, in evaluating the potential 

impact of biofuel policies. 

On 23 April 2009, the European Union adopted the Renewable Energy Directive (RED) which included 

a 10% target for the use of renewable energy in road transport fuels by 2020. It also established the 

environmental sustainability criteria that biofuels consumed in the EU have to comply with. This 

includes a minimum rate of direct GHG emission savings (35% in 2009 and rising over time to 50% in 

2017) and restrictions on the types of land that may be converted to production of biofuels 

feedstock crops. The latter criterion covers direct land use changes only. The revised Fuel Quality 

Directive (FQD), adopted at the same time as the RED, includes identical sustainability criteria and 

targets a reduction in lifecycle greenhouse gas emissions from fuels consumed in the EU by 6% by 

2020. Moreover, the Parliament and Council asked the Commission to examine the question of 

indirect land use change (ILUC), including possible measures to avoid this, and report back on this 

issue by the end of 2010. In that context, the Commission launched four studies to examine ILUC 

issues, including the present study. 

The primary objective of this study is to analyse the impact of possible changes in EU biofuels trade 

policies on global agricultural production and the environmental performance of the EU biofuel 

policy as concretised in the RED. The study pays particular attention to the ILUC effects, and the 

associated emissions, of the main feedstocks used for first-generation biofuels production. 

This is the only study, out of the four launched by the Commission, that uses a global computable 

general equilibrium model (CGE) to estimate the impact of EU biofuels policies, in this case an 

9


extensively modified version of the existing MIRAGE model. Primary among major methodological 

innovations introduced in the model is the new modeling of energy demand which allows for 

substitutability between different sources of energy, including biofuels. The underlying global Global 

Trade Analysis Project (GTAP) database has been extended to separately identify ethanol (with four 

subsectors), biodiesel, five additional feedstock crops sectors, four vegetable oils sectors, fertilizers, 

and the transport fuel sectors. This extension has been introduced using innovative tools to ensure 

the consistency in both value and volume for the sectors of interests. The model was also modified to 

account for the co-products generated in the ethanol and biodiesel production processes and their 

role as inputs to the livestock sector. Fertilizer modeling was also introduced to allow for substitution 

with land under intensive or extensive crop production methods. Finally, another major innovation is 

the introduction of a land use module which allows for substitutability between land classes, 

classified according to agro-ecological zones (AEZs), and land extension possibilities. We assess the 

greenhouse gas emissions (focusing on CO2) associated with direct and indirect land use changes as 

generated by the model for the year 2020, and separately quantify the marginal ILUC for each 

feedstock crop. 

The modelling starts from a baseline scenario that excludes the EU biofuels policies introduced by 

the RED. In that baseline, EU biofuels consumption is kept stable between 2009 and 2020, at the 

2008 level of a 3.3% share in the mix of biofuels and fossil fuels. This baseline scenario incorporates 

the latest forecasts of energy prices by the IEA, and OECD economic growth. It also maintains the EU 

anti-dumping levy on biodiesel imports from the US. The baseline takes into account the biofuels 

mandates in other economies but we have limited this to a conservative case (5% mandates for 

China, Canada, Japan, Australia, New Zealand, Switzerland, Indonesia and Indonesia). 

We then introduce a first-generation land-using biofuels share of 5.6% in the overall EU renewable 

energy target of 10% for road transport fuels (by 2020) in a central policy scenario, and calculate the 

impact of this policy measure on agricultural production, trade, incomes and carbon emissions. The 

5.6% figure is obtained by deducting the expected share in 2020 of other renewable road transport 

fuels from the 10% target. We also examine the impact of a change in the EU biofuels trade policy 

regime, with an elimination of import tariffs, in a full multilateral scenario and in a bilateral scenario 

(with the MERCOSUR countries only). Finally, sensitivity analyses are conducted to assess the 

robustness of the model results to alternative assumptions about the size of the EU biofuels policy 

target and on several parameter settings. 

The central policy scenario translates the 5.6% first-generation biofuels mix in road transport fuels in 

2020 into an increase in biofuels consumption in the EU to 17.8 Mtoe. The required increase in 

biodiesel production is mostly domestic in the EU while the increase in bioethanol production is 

10


mostly concentrated in Brazil. It implies a considerable increase in EU imports of bioethanol, despite 

the duties. Brazil's real income increase marginally (+0.06%), and even less so for the EU; all other 

regions lose marginally. World cropland increases by 0.07%, showing that there is indeed indirect 

land use change associated with the EU biofuels mandate. Direct emission savings from biofuels are 

estimated at 18 Mt CO2, additional emissions from ILUC at 5.3 Mt CO2 (mostly in Brazil), resulting in 

a global net balance of nearly 13 Mt CO2 savings in a 20 years horizon. 

The multilateral and the bilateral trade liberalization scenarios show very similar results, primarily 

because Brazil is the main beneficiary in both scenarios. Trade opening is beneficial for the 

environment. Elimination of tariffs on biofuels imports, especially for bioethanol, leads to slightly 

higher ILUC effects because of land extension outside the EU, especially in Brazil. But direct emissions 

are reduced because production and consumption moves towards a more emission-efficient biofuel 

(sugar cane ethanol from Brazil). The emissions saving rate is improved and the overall emission 

balance is positive in terms of CO2 reduction (between 43 and 47 gCO2 saved by MJ of biofuels). This 

effect is based on the assumption that the share of ethanol in EU biofuel consumption can increase 

from 19% to the maximum level of 45% by 2020. 

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. 

The EU biofuels policy also has no significant real income consequences for the EU, though some 

countries may experience a slight decline in real income: -0.11% to -0.18% by 2020 among oil 

exporters, and -0.12% for Sub-Saharan Africa, due to a decline in fossil oil prices and a rise in food 

prices, respectively. 

Analysis of ILUC effects by crop indicates that ethanol, and particularly sugar-based ethanol, will 

generate the highest potential gains in terms of net emission savings. For biodiesel, palm oil remains 

as efficient as rapeseed oil, even if peatland emissions are taken into account. The model also 

indicates that the ILUC emission coefficients could increase with the size of the EU mandate. 

Simulations for EU biofuels consumption above 5.6% of road transport fuels show that ILUC 

emissions can rapidly increase and erode the environmental sustainability of biofuels. 

There are important uncertainties with respect to a number of behavioral parameters in the model. 

Still, the main conclusions of the study remain robust with respect to the sensitivity analyses 

performed. Yield response and land elasticities play a critical role in the assessment. We have also 

confirmed the importance of having a high quality database that links the value and the quantity 

matrix to feed the model with technically relevant marginal rates of substitution. . 

11


We conclude by emphasizing critical areas for further research to improve the evidence base for 

policy makers. 

It is important to investigate the assumptions regarding the 45%/55% ratio between biodiesel and 

bioethanol that we use in this study (as a function of vehicle fleet composition) since this strongly 

influences the results. It pushes biofuel demand towards bioethanol, where sugar ethanol provides 

important net emissions savings and accounts for the strong benefits from trade liberalization.. 

There is also a critical need to improve the overall quality of data for the EU27 SAM. Considerable 

effort was spent on correcting some inconsistencies and updating the GTAP7 database. However, the 

quality of the original EU social accounting matrix in the GTAP7 database is poor. Moving to the 

latest GTAP7.1 database (released in mid-February 2010) that includes updated EU SAMs could 

improve the analysis. 

Finally, considerable uncertainty remains regarding the impact of the sustainability criteria on 

biofuels markets. The role of certification and the emergence of differentiation in biofuels, feedstock 

crops and land prices, based on carbon content and the respect of sustainability criteria, require 

more empirical research. More research on the situation and likely evolution of the share of 

different production pathways could reduce uncertainties regarding direct emission savings. It would 

help to get a better understanding of the actual impact of the sustainability criteria in the EU RED on 

emissions and the market for biofuels. 

12


1 Introduction 

Global demand for biofuels has risen sharply over the last decade, driven initially by oil price hikes 

and the need for greater energy security. Support measures were established in many countries in 

recognition of the potential of biofuel development in reducing dependence on fossil fuels, 

increasing farm revenues, and generating less environmental damage through lower greenhouse gas 

(GHG) emissions compared to non-renewable fuel sources. Over the last three years, however, 

scepticism about the positive impact of first-generation biofuels has escalated as the trade-offs 

between food, feed, and fuels and their impact on global agricultural markets became more evident, 

eventually leading to the debate over the extent of the role of biofuels in the 2007-08 food price 

crisis. Furthermore, several studies (e.g. Fargione et al, 2008; Searchinger et al., 2008; RFA, 2008) 

have raised serious concerns about the negative environmental impact of the unintended 

consequences of first-generation biofuels that are based on feedstock fit for human food 

consumption and compete for land use with food crops. The indirect land use change (ILUC) impact 

of these biofuel feedstocks could release more carbon emissions as forests and pristine lands are 

converted to cropland due to biofuel expansion. This has led to the current debate over whether, 

and how, the ILUC effects should be accounted for, along with the direct land use change effects, in 

evaluating the potential impact of biofuel policies. 

The adoption of targets for the use of biofuels in road transport fuels is a key component of the 

European Union's response to achieving its Kyoto targets of GHG emissions. In 2003 the European 

Union first set a target of 5.75% biofuels use in all road transport fuels by the end of 2010. The 

proposal to adopt a 10% target for a combination of first and second generation biofuels use in road 

transport fuels by 2020 was made in the Renewable Energy Roadmap (CEC, 2006) as part of an 

overall binding target for renewable energy to represent 20% of the total EU energy mix by the same 

date. On 23 April 2009, the European Union adopted the Renewable Energy Directive (RED) which 

includes a 10% binding target for renewable energy use in road transport fuels and also establishes 

the environmental sustainability criteria for biofuels consumed in the EU (CEC, 2008). A minimum 

rate of GHG emission savings (35% in 2009 and rising over time to 50% in 2017), rules for calculating 

GHG impact, and restrictions on land where biofuels may be grown are part of the environmental 

sustainability scheme that biofuel production must adhere to under the RED. The revised Fuel Quality 

Directive (FQD), adopted at the same time as the RED, includes identical sustainability criteria and it 

targets a reduction in lifecycle greenhouse gas emissions from fuels consumed in the EU by 6% by 

2020. The adoption of the RED includes a requirement for the Commission to report, by 31 

December 2010, on the impact of ILUC on GHG emissions and address ways to minimize that impact. 

13


It is against this background that this study seeks to clarify the interactions between different policy 

scenarios and their potential impact on global agricultural markets and on the environment, 

particularly on GHG emissions from direct and indirect land use change. 

This study was commissioned by the Directorate General for Trade of the European Commission (DG 

TRADE). The initial objective was to examine the potential economic and environmental impact of 

various EU trade policy options with respect to biofuels. However, the model developed for this 

purpose was also a very useful contribution to the Commission's impact assessment and report on 

ILUC and to possible Commission proposals on the methodology to deal with ILUC under biofuel 

production. The objective of the study was thus expanded to analyse the global agricultural 

production, trade and environmental impact of the EU biofuel policy as concretised in the RED. The 

study pays particular attention to the ILUC effects of the main biofuel feedstocks. 

This quantitative analysis of the global economic and environmental impact of biofuel development 

is conducted using an extensively modified version of the MIRAGE global computable general 

equilibrium model (CGE) 1 . Primary among major methodological innovations introduced in the model 

is the new modeling of energy demand which allows for substitutability between different sources of 

energy, including biofuels. This is facilitated by the extension of the underlying global Global Trade 

Analysis Project (GTAP) database which separately identifies ethanol with four subsectors, biodiesel, 

five additional feedstock crops sectors, four vegetable oils sectors, fertilizers, and the transport fuel 

sectors. The model was also modified to account for the co-products generated in the ethanol and 

biodiesel production processes and their role as inputs to the livestock sector. Fertilizer modeling 

was also introduced to allow for substitution with land under intensive or extensive crop production 

methods. Finally, another major innovation is the introduction of a land use module which allows for 

substitution between land classes, classified according to Agro-Ecological Zones (AEZs), and land 

extension possibilities. We assess the greenhouse gas emissions (focusing on CO2) associated with 

direct and indirect land use changes as generated by the model for the year 2020, and separately 

quantify the marginal ILUC for each feedstock crop. 

The impact of the EU biofuels policy are assessed under alternative trade policy assumptions: 

business as usual trade policy; full multilateral trade liberalization in biofuels; and bilateral trade 

liberalization between the EU and MERCOSUR. These trade policy alternatives are calculated against 

a baseline scenario which incorporates the latest forecasts of energy prices by the IEA and OECD 

1 The MIRAGE (Modeling International Relationships in Applied General Equilibrium) model was developed at 

the Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Documentation of the standard 

model is available in Bchir et al. (2002) and Decreux and Valin (2007). Model equations for the extensively 

modified version developed at IFPRI and used in this study are provided in a separate document as Appendix A. 

14


economic growth. Sensitivity analyses are conducted to assess the robustness of the model results to 

alternative assumptions about the size of the EU biofuels policy target, and on several parameter 

settings. 

A brief review of previous studies that have quantified the potential economic and environmental 

impact of biofuel development is provided in the next section of the report. Section 3 includes an 

overview of the data development and model development involved in the study. More detailed 

discussions of the various components of the methodology are relegated to annexes. The baseline 

scenario and alternative trade policy scenarios analyzed in the study, along with the variations 

considered for sensitivity analyses, are presented in Section 4. Results and discussions are provided 

in Section 5 and concluding remarks are given in Section 6. 

15


2 Review of Recent Studies 

Although it is a relatively new area of study, research on the impact of policies to promote 

biofuels has been particularly intense in recent years. The growing literature reflects the evolution of 

issues regarding the impact of biofuel development and support policies on agricultural markets and 

the environment. Most quantitative assessments focused initially on the impact of biofuels on 

agricultural markets and its contribution to the food price crisis, but more recent studies have 

centered on the impact of biofuels on global land use change and greenhouse gas emissions. In this 

section of the report, we provide a brief survey of recent studies focusing on the quantitative 

assessments of the impact of biofuel support policies on global trade and the environment, 

specifically on land use and GHG emissions. 

2.1 Impact on Production, Prices, Trade 

Much of the early literature on biofuels came out in the mid 2000s and emphasized the potential 

benefits of biofuels development in reducing dependence of fossil fuels, providing opportunities for 

agricultural and rural development, and reducing environmental damage due to lower greenhouse 

gas emissions compared to biofuels. However, concerns about the impact on food security quickly 

emerged due to the rising competition between biofuel feedstock crops, food crops, and feed crops, 

thereby giving rise to the debate on food versus fuel. 

In their review of early work on the economic, environmental, and policy aspects of biofuels, 

Rajagopal and Zilberman (2007) found that the current generation of biofuels from food crops is 

intensive in land, water, energy and chemical inputs. The authors' synthesis of economic studies 

revealed that most models predict that biofuels development will result in higher food prices, a 

decline in cereal exports of the United States and European Union, a decline in farm support 

programs, an increase in rural jobs, and an ambiguous effect on the livestock sector. 

The impact of biofuels on food prices became a hotly debated issue during the food price crisis of 

2008. Several researchers sought to quantify the impact of biofuel polices on food prices. Studies 

range from back of the envelope calculations, such as those from the JRC (de Santi, 2008) and the 

World Bank (Mitchell, 2008) to extensive modeling exercises. In assigning the largest proportion of 

the blame on biofuels, Mitchell (2008) concluded that higher energy costs and exchange rate changes 

contributed between 25-30% of the rise in food prices, while the other 70-75% was due to biofuels 

along with the associated low grain stocks, large land use shifts, speculative activity and export bans. 

Although results vary, there is broad agreement from these studies that the price increases are due 

to several factors including but by no means restricted to biofuels (Sheeran, 2008, Von Braun, 2008). 

16


In a partial equilibrium exercise using IFPRI's IMPACT model, Rosegrant (2008) also addressed the 

question of the extent to which biofuel production contributed to the high food prices in 2008. Based 

on a comparison of the simulations of market developments between 2000-2007, with and without 

the sudden increase in biofuel production, Rosegrant estimated that biofuel growth accounted for 

30% of the food price increases seen in the period. The level varied from 39% for maize to 21% for 

rice. A simulation of the future impact of freezing biofuel production at 2007 levels indicated that 

maize prices are likely to decline by 6% in 2020 and 14% in 2015. 

Based on their review of 25 studies, Abbott et al. (2008) identified three broad sets of forces that 

drove up food prices in 2008, namely: the global changes in production and consumption of key 

commodities, the depreciation of the dollar, and the growth in the production of biofuels. Even in 

their follow-up study after the financial crisis, Abbott et al (2009) found that the key drivers of food 

prices remain the same: crop supply and utilization, the exchange rate and world macroeconomic 

factors and the agricultural-energy linkage through the biofuels market. 

Similarly, in their synthesis of several studies that assessed the impact of biofuels development on 

current and projected food prices, Gerber et al. (2009) found that although there are considerable 

differences in the methodology and assumptions, and thus in the projection results, the studies 

predict some common trends: the EU and US biofuel programs are expected to raise prices of 

vegetable oils the most with smaller price increases for corn, wheat and soybean and price declines 

for oilseed meals. 

2.2 Modeling Bioenergy 

The economic studies that assess the impact of biofuel production are based either on partial 

equilibrium (PE) or computable general equilibrium (CGE) models. These models explain the 

relationship between supply, demand, and prices through market clearance using a system of 

equilibrium equations. In partial equilibrium models, clearance in the market of a specific good or 

sector is obtained assuming that prices and quantities in other markets remain constant, thus 

providing better indication of short term response to shocks. PE models often provide a detailed 

description of the specific sector of interest but do not account for the impact of expansion in that 

sector on other sectors of the economy. Several examples of partial equilibrium models used in the 

assessment of the impact of biofuel development include AGLINK/COSIMO, ESIM, FAPRI, and the 

IMPACT model. Witzke et al. (2008) provide a review of the methodologies in modeling energy crops 

in agricultural sector PE models. 

CGE models determine equilibrium by simultaneously taking into account the linkages between all 

sectors in the economy. The modeling framework provides an understanding of the impact of 

17


biofuels on the overall economy by accounting for all the feedback mechanisms between biofuels 

and other markets, and by capturing factor market impact. As the Gallagher review (RFA, 2008) 

points out, CGE models provide a better global assessment, taking linkages in the economy into 

account and predicting outcomes that are more representative of medium and long term impact. 

Since the present employs a global CGE model, this section focuses on a review of bioenergy 

modeling in CGE models. 

Kretschmer et al. (2008) classified CGE studies according to three different categories based on the 

approach used in integrating bioenergy in CGE models. The implicit approach avoids explicit 

modeling of bioenergy production technology and employs an ad-hoc procedure of determining the 

quantities of biomass necessary to achieve certain production targets. Classified under this category 

is the study of the economy-wide effects of replacing petroleum with biomass in the US using (Dixon 

et al. 2007)USAGE, a dynamic CGE model of the US economy. Banse et el. (2008) also used a implicit 

approach in introducing biofuels in their extended version of the GTAP-E CGE model. Ethanol is 

introduced in a ‘Fuel’ nesting, substituting with vegetable oil, oil, and petroleum products. It is 

produced only from crop inputs (sugarcane\beet and cereals) thereby capturing only a part of 

ethanol production technology. 

The second category identified by Kretschmer et al. (2008) is the latent technology approach that 

focuses on production technologies that are existent but not active during the base year of the 

model but can become active at a later stage. Information on the inputs and costs structures of the 

different types of biofuels are required in modeling latent technologies. Reilly and Paltsev (2007) and 

Gurgel, Reilly and Paltsev et al. (2007) employed this approach in looking at the potential land use 

implications of a global biofuels development focusing on second generation biofuels . Boeters et al. 

(2008) and Kretschmer et al. (2008) both incorporate the European emissions trading scheme (ETS) in 

assessing the impact of a 10 percent EU biofuels target. 

This study falls under the third category of CGE biofuel studies identified by Kretschmer et al. (2008) 

which are studies that actually disaggregate the bioenergy production sectors in the social 

accounting matrix (SAM) of the GTAP database, which provides the underlying structure to global 

CGE models. Since bioenergy sectors are not explicitly identified in the GTAP database, Taheripour et 

al. (2007) introduced ethanol (from aggregated coarse grains and sugarcane) and biodiesel from an 

aggregated oilseeds sector. External data on production, cost structure and trade are used to extract 

these bioenergy sectors from existing food processing sectors in the 2001 GTAP 6 database. 

Bioenergy is modeled through an extended version of the GTAP-E model (Burniaux and Truong, 

2002). The applications of this approach include Birur, Hertel, and Tyner (2008) looking at the impact 

of biofuels production on the global agricultural market; Hertel, Tyner, and Birur (2008) looking at 

18


the impact of both US and EU biofuel support policies; and Taheripour et al. (2008) which compares 

the impact of adding by-products to the results of Hertel et al (2008). 

In a recent study, Britz and Hertel (2009) linked the European CAPRI PE model with the GTAP CGE 

model to look at the impact of the EU biofuels directive on global markets and on the environment. 

Starting from a modified GTAP model that includes a ‘parsimonious summary’ of the regional supply 

models of CAPRI, the authors then take the resulting equilibrium price changes from the global 

model and apply them to the supply models of CAPRI to obtain highly disaggregated results in terms 

of changes in farming practice and their impact in the EU. 

The production of biofuels results in several by-products which have potential or existing markets. 

Producing ethanol from corn results in a by-product – Dried Distillers Grains with Solubles (DDGS) 

which is used as animal feed. Its sale represents 16% of ethanol revenues in the US (Hertel et al., 

2008). Biodiesel production from vegetable oil produces seed meals which can be used as animal 

feed. Farrell et al. (2006) pointed out the importance of integrating by-products in assessments of 

the energy balance of biofuels. In particular they found that studies which didn’t take by-products 

into account concluded that biofuels had a negative energy balance because they failed to take 

account of the energy use which the by-products offset. 

The increased availability of by-products also have beneficial side effects in other areas of 

agriculture. The Commission’s impact assessment of the biofuels mandate pointed out the positive 

impact on livestock production in terms of reduced prices for animal feed, with soymeal prices 

predicted to fall by 25% and rapemeal by 40% by 2020 (CEC, 2007). In a CGE assessment of the 

impact of including biofuel by-products, Taheripour et al. (2008) also found significant differences in 

feedstock output and prices depending on whether the existence of by-products is taken into 

account. In terms on the land use impact of accounting for by-products, Kampman et al (2008) 

estimated that incorporating by-products into the calculations for land requirements of biofuels 

reduced the land demand by 10-25%. Croezen and Brouwer (2008) found that scenarios which 

include 2G biofuels resulted in substantial reductions of almost half in the amount of avoided land 

use. It is clear that the integration of by-products is key to properly estimating changes in prices and 

land use, as well as energy balance. 

2.3 Land Use Modeling 

Although extensive research and literature exists about local drivers of land use changes concerning 

deforestation processes, arable land conversion, pasture expansion, and the associated 

methodological challenges and development of land-use indicators, the boom in biofuel production 

19


is a recent phenomenon and as such has not yet been included as a factor driving land-use change 

(Gnansounou and Panichelli , 2008). 

It is clear however, that increased demand for biofuels will have impact on the demand for land and 

will result in potentially significant land use changes. The increased demand for land for biofuels is 

estimated to be lower than the increased demand for land for food, however estimates vary. Based 

on their review of the literature, Kampman et al. (2008) estimated that land for food and feed will 

expand between 200-500 Mha by 2020, whereas increased demand for biofuels could result in total 

demand of between 73-276 Mha (up from 13.8 Mha today). Eickhout et al. (2008) estimated the land 

requirements of the EU’s mandate alone as being between 20-30 Mha. There are high levels of 

uncertainty in these estimates as much depends on development of demand, but also on the extent 

to which high yield crops (such as sugar cane) are used, the share of second generation biofuels (land 

demand is 30-40% less in scenarios with 2G biofuels) and on crop yield. The FAO reported estimates 

of the difference between the land required for different sources of first generation biofuels if they 

were to replace 25% of global transport needs. This varies from 17% of available land (estimated at 

2.5 bn ha) if the source were sugar cane to 200% for soybean (FAO, 2008a). 

Many CGE models use the constant elasticity of transformation (CET) approach to capture the 

conversion of land to other uses due to the expansion of bioenergy production. Under the CET, 

different types of land can be transformed to other uses with the ease of transformation determined 

by the elasticity of transformation. Using the WorldScan CGE model to assess the impact of the 10% 

EU biofuels target, Boeters et al. (2008) used the CET framework to allow for transformation of 

different types of arable land use. The authors assess the sensitivity of the elasticity value of the CET 

by allowing for lower and higher-end values of 0.5 and 15, respectively, aside from the default value 

of 2. They found that their results for arable land rents and economic welfare are quite robust to the 

value to the CET. 

Banse et al. (2008) also used the CET approach but employed a 3-level CET nesting structure that 

allows for different degrees of land use transformation across types of land use. The third level nest 

distinguishes between land in wheat, coarse grains, and oilseeds. This aggregate is distinguished from 

land in sugar and pasture in the second level nest. Together, as Field Crops\Pasture, they are 

distinguished from Horticulture and Other Crops at the top-level nest. The authors also introduced a 

land supply curve which allows for endogenous processes of land conversion and land abandonment. 

In their analysis of the impact of implementing biofuels mandate on a global scale, the authors found 

that compared to a reference scenario of trade liberalization, the EU’s land use falls by less under 

both an EU and global biofuel scenario. All other key regions expand land use under a global biofuels 

20


scenario, in particular Central and South America which increases agricultural land use by almost 10 

percentage points compared to the reference scenario. The study also shows, however, that the 

majority of the expansion in land use in most regions is due to the liberalization of trade and other 

projected changes in the world economy. These are modeled under the reference scenario, where 

global agricultural land increases by18% compared to 21% under the global biofuels scenario. 

Although biofuels contributes to greater land cultivation it is not, therefore, projected to be the 

major driver. 

Birur et al. (2008), Hertel et al. (2008) and Taheripour et al. (2008) also employ the CET approach. 

These studies use a GTAP-E model adapted to include Agro-Ecological Zones (AEZs). In other words 

they take account of the fact that land types differ and substitutability is only possible within limited 

zones. Hertel et al. (2008) find substantial impact on land use from the EU and US mandates. In the 

US, coarse grains acreage increases by 10% at the expense of other cropland, as well as pasture land 

and forests. The biggest global impact are however seen as a result of the boom in oilseeds 

production due to EU demand for biodiesel. Here increases range from 11-16% in Latin America, 14% 

in SE Asia and Africa and 40% in the EU itself. The model restricts the potential land sources of 

increased biomass production to pastureland or forests as it does not take into account idle land. 

This tends to over-estimate the impact. The largest impact are for pastureland in Brazil where the 

acreage is estimated to reduce by 11%, of which 8% is due to the EU mandate. Reductions in forestry 

cover are highest for the EU (-7%), Canada (-6%) and Africa (-3%). The model (like that used in this 

study) does not take account of the potential impact of biofuel by-products which the authors 

acknowledge to be an important limitation which overestimates the impact of the mandates on corn 

and livestock markets. Greenhouse Gas Emissions (ILUC) 

A major reason behind the adoption of biofuels is based on the assumption that they are a more 

environmentally friendly fuel source, as the GHG emissions associated with their production and use 

are lower than those associated with traditional fossil fuels. This assumption is not based on the 

GHGs impact from the use of biofuels, as the GHGs emitted from burning them are not noticeably 

different to those of other fuels. There is a reduction in certain pollutants, with a possible increase in 

others (Worldwatch Institute, 2006). Rather their advantage over fossil fuels is based on the idea that 

the production of biofuels absorbs CO2 and therefore offsets large percentages of the future 

emissions from using them. 

This assumption is far from being universally accepted. Early estimates from the International Energy 

Agency indicated that the use of biofuels resulted in net GHG savings – between 20-90% for ethanol 

from crops (with most crops in the lower levels. The higher figures are for cellulosic ethanol) and 

around 50% for biodiesel from oilseeds (IEA, 2004). 

21


Although the logic for these figures is intuitively attractive, several researchers have pointed out that 

such estimates are incomplete as several aspects of the lifecycle and indirect effects of biofuels are 

not properly taken into account. Even early, positive analysis of biofuels’ potential, such as that from 

the Worldwatch Institute warned that there were limits to its potential benefits. In particular, 

biofuels that are produced from low yielding crops, or grown on previous forested or grasslands or 

produced using large inputs of fossil fuel, could easily have a negative GHG balance (Worldwatch 

Institute, 2006). The fact that biofuel’s GHG balance varies widely depending on these factors is 

increasingly taken into account in analyses. 

A recent review conducted by the US Government Accounting Office (US GAO, 2009) found that 

although there is general consensus on the approach for measuring the direct effects of increased 

biofuels production, there is disagreement about assumptions and assessment methods for 

estimating the indirect effects of global land-use change. The twelve scientific studies that the GAO 

reviewed provided a wide range of estimates on the lifecycle GHG emissions of biofuels relative to 

fossil fuels: from a 59 percent reduction to a 93% increase in emissions for conventional corn starch 

ethanol, a 113% reduction to a 50 % increase for cellulosic ethanol, and a 41% to 95% reduction for 

biodiesel. The differences in assumptions about the agricultural and energy inputs used in biofuel 

production and how to allocate the energy used in this production to co-products, such as DDGS, 

primarily explain why large differences in the GHG emission estimates among the studies. 

One key issue is that producing biofuels requires energy and the assumptions on where that energy 

comes from can make a large difference to the calculated relative efficiency of different biofuel 

sources. Mortimer et al. (2008) note the large difference between the CO2 emissions in the 

production of corn based ethanol in the US and France (0.108 kg eq/MJ compared to 0.049 kg eq/MJ 

respectively), which is largely due to the assumption that coal is used for ethanol processing in the 

US compared to natural gas in France. Biofuels that use plant waste to fuel their processing, such as 

those based on switchgrass and sugarcane are clearly the most efficient. 

In their research for the Gallagher Review, Mortimer et al. (2008) provide estimates of the 

percentage of GHGs emissions by various sources of biofuels compared to standard fossil fuels. Their 

results are fairly consistent with other sources in highlighting the relative efficiency of Brazilian sugar 

cane (which generally uses bagasse as the fuel source) and the relative inefficiency of maize which 

the study found to be more intensive in GHG emissions than the fuels it seeks to replace. 

The above results take into account the ‘credit’ represented by the by-products of the various 

processes and the N2O emissions from the soil where the crops are grown. This latter issue is one of 

the most contentious and difficult to integrate in relation to the biofuels debate. N2O is a 

22


greenhouse gas which is far more detrimental to global warming than CO2 (296 times according to 

Mortimer at al (2008)). For this reason, although emissions are far lower by weight than CO2, they 

are potentially very damaging. 

A key input to the debate on NO2 emissions and biofuels is a paper by Crutzen et al. (2007). This 

paper claims that the manner in which the UN’s Inter-governmental Panel on Climate Change (IPCC) 

integrates N2O emissions into its assessments underestimates N2O emissions from crops by a factor 

of 3 to 5. The paper has been criticized and its accuracy called into question. Mortimer et al. (2008) 

have undertaken an exhaustive review of the paper and conclude that while it raises an important 

issue ‘…it cannot be regarded as resolving the problems and assisting the objective evaluation of 

biofuels.’ (Mortimer et al, 2008, p. 29). For the moment, as their review makes clear, it is impossible 

to accurately measure the extent of N2O emissions related to a given biofuel from a given source. 

For this reason and due to the complexities of seeking to integrate it in the model, this research does 

not seek to assess the indirect effects, related to land use, of biofuels on GHGs other than CO2. But 

direct effects related to CO2 and N2O are accounted for as they are incorporated in the coefficients. 

The other key issue which has emerged as controversial in recent months is the question of the 

‘credit’ attributable to biofuels from the ‘carbon uptake’ of the crops used to produce them. A key 

paper in this debate is that by Searchinger et al. (2008). His main point is that earlier assessments of 

the carbon impact of biofuels have been biased because they have not taken account of the land use 

impact. In short they have counted the carbon benefits of using land for biofuels but not the carbon 

costs – the carbon storage and sequestration which is sacrificed when the land is diverted from its 

former use (direct GHG effects) or when land is cleared for growing food to replace land which has 

been diverted into biofuel production (indirect GHG effects). Searchinger et al. (2008) used the 

Greenhouse gases, Regulated Emission and Energy use in Transportation (GREET) model to calculate 

the total GHG emissions from various biofuel sources. The model indicates that, without taking into 

account land use changes, replacing gasoline by corn-based ethanol reduces GHG emissions by 20% 

by 2015. Once they account for land use change, however, the picture changes significantly and they 

find that corn based ethanol more than doubles GHG emissions over a 30 year timescale and 

increases GHGs for 167 years. On the other end of the spectrum, Brazilian sugarcane production is 

estimated by their model to provide GHG savings of 86%. If this sugarcane production converts only 

tropical grassland, the payback for GHG emissions would be only 4 years, although this would rise to 

45 years if displaced ranches were to convert forest to grazing land. 

In their review of the Searchinger paper and the GREET model to assess its applicability to the EU/UK 

context, Mortimer et al. (2008) concluded that the model is too US specific to be readily useable 

outside that context. The US Department of Energy has itself issued a rebuke criticizing many aspects 

23


of the study, which it considers to also misrepresent the US case, by overestimating corn ethanol 

production and making several invalid assumptions (DOE, 2008). Nevertheless the key point which 

Searchinger makes – that land use changes and their impact on GHG emissions are key to assessing 

the true impact of biofuels - is a valid one which needs to be taken into account in analysis. The 

Gallagher review acknowledges this, particularly in its recommendation that policies should seek to 

direct biofuels production towards suitable idle land or appropriate wastes and non-food products. 

This recommendation is based on a series of calculations on the net impact of the conversion of 

various types of land on GHG emissions which concur with the broad conclusions of Searchinger’s 

paper (Mortimer et al., 2008). The analysis finds that, apart from the lowest estimate of ethanol from 

sugar beet, all current biofuel production on converted UK grasslands would increase GHG emissions, 

in some cases emitting twice the level of fossil fuels. The figures calculated for biofuels from overseas 

sources are even worse. Of all sources analysed - oil palm in Malaysia, soy biodiesel in Brazil, maize 

ethanol in the US and sugar cane ethanol in Brazil - only the latter showed a net saving and the 

others showed large net losses, topping 30,000% for biodiesel from soy converted from Brazilian 

rainforest. 

The calculation for the impact of using fallow land is slightly different, as it assumes that the N2O 

emissions which would have been emitted by this land are avoided by its cultivation, thus adding an 

additional ‘credit’ to the calculation. The results are generally positive i.e. the production of biofuels 

in the UK from fallow land is calculated to emit less GHG than fossil fuels, although the percentage 

varies from 88-55%. The figures are similar for biodiesel and ethanol, although they tend to be lower 

for the former, especially in the long term and when rotational set-aside land is used. 

The JRC report also looked at the issue of land use change and its impact on GHG emissions (de Santi, 

2008). They made the point that looking at direct effects alone was probably legitimate when rates of 

substitution by biofuel were low and most biofuel feedstock could come from set-aside or other 

unused arable land. However the 10% target means that most of the EU biofuel feedstock will be 

removed from the world commodity markets either by reduced EU exports or increased EU imports. 

They looked at the alternative sources of these extra biofuels and in most cases found significant 

negative effects. For example using EU permanent grassland would result in an initial emission of 

carbon which would take 20 to 110 (+/- 50%) years to recover through biofuel production. The 

carbon losses from drained peat forest, which is used for palm oil production in South East Asia, are 

so high that if even 2.4% of the EU’s biodiesel needs are met directly or indirectly by palm oil grown 

in peatland all GHG savings from EU biodiesel would be cancelled out. Palm oil is a key alternative to 

rapeseed for the food industry, so EU imports are likely to increase once the latter is diverted to 

24


biofuel production. The calculations in the report indicate that the level of EU imports of palm oil 

produced on peatland is likely to be considerably higher than 2.4%. Although local regulations could 

be set in place to avoid such negative indirect effects, the report is dubious about the potential of 

certification schemes to assure sustainability. The report concludes ‘Indirect land use change could 

potentially release enough greenhouse gas to negate the savings from conventional EU biofuels.’ (De 

Santi, 2008). 

Finally a key question which is frequently ignored in the biofuels debate is whether the use of 

biomass for biofuels is the most efficient means to use the limited biomass resources at our disposal 

to reduce GHGs. A recent JRC report pointed out that while the efficiency of fuel burners for heating 

and electricity is 21 almost as high as that of fossil fuels, the energy efficiency of converting biomass 

to liquid fuels is only 30-40% (de Santi, 2008). Their cost benefit analysis indicates that the decision 

to specifically target GHG reductions in the transport sector reduces the benefits that could be 

achieved in other ways. The European Environment Agency has furthermore expressed concern that 

diversion of biomass to biofuel will make it difficult for the EU to meet its objectives for renewable 

energy sources in energy production (EEA, 2004). 

A related point is that support for biofuels is a very expensive means of reducing CO2 emissions. The 

OECD has estimated that policy support to biofuels would cost taxpayers and consumers between 

$960 and $1 700 per ton of CO2 emissions avoided (OECD, 2008). The exact figures can be debated 

as they are based on a series of assumptions and indeed are far higher than the figures used in the 

Commission’s impact assessment of the Renewable Energy Directive5 or even the high end estimates 

(over €300/ton) referred to in the Economic and Social Committee’s report (EESC, 2008). However 

the fundamental point of the OECD work – that reducing CO2 emissions through measures in support 

of biofuel production is an expensive option – is a valid one, reiterated both in that report (EESC, 

2008) and in the work of the JRC (2007). 

25


3 Data and Methodology 

The MIRAGE model 2 , a computable general equilibrium model originally developed at CEPII for trade 

policy analysis, was extensively modified at IFPRI 3 in order to address the potential economic and 

environmental impact of biofuels policies. The key adaptations to the standard model are the 

integration of two main biofuels sectors (ethanol and biodiesel) and biofuel feedstock sectors, 

improved modeling of the energy sector, the modeling of co-products and the modeling of fertilizer 

use. The land use module which includes the decomposition of land into different land uses, and the 

quantification of the environmental impact of direct and indirect land use change (ILUC), was 

introduced in the model at the Agro-Ecological Zone (AEZ) level, allowing for infra-national modeling. 

The latter feature is particularly valuable for large countries where production patterns and land 

availability are quite heterogeneous. The overall architecture of the model has been modified to 

allow for various sensitivity analyses, as well as for the computation of marginal ILUC under specific 

assumptions. The full set of model equations are provided in a separate document as Appendix A. 

Data enhancements, model modifications, and the land use module are discussed in this section of 

the report. 

3.1 Global Data Base 

The MIRAGE model relies on the Global Trade Analysis Project (GTAP) database for global, economywide 

data. The GTAP database combines domestic input-output matrices which provide details on 

the intersectoral linkages within each region, and international datasets on macroeconomic 

aggregates, bilateral trade, protection, and energy. We started from the latest available database, 

GTAP 7, which describes global economic activity for the 2004 reference year in an aggregation of 

113 regions and 57 sectors (Narayanan and Walmsley, 2008). The database was then modified to 

accommodate the sectoral changes made to the MIRAGE model. 

Twenty-three new sectors were carved out of the GTAP sector aggregates -- the liquid biofuels 

sectors (an ethanol sector with four feed-stock specific sectors, and a biodiesel sector), major 

feedstock sectors (maize, rapeseed, soybeans, sunflower, palm fruit and the related oils), co- and by- 

2 Decreux and Valin (2007). 

3 The development of the model for this study was undertaken by a joint team of IFPRI researchers and visiting 

fellow under a larger research framework including Hugo Valin (land use, biofuel mandate, co-products), 

Antoine Bouet (energy representation) and David Laborde (value chain, trade). 

26


products of distilling and crushing activities, the fertilizer sector, and the transport fuels sector. For 

the last two sectors, we split the existing GTAP sectors with the aid of the SplitCom software. 4 

However, after several tests, we found that limitations of the SplitCom software and the initial data 

lead to very unsatisfactory results in our splitting of several feedstock crops, vegetable oils, and 

biofuel sectors. We therefore developed an original and specific procedure aiming at providing a 

database that is consistent in both values and quantities: 

1. Agricultural production value and volume are targeted to match FAO statistics. A world price 

matrix for homogenous commodities was constructed in order to be consistent with 

international price distortions (transportation costs, tariffs, and export taxes or subsidies); 

2. Production technology for new crops is inherited from the parent GTAP sector and the new 

sectors are deducted from the parent ones; 

3. Vegetal oil sectors are built with a bottom-up approach based on crushing equations. Value 

and volume of both oils and meals are consistent with the prices matrix, the physical yields, 

and the inputs quantity; 

4. Biofuels sectors are built with a bottom-up approach to respect the production costs, input 

requirements, production volume, and for the different type of ethanols, the different byproducts. 

Finally, rates of profits are computed based on the difference between production 

costs, subsidies and output prices; 

5. For steps 2, 3 and 4, the value of inputs is deducted from the relevant sectors (Other Food, 

Vegetal Oils, Chemical products, Fuel) in the original SAM, allowing resources and uses to be 

extracted from different sectors if needed (mapping n to n). 

6. At each stage, consumption data are adjusted to be consistent with production and trade 

flows. 

It is important to emphasize that this procedure, even if time consuming and delicate to operate with 

so many new sectors, was crucial and differs from a more simplistic approach used in the literature 

until now. Indeed, each step allows addressing several issues. For instance, step 1 allows us to have a 

more realistic level of production than using the GTAP database that performs production targeting 

only for OECD countries, with some flaws, and therefore has an outdated agricultural production 

4 SplitCom, a Windows program developed by J. Mark Horridge of the Center for Policy Studies, Monash 

University, Australia, is specifically designed for introducing new sectors in the GTAP database by splitting 

existing sectors into two or three sectors. Users are required to supply as much available data on consumption, 

production technology, trade, and taxes either in US dollar values or as shares information for use in splitting 

an existing sector. The software allows for each GTAP sector to be split one at a time, each time creating a 

balanced and consistent database that is suitable for CGE analysis. 

27


structure for many countries. Building a consistent dataset in value and volume – thanks to the price 

matrix – is also critical. Targeting only in value often generates inconsistencies in the physical linkage 

that thereby leads to erroneous assessments (e.g. wrong yields for extracting vegetal oil). Even more 

important is the role of initial prices, and price distortions, in a modeling framework using CES and 

CET functions. Indeed, economic models rely on optimality conditions and, in our case, as in all the 

CGE literature, our modeling approach leads to equalization of the marginal rate of substitution (CES 

case) to relative prices. It means that the physical conversion ratio is bound to the relative prices. 

Wrong initial prices, or incorrect price normalization, will lead to convert X units of good i (e.g. 

imported ethanol) in Y units of good j (e.g. domestic produced ethanol). In the case of a homogenous 

good, we need to have an initial price ratio equal to one and to ensure with a high elasticity of 

substitution that this ratio will remain close to one. Otherwise, misleading results appear, e.g. one 

ton of palm oil will replace only half a ton of sunflower oil, one ton of imported ethanol can replace 

1.5 tons of domestic ethanol, etc. This mechanism may be neglected in many CGE exercises where 

the level of aggregation easily explains the imperfect substitution. In the case of this study, however, 

we found it imperative to directly address this challenge since we deal with a high level of sector 

disaggregation, a high level of substitution (among ethanols produced from different feedstocks, 

among vegetal oils, or among imported and domestic production), and with the critical role of 

physical linkages, from the crop areas to the energy content of different fuels and meals. 

Finally, a flexible procedure is needed (see 5) since some of our new sectors can be constructed from 

among several sectors in GTAP. SplitCom allows only a 1-ton disaggregation which is rather 

restrictive for the more complex configuration that we face with the data. For instance, Brazilian 

ethanol trade data falls under the beverages and tobacco sector while its production is classified 

under the chemical products sector. For the vegetal oils, we face similar issues since the value of the 

oil is in the “Vegetable Oil” sector but the value of the oil meals are generally under in the food 

products sector. 

The specific data sources, procedures and assumptions made in the construction of each new sector 

are described in Annex I. 

3.2 Global Model 

Extensive model modifications were done to adapt the MIRAGE trade policy focused CGE model for 

an assessment of the trade and environmental impact of biofuels policies. Some of the changes were 

already introduced by Bouet et al.(2008) and Valin et al.(2008). In this section, we first provide a brief 

description of the standard MIRAGE model. This is followed by the adaptations and innovations 

made in the areas of energy modeling, the modeling of co-products of ethanol and biodiesel 

28


production, and the description of fertilizer use. More detailed explanations of the various modeling 

changes are provided in the annexes. 

3.2.1 Standard MIRAGE Model 

The work starts with the MIRAGE model, initially developed at CEPII. This section summarizes the 

features of the standard version relevant for this study. MIRAGE is a multisector, multiregion 

Computable General Equilibrium Model for trade policy analysis. The model operates in a sequential 

dynamic recursive set-up: it is solved for one period, and then all variable values, determined at the 

end of a period, are used as the initial values of the next one. Macroeconomic data and social 

accounting matrixes, in particular, come from the GTAP 7 database (see Narayanan, 2008), which 

describes the world economy in 2004. From the supply side in each sector, the production function is 

a Leontief function of value-added and intermediate inputs: one output unit needs for its production 

x percent of an aggregate of productive factors (labor, unskilled and skilled; capital; land and natural 

resources) and (1 – x) percent of intermediate inputs. 5 The intermediate inputs function is an 

aggregate CES function of all goods: it means that substitutability exists between two intermediate 

goods, depending on the relative prices of these goods. This substitutability is constant and at the 

same level for any pair of intermediate goods. Similarly, in the generic version of the model, valueadded 

is a constant elasticity of substitution (CES) function of unskilled labor, land, natural resources, 

and of a CES bundle of skilled labor and capital. This nesting allows the modeler to introduce less 

substitutability between capital and skilled labor than between these two and other factors. In other 

words, when the relative price of unskilled labor is increased, this factor is replaced by a combination 

of capital and skilled labor, which are more complementary. 6 

Factor endowments are fully employed. The only factor whose supply is constant is natural 

resources. Capital supply is modified each year because of depreciation and investment. Growth 

rates of labor supply are fixed exogenously. Land supply is endogenous; it depends on the real 

remuneration of land. In some countries land is a scarce factor (for example, Japan and the EU), such 

5 The fixed-proportion assumption for intermediate inputs and primary factor inputs is especially pertinent to 

developed economies, but for some developing economies that are undergoing dramatic economic growth and 

structural change, such as China, the substitution between intermediate inputs and primary factor inputs may 

be significant. 

6 In the generic version, substitution elasticity between unskilled labor, land, natural resources, and the bundle 

of capital and skilled labor is 1.1, whereas it is only 0.6 between capital and skilled labor. This structure has 

been modified for the present exercise (see 4.2). 

29


that elasticity of supply is low. In others (such as Argentina, Australia, and Brazil), land is abundant 

and elasticity is high 7 . 

Skilled labor is the only factor that is perfectly mobile. Installed capital and natural resources are 

sector specific. New capital is allocated among sectors according to an investment function. Unskilled 

labor is imperfectly mobile between agricultural and nonagricultural sectors according to a constant 

elasticity of transformation (CET) function: unskilled labor’s remuneration in agricultural activities is 

different from that in nonagricultural activities. This factor is distributed between these two series of 

sectors according to the ratio of remunerations. Land is also imperfectly mobile between agricultural 

sectors. 

In the MIRAGE model there is full employment of labor; more precisely, there is constant aggregate 

employment in all countries, combined with wage flexibility. It is quite possible to suppose that total 

aggregate employment is variable and that there is unemployment; but this choice greatly increases 

the complexity of the model, so that simplifying assumptions have to be made in other areas (such as 

the number of countries or sectors). This assumption could amplify the benefits of trade 

liberalization for developing countries (see Diao et al. 2005): in full-employment models, increased 

demand for labor (from increased activity and exports) leads to higher real wages, such that the 

origin of comparative advantage is progressively eroded; but in models with unemployment, real 

wages are constant and exports increase much more. 

Capital in a given region, whatever its origin, domestic or foreign, is assumed to be obtained by 

assembling intermediate inputs according to a specific combination. The capital good is the same 

whatever the sector. MIRAGE describes imperfect, as well as perfect, competition. In sectors under 

perfect competition, there is no fixed cost, and price equals marginal cost. Imperfect competition is 

modeled according to a monopolistic competition framework. It accounts for horizontal product 

differentiation linked to product variety. Each firm in sectors under imperfect competition produces 

its own unique variety, with a fixed cost expressed as a fixed quantity of output. According to the 

Cournot hypothesis, each firm supposes that its decision of production will not affect the production 

of other firms. Furthermore, the firms do not expect that their decision of production will affect the 

level of domestic demand (which would be what modelers call a “Ford effect”). 

The demand side is modeled in each region through a representative agent whose propensity to save 

is constant. The rest of the national income is used to purchase final consumption. Preferences 

between sectors are represented by a linear expenditure system–constant elasticity of substitution 

7 This assumption that applies to the standard model is modified in the version of MIRAGE used in this biofuels 

study. 

30


(LES-CES) function. This implies that consumption has a non-unitary income elasticity; when the 

consumer’s income is augmented by x percent, the consumption of each good is not systematically 

raised by x percent, other things being equal. 

The sector sub-utility function used in MIRAGE is a nesting of four CES functions. In this study, 

Armington elasticities are drawn from the GTAP 7 database and are assumed to be the same across 

regions. But a high value of Armington elasticity, i.e. 10, is assumed for all homogenous sectors 

(single crops, single vegetal oils, ethanol). For biodiesel, we assume the same elasticity as that for 

other fossil fuels. Macroeconomic closure is obtained by assuming that the sum of the balance of 

goods and services and foreign direct investments (FDIs) is constant and equal to its initial value. 

3.2.2 Energy Modeling 

Most significant of these model modifications is the modeling of the energy sector to introduce 

energy products, including biofuels, as components of value-added in the production process. 

Following a survey of energy modeling approaches, the MIRAGE model was modified following a topdown 

approach, similar to the approach taken with the GTAP-E model (Burniaux and Truong, 2002) 

wherein energy demand is derived from the modeling of macroeconomic activity. However, beyond 

what is in the GTAP-E model, the MIRAGE model was revised to include a better representation of 

agricultural production processes to better capture the potential impact of biofuels development on 

agricultural production. The possibility of either intensive or extensive production of crops and 

livestock was introduced in the model. The characterization of demand for energy in non-agricultural 

sectors, particularly the elasticity of substitution between different energy sources, was also 

modified. Further details about the energy modeling developed for this study are in Annex II. 

In addition to the extensive modifications made to address the shortcomings of the MIRAGE global 

trade model in characterizing the energy sector, modifications were also made in the MIRAGE 

demand function for final consumption. The Linear Expenditure System - Constant Elasticity of 

Substitution (LES-CES), which captures non-homothetic behaviour in response to changes in income, 

was improved through the introduction of new calibration to USDA income and price elasticities 

(Seale et al., 2003). For China and India, some complementary information was sourced from FAPRI. 

The LES-CES demand structure was further modified to allow for a separate characterization of 

demand for fuel relative to demand for other goods. A new CES level is introduced to allow for the 

lower elasticity of fuel demand to prices. Further details on this energy demand structure 

modification is provided in Annex III. 

31


3.2.3 Fertilizer modeling 

Fertilizers are explicitly introduced in the global database and MIRAGE model to capture potential 

crop production intensification, using more fertilizers, in response to increased demand for biofuel 

feedstock crops. The characterization of the crop production response to prices resulting from 

increased bioenergy demand is particularly important. Through improved modeling of fertilizers and 

its impact on crop yield, we introduce a better representation of yield response to economic 

incentives while taking into account biophysical constraints and saturation effects. The degree of 

crop intensification depends on the relative price between land and fertilizers. Further details on the 

fertilizer modeling are provided in Annex IV. 

In this context, crop yields in the model increase through three channels: 

Exogenous technical progress (see baseline section); 

Endogenous “factor” based intensification: land is combined with more labor and capital; 

Endogenous “fertilizers” (intermediate consumption) based intensification, the mechanism 

described above. 

The model does not include endogenous technical progress based on private or public research and 

development expenditures in response to relative price changes. However, the increase of capital 

and labor by unit of land (effect ii) plays a similar role. 

3.2.4. Modelling of biofuel sectors 

The biodiesel and ethanol sectors are modeled in slightly different ways. Biodiesel production, which 

does not produce by-products, uses four kind of vegetal oils (palm oil, soybean oil, sunflower oil and 

rapeseed oil) as primary inputs (see Figure 1). These are combined with other inputs (mainly 

chemicals and energy) and value-added (capital and labour). Intermediate consumption are modeled 

using a CES nested structure with high substitutable (elasticity of substitution equals to 8) assumed 

among the vegetal oils. The initial dataset and the calibration of the model were set to allow for an 

initial marginal rate of substitution equal to 1 (e.g. one ton of rapeseed oil may be replaced by one 

ton of palm oil). The feedstock aggregate is then combined with a bundle comprised of the other 

components of intermediate consumption assuming complementarity (with elasticity of substitution 

equal to 0.001). As the only output of this sector, biodiesel can be exported or consumed locally. The 

share of the different vegetal oils is given by initial data but evolve endogenously through the CES 

aggregate. However, in this framework, a country that does not produce biodiesel initially will never 

produce biodiesel and if a biodiesel sector in one country does not initially use a type of vegetal oil as 

feedstock, it will never switch to such feedstock.For the ethanol sector, we first model four 

32


subsectors, each using only one of the following as specific feedstock -- wheat, sugar cane, sugar 

beet, or maize. This main input is combined with other production inputs and value-added assuming 

complementarity. Each subsector produces a specific by-product (DDGS with different properties and 

prices), except for the sugarcane-based ethanol sector, as well as the main output ethanol. These 

different types of ethanol are blended into one homogenous good that is exported or consumed 

locally. In addition, we allow for Central America and Caribbean regions the possibility to use 

imported ethanol for Brazil as an input into their own ethanol production sector. 8 Each type of DDGS 

is also directly traded or consumed by local livestock industries. It is important to emphasize that no 

other DDGS production is modeled outside of the production of ethanol. It means that the size of 

DDGS market is more restricted in the model than in the real world and will be totally dependent on 

the evolution of the ethanol production sectors. It is quite different from the production of meals 

wherein the vegetal oil production process itself generates oilcakes. Since the biodiesel sector is a 

limited destination for the overall vegetal oil sectors, the effects of biodiesel policies are much more 

limited on these markets. 

3.2.4 Modeling of Co-products and Livestock Sectors 

Co-products of the biofuels industry, such as Dried Distillers Grains with Solubles (DDGS), soy meal, 

and rapeseed meal, are used as substitutes for feedgrains in livestock production. It is therefore 

recognized that in assessing the impact of biofuels development on agricultural markets, co-products 

should be taken into account since they could lessen the unfavorable impact of biofuels: they reduce 

the need of land reallocation/extension to replace the crops displaced from the feed and food 

sectors to bio-energy production. Biofuel co-products are also recognized for their role in potentially 

mitigating the land use impact of biofuels as demand for feedgrains are reduced. Kampman et al. 

(2008) estimated that incorporating by-products into the calculations for land requirements of 

biofuels reduced land demand by 10-25%. 

Accounting for co-products was only recently introduced in CGE assessments of the impact of 

biofuels development. Taheripour et al. (2008) analysed the impact of including biofuel by-products 

(DDGS) in an analysis based on the GTAP CGE model. They found significant differences in feedstock 

output and prices depending on whether the existence of by-products is taken into account. 

Inclusion of co-products has become a prerequisite to the modeling of biofuel policy impact. 

8 The consumption of other inputs are corrected from the share of imported ethanol used in the processing of 

domestic ethanol under the assumption that transformation of processing of imported ethanol is performed at 

a low cost. However, only the existence of tariff preferences on the US and EU markets justify these indirect 

exports from Brazil. 

33


Significant efforts have been made in the US Environmental Protection Agency and the California Air 

Resource Board (CARB) assessments of indirect land use impact to take co-products into account at 

the US level (DDGS from corn ethanol production). 

Figure 1 Biofuel Feedstock Schematic 

Crops 

Sunflower 

seed 

Soybean 

Rapeseed 

Palm fruit 

& Kernel 

Oil sector 

(+meals) 

Sunflower 

oil 

Soybean oil 

Rapeseed 

oil 

Palm oil 

Biofuel 

Biodiesel 

Crops 

Biofuels 

(+ DDG) 

Blending 

Wheat 

Ethanol W 

Maize 

Ethanol M 

Sugar Beet 

Ethanol B 

Ethanol 

Sugar Cane 

Imported 

Ethanol* 

Ethanol C 

Imported 

Ethanol* 

*Only for Central America and Caribbean regions to represent the re-export channel of Brazilian ethanol in the 

region. 

Note: Other inputs and Value added are not displayed here. 

34


Co-products play a different role in the ethanol and in the biodiesel production pathways. For 

ethanol, distillers grains and sugar beet pulp are low value materials that are not profitable without 

the benefits from ethanol sale (the share of ethanol by-products in total production value is below 

20%). On the other hand, the production of oilseed meals is at the heart of oilseed market dynamics 

in biodiesel production. Oil and meals are co-products that can be valued independently and the 

demand for one of them directly affects the price of the other. This difference in treatment of coproducts 

of ethanol and biodiesel production is reflected in the modeling of co-products introduced 

in this study. For ethanol, co-products are represented as a fixed proportion of ethanol production, 

with the shares based on cost shares data for co-products for selected ethanol feedstocks in the USA 

and EU. For biodiesel, we consider as co-products the oilcakes\meals that are produced in the 

crushing of oilseeds to produce vegetable oils that are then processed for biodiesel production. We 

rely on cost share information for oilcakes in the vegetable oil production process. Co-products are 

then introduced in the model as substitutes for feedgrains in livestock production. Substitution 

between oilcakes, based on the protein content of the different oilcakes, is first introduced. The 

composite of oilcakes is then introduced as substitute for animal feed and DDGS as feed inputs to the 

livestock sector based on their energy content. However, we do not model the co-products of the 

biodiesel trans-esterification process, i.e. glycerol and similar products that can be used as additives 

to the feeding process. 

With the introduction of co-products in the model, the modeling of livestock production was also 

significantly modified to allow for intensification through substitution of livestock feed, including 

ethanol and biodiesel co-products, with land. This is treated using a similar approach to our modeling 

of crop intensification through substitution of fertilizer for land, and is assessed as an alternative case 

in the sensitivity analyses. Further details on the modeling of co-products are given in Annex V. 

3.3 Land Use Module 

To capture the interactions between biofuels production and land use change, we introduce a 

decomposition of land use and land use change dynamics. Land resources are differentiated between 

different agro-environmental zones (AEZ). The possibility of extension in total land supply to take 

into account the role of marginal land is also introduced. The modeling of land use change captures 

both the substitution effect involved in changing the existing land allocation to different crops and 

economic uses, and the expansion effect of using more arable land for cultivation. Detailed 

documentation of the land use module including data on AEZs and land use change modeling are 

35


available in Annex VI. Land extension takes place at the AEZ level allowing capturing different 

behaviour across different regions of large countries (e.g. Brazil). 

To determine in which biotope cropland occurs, we follow the marginal land extension coefficients 

computed by Winrock International for the US EPA, wherein the extent of land use change over the 

period 2001 to 2004 was determined using remote sensing analysis. For Brazil, these coefficients are 

defined at the AEZ level to capture that deforestation occurs in specific regions. This feature is 

particularly important since sectoral distribution will lead to different deforestation behaviour: for 

instance, soya crops are closer to the deforestation frontier than sugar cane plantations. Although 

the historical trends for land use change are followed in the baseline, changes in land use allocation 

in the scenarios come from the endogenous response to prices through the substitution effects. 

Therefore, historical land use changes do not affect the distribution of land under economic use 

across their alternative uses (cropland, pasture, managed forest). 

We also introduce a mechanism for expansion or retraction of pasture land in response to changes in 

demand for cattle. Alternative assumptions regarding the links between demand for cattle and for 

pastureland and for the possibility of intensification are accommodated in the revised modeling of 

land use expansion and discussed in Annex VI. 

3.4 GHG Emissions and Marginal ILUC Measurement 

A critical component of this study is the assessment of the of balance in CO2 emissions between (a) 

direct emission savings induced by the production and use of biofuels and (b) possible increases in 

emissions as a result of indirect land use changes (ILUC) induced by biofuels production. 

Direct emissions savings for each region, are calculated primarily using the typical direct emission 

coefficients for various production pathways as specified in the EU Renewable Energy Directive (see 

in Annex VII). Additional sources were used for the relevant emissions coefficients data for other 

regions (EPA, 2009). We also perform sensitivity analysis on these values. The values of these 

coefficients are critical to the determination of direct emission savings and the net emissions effects 

of biofuels. We do not model each production pathway separately in the model but calculate an 

average composition of the biofuels production sector. Data on that composition remain sparse 

however; consequently the current average composition of production capacity in the industry 

remains uncertain as well. Moreover, there are major uncertainties with regard to (a) the future 

weight of each of these production pathways in total production and (b) the possibility for 

substitution between different pathways to comply with the sustainability criteria defined in the RED. 

As a result, major uncertainties remain regarding the direct emission savings in the biofuels industry. 

36


We use the consumption approach to allocate direct emission savings: the emission credit is given to 

the country that consumes the biofuels, not to the producer country. In this we follow the RED 

directive even though this may appear to be in contradiction with the UNFCCC and Kyoto Protocol 

emission accounting rules that allocate credits for reductions to the producer country. 

In calculating the GHG emissions from indirect land use change, the study considered emissions from 

(a) converting forest to other types of land, (b) emissions associated with the cultivation of new land 

and (c) below-ground carbon stocks of grasslands and meadows. We rely on IPCC coefficients for 

these different ecosystems. We also include two different treatments. For the EU, the carbon stock 

of forest is limited to 50% of the value for a mature forest. It is considered that no primary forest will 

be affected by the land extension in the EU and only the areas recently concerned by afforestation 

will be impacted. 

For Indonesia and Malaysia, we include in addition to the carbon stocks (above and below ground), 

the emissions from peatlands converted to palm tree plantations. We assume a marginal coefficient 

of extension of palm tree plantations on peatlands of 10% for Malaysia and 27% for Indonesia, based 

on statistics provided by Wetlands International 9 . We use two sets of emissions coefficients for 

peatlands, from IPCC – AFOLU and from Couwenberg (2009), since the literature displays a wide 

range of coefficients (from 5 to 40 tonnes of CO2 by hectare). Recent trends emphasize the 

underestimation of past values. 

In this study, we compute the overall effect of the mandate using average ILUC, as well as marginal 

ILUC (the effect of an additional unit of biofuels). The two notions differ from each other due to the 

non-linearity of marginal ILUC in the model. 10 

We estimate the marginal ILUC effects for each feedstock, measured in tons of CO2 emissions per 

metric ton and per Giga Joule of biofuel, resulting from a marginal extra demand of 10 6 GJ, i.e. 

around 0.1% of the consumption level at this stage, applied to the EU mandate level. Further details 

are provided in Annex VII. 

9 http://wetlands.org/. 

10 The distinction between the concept of average (mean) and marginal ILUC is discussed in Tipper et al. (2009). 

37


4 Baseline, Trade Policy Scenarios, and Sensitivity Analysis 

This section provides a description of the baseline scenarios, the alternative trade policy scenarios, 

and the sensitivity analyses conducted on some parameters used in the model. The baseline scenario 

provides a characterization of growth of the global economy up to 2020 but without the biofuels 

policy scenarios of interest in the study. We then introduce the EU biofuels mandate as a policy 

scenario and examine the resulting changes compared to the baseline scenario. We also introduce 

alternative trade policy scenarios around this EU biofuels mandate scenario impact. Moreover, since 

the values of some parameters used in the model are uncertain, sensitivity analyses are performed 

by simulating the policy scenarios using alternative values of key parameters. 

4.1 Sectoral and Regional Nomenclature 

Even if the database has been developed at a detailed level (57 sectors and 35 regions), it is not 

practical to run the scenarios at this highly detailed level due to the much larger size of this model 

(now twice the number of equations/variables than the normal MIRAGE model) and the modeling of 

land extension at the detailed AEZ level. Focusing on the sectors and regions of interest in this study 

on biofuels and agricultural production and trade from an EU point of view, we limit the size of our 

aggregation to the main players (11 regions) and 43 sectors. Details are provided in Table 1 and 2. 

The sectoral disaggregation covers agricultural feedstock crops and processing sectors, energy 

sectors and other sectors that also use agricultural inputs. 

Table 1 Regional Aggregation 

Region 

Brazil 

CAMCarib 

China 

CIS 

EU27 

IndoMalay 

LAC 

RoOECD 

RoW 

SSA 

USA 

Description 

Brazil 

Central America and Caribbean countries 

China 

CIS countries (inc. Ukraine) 

European Union (27 members) 

Indonesia and Malaysia 

Other Latin America countries (inc. Argentina) 

Rest of OECD (inc. Canada & Australia) 

Rest of the World 

Sub Saharan Africa 

United States of America 

38


Table 2. Sectoral Aggregation 

Sector Description Sector Description Sector Description 

Rice Rice SoybnOil Soy Oil EthanolW Ethanol - Wheat 

Wheat Wheat SunOil Sunflower Oil Biodiesel Biodiesel 

Maize Maize OthFood Other Food sectors Manuf Other Manufacturing 

activities 

PalmFruit Palm Fruit MeatDairy Meat and Dairy WoodPaper Wood and Paper 

products 

Rapeseed Rapeseed Sugar Sugar Fuel Fuel 

Soybeans Soybeans Forestry Forestry PetrNoFuel Petroleum products, 

except fuel 

Sunflower Sunflower Fishing Fishing Fertiliz Fertilizers 

OthOilSds Other oilseeds Coal Coal ElecGas Electricity and Gas 

VegFruits Vegetable & Oil Oil Construction Construction 

Fruits 

OthCrop Other crops Gas Gas PrivServ Private services 

Sugar_cb Sugar beet or OthMin Other minerals RoadTrans Road Transportation 

cane 

Cattle Cattle Ethanol Ethanol - Main AirSeaTran Air & Sea 

OthAnim 

Other animals 

(inc. hogs and 

poultry) 

sector 

transportation 

EthanolC Ethanol - Sugar Cane PubServ Public services 

PalmOil Palm Oil EthanolB Ethanol - Sugar Beet 

RpSdOil Rapeseed Oil EthanolM Ethanol - Maize 

4.2 Baseline Scenario 

It is important to emphasize that the underlying GTAP database is first updated from the 2004 data 

reference year to 2008 through a simulation that uses external macroeconomic variables (GDP, 

population, labor force) over that period, as well as by targeting observed biofuel production and 

consumption data for 2008. Endogenous variables (mandate) are used to reach these levels. After 

2009, we let the model evolve freely in the baseline except for the macroeconomic variables and oil 

prices that are still targeted. 

An exhaustive description of the baseline scenario is provided in the Excel workbook that 

accompanies this report: Details_baseline_CentralScenario.xlsx. 

4.2.1 Macroeconomic Trends 

The baseline scenario reflects recent International Energy Agency forecasts (2008) with oil prices 

reaching $120 a barrel in 2030 current prices. Economic growth projections, now taking into account 

the effects of the economic crisis, have also been updated with projections data from the World 

Economic Outlook (April 2009) of the International Monetary Fund. In this context, EU consumption 

39


of energy for road transportation is estimated to reach 316 Mtoe in 2020. This figure is in line with 

the latest projections of DG ENER. However, this number may appear too high when new EU policies 

aimed a reducing energy consumption are taken into account. 

4.2.2 Technology 

The average total factor productivity (TFP) in the economy is computed endogenously to reach the 

real GDP target in the baseline. 

In agriculture, we introduce country and sector specific TFP rates based on estimates from Ludena et 

al. (2006). It is important to note that no exogenous growth in palm tree yield is assumed due to the 

lack of data at our disposal. Therefore, compared to other crops, palm oil tends to suffer from a 

disadvantage in the baseline. Yields in the palm fruit sector can only increase through an endogenous 

process (intensification). (See table B9 of the Baseline Excel workbook for details). We do not assume 

changes in the yield of the crushing, distilling and biofuel production activities. 

It is important to notice that these projections assume very low exogenous productivity increases in 

EU agriculture, both when comparing agriculture to other sectors in the EU and also comparing EU 

agriculture to its main competitors (up to +5% only for main crops in the EU whereas yields increase 

by more than 30% in Brazil). This assumption is based on Ludena et al. (2006) but leads to losses of 

competitiveness of EU agriculture in the baseline and will have adverse consequences on 

endogenous yield growth. Indeed, since agricultural sectors are below EU average in terms of 

productivity growth, capital will tend avoid these sectors as expected returns are higher in other 

sectors. Less capital accumulation leads to low yield increases through factor intensification. 

4.2.3 Trade Policy Assumptions 

The baseline scenario leaves the trade policies that were in place by end 2008 unchanged. The 

Economic Partnership Agreements (EPAs) between the EU and the ACP countries, negotiated in 2008, 

are implemented either as ratified interim agreements or a complete EPA (e.g. with CARICOM), 

depending on the status of the agreement. Negotiations on trade agreements that were not finalized 

by end 2008 are not included: the Doha Development Agenda, an EU-ASEAN agreement and an EU- 

Ukraine agreement. 

The baseline scenario includes the full ad-valorem equivalent (AVE around 48%) of the prevailing EU 

MFN duty on EU bioethanol imports from countries that do not benefit from bilateral or unilateral 

(GSP) preferential schemes. In reality, this is likely to be an overestimate of the effective AVE. 

Significant quantities of bioethanol are imported under temporary suspensions of duties and, in the 

form of denatured ethanol, as chemical products for which a lower duty applies. In the absence of a 

40


specific EU tariff line for bioethanol, there are no trade statistics available that permit us to estimate 

the effective trade-weighted tariff on bioethanol. 

Another critical trade policy measure that we incorporate in the baseline scenario are the antidumping 

duties that the EU imposed on US exports of biodiesel in March 2009. Over the last few 

years, the US has emerged as the major biodiesel exporter to the EU (with more than 80% of market 

share among all exporters), supplying about 19% of the EU domestic market for biodiesel. However, 

due to the tax credit given to the US blenders, and the splash’n dash practice, the EU initiated antidumping 

measures and countervailing duties in March 2009. This contingent protection has reduced 

US biodiesel exports to the EU to negligible quantities. Allegedly, some of these US exports may now 

have been replaced partially by exports from Indonesia and Malaysia and Argentina and growing 

trade flows from Canada. 11 In the model, the bulk of the adjustment to the antidumping duty is 

achieved through increased in EU biodiesel production (based on EU produced and imported 

feedstocks). Figure 2 shows the change in EU imports in 2008 and 2020. 

Figure 2 EU biodiesel imports by source, Mtoe, in the baseline 

EU imports, Mtoe 

2.00 

1.80 

1.60 

1.40 

1.20 

1.00 

0.80 

0.60 

0.40 

0.20 

0.00 

2008 2020 

USA 

RoOECD 

LAC 

IndoMalay 

China 

Brazil 

Source: Authors’ calculations 

4.2.4 Agricultural and Agri-Energy Policies 

For the EU, we implement two policy elements in the baseline: 

i. The sugar reform market; 

ii. 

The end of the land set-aside policy. 

11 These flows can be re-exported US production and in some cases, double splash’n go has been detected (tax 

credit in the US then in Canada). 

41


These two assumptions have overall limited effects in the baseline. First, we remove the land setaside 

constraint by 2008 (full use of EU land). The main effect is to lead to a fall in EU yields from 

2007 to 2020 by an average of 10 percent. This result is quite strong and will be translated into a 

proportional fall in land prices. Indeed, we force EU farmers to use all set-aside land (10% of the 

overall croplands in our baseline) when overall demand for crops will not change during the same 

period. Therefore, EU production will not change when the harvested area will increase by 10% and 

yield decreases. Since the relative price between land and fertilizers determines the use of fertilizers 

in this model, another yield-depressing effect appears: lower land prices reduce intensification 

behaviour and yield. The effects are differentiated between crops depending on existing tensions on 

markets during the period in the baseline: stronger for crops with low demand (other crops -15%), 

weaker for crops with high demand (-5%). The combination of this with our assumptions on EU 

agricultural productivity (Section 4.2.2) leads to the decline in EU yields in the baseline. This is a 

crude modelling solution for land set-aside and it should be improved. In particular, forcing farmers 

to use all the land set-aside has a strong mechanical effect. In reality, it appears that these lands have 

lower yields than average and that only a share of it has been used in 2008, even during crop price 

surges. 

Second, since we do not explicitly model the existing sugar policy tool, we mimic the sugar market 

reform by reducing the EU MFN tariff to reproduce the price decrease. Overall, the EU sugar 

production decrease by 5% between 2008 and 2020 when the world production increases by 47%. 

The effects of the reform are slightly absorbed by the ethanol industry since the sugar-beet ethanol 

industry is the most resilient in the baseline (see next paragraph for the evolution of the biofuels 

sector in the baseline). 

In the baseline, no additional EU bioenergy mandate is implemented. The status-quo is assumed to 

prevail until 2020, with biofuel blending levels not exceeding the 3.3% level in 2008. The previous EU 

target of 5.75% blending is not implemented. We do this to capture the impact of the EU mandate 

against a baseline where biofuel use remains at the 2008 blending levels (3.3%). It implies that EU 

consumption reach 9.75 Mtoe in 2020 with a 90% share for biodiesel. At the same time, production 

increases by 22% while imports fall by 68% with the exclusion of the US from the market (see Figure 

2). Interestingly, EU production of bioethanol falls by 20% under the pressure of foreign competitors 

(Brazil). Indeed it appears that the EU has no dynamic comparative advantage in this sector, contrary 

to biodiesel. 

This result is quite strong and has several explanations. First, the relative price of cereals compared 

to sugar cane/sugar beet increases. This is due mainly to the evolution of world demand and the role 

42


of cereals in cattle feeding but also demand from agribusiness sectors (flours etc.). This price gap 

leads to a loss of competitiveness of EU ethanol (except for sugar beet). Second, as discussed 

previously, EU yields will progress – exogenously and endogenously - very slowly compared to Brazil. 

In addition, the land constraint is tighter in the EU than in Brazil. We have also a clear dichotomy 

between EU and Brazil agricultural supplies since in the former land is scarce and intensification 

already high, when in the latter both extensive and intensive growth appear to be very easy. This 

undermines the overall competitiveness of EU ethanol. Last, we have a CGE effect: with the loss of 

competitiveness of the EU ethanol sector, capital accumulation will slow, other sectors being more 

attractive, and the ethanol sectors will shrink in the EU. 

Since there are already strong political commitments in place in these countries, we implement the 

US and Brazilian biofuel targets in the baseline. 12 The US mandate will lead to the consumption of 40 

Mtoe of ethanol by 2020. The US production of ethanol will increase by 128% in twelve years while 

the US biodiesel sectors will expand by 193% (but will represent only 12% of the ethanol sector). 

With the Brazilian blending target fixed at 24.4% over the period, its ethanol production rises by 

139%. We also include a 5% mandate for Indonesia, Malaysia, Rest of OECD and China. This 

assumption is aimed to maintain a minimal consumption target in these countries in the baseline and 

in the scenarios. It is important to take other countries' bioenergy consumption targets into account 

since they affect the amount of foreign feedstock and biofuels production that the EU will be able to 

import and thus the future domestic production in the EU. 

4.2.5 Other Baseline Evolutions 

As described previously, oil prices follow trends proposed by IEA in the recent World Energy Outlook 

with an oil price stable at $83.8 a barrel by 2010 and increasing slowly up to $96.4 in 2015, and $109 

in 2020 (values are given in 2004 constant dollars). Oil production is forecast to experience 

constraints with an increase of only 32% on the period 2010-2020. 

Demand for all crops increases only marginally (+27% in world production) over the same period. The 

highest increases in demand are for palm fruit (60%) and for sugar cane, sugar beet and soybeans 

sectors (+47%). Demand for cereals faces limited increases (about 20% for both wheat and maize). 

These figures are above the FAO-Aglink projections and are mainly driven by a relatively inelastic 

demand for agricultural products by other sectors (services, agri-business, chemistry) and are 

intrinsic to the CGE exercise. This forecast is based on the assumption that no major changes occur in 

the diet of the world population. 

12 A survey of biofuels policies in the EU, US and Brazil is provided in Annex X. 

43


Given these forecasted changes, cropland expansion is expected to be 1 Mios of km2 between 2008 

and 2020 (+9% for crops), with substantial expansion in Brazil (+36%) and Africa (+22%). In Europe, 

the cropland surface will increase by 5% between 2008 and 2020. 

4.3 Central and Alternative Trade Policy Scenarios 

Against this baseline scenario, we evaluate the impact of three different trade policy scenarios. In the 

central scenario, we introduce a biofuels policy shock that assumes that the EU will consume 17.76 

Mtoe of bioethanol and biodiesel by 2020 in order to achieve the mandate target of 10% renewable 

energy in road transport fuels. This figure is taken from an intermediate biofuels demand scenario by 

DG ENER, based on the PRIMES model, that combines various renewable energy sources, including 

second generation biofuels and increased use of electric cars powered by renewable electricity. 

Furthermore, the model uses a target ratio for 2020 of 55% ethanol and 45% biodiesel, based on DG 

AGRI projections. 13 

However, the current baseline does not include new projections for total road transport fuel 

consumption in the EU in 2020, taking into account new EU energy and emission policy initiatives. 

For this reason, we stick to the existing PRIMES figure of 316 Mtoe by 2020, and derive a biofuels 

incorporation ratio of 5.6% 14 . As a result, the denominator of that ratio is probably too high. We do 

however test the sensitivity of the outcomes for other values of this ratio (see below 4.4.1) 

The mandate target is achieved in the model by mandatory regulation (explicit biofuels mix 

constraints build into the supply of road transport fuels) and not by means of explicit subsidies or tax 

credits. 

Our trade policy scenarios are: 

MEU_BAU: Implementation of the EU biofuels mandate of achieving 5.6% consumption of 

ethanol and of biodiesel in 2020 under a Business as Usual trade policy assumption; 

MEU_FT: Implementation of the EU biofuels mandate of achieving 5.6% consumption of 

ethanol and of biodiesel in 2020 with the assumption of full, multilateral, trade liberalization 

in biofuels. Contingent protection on US biodiesel remains; 

13 “Impact Assessment of the Renewable Energy Roadmap - March 2007”, DG AGRI, AGRI G-2/WM D(2007). 

These targets are still very close to the latest estimates of the JRC ISPRA. The ratio of bioethanol to biodiesel is 

largely determined by the car fleet composition. Diesel cars cannot use petrol, and vice versa. We assume that 

the fleet composition is exogenous to the model and not influenced by EU biofuels policies. 

14 Note that this estimated 5.6% target for biofuels in 2020 is actually below the previous target of 5.75% for 

2012. These 5.6% include land-using first-generation biofuels only. Non land-using first generation biofuels 

such as recycled waste oil and animal fats are not included. 

44


MEU_MCS: Implementation of the EU biofuels mandate of achieving 5.6% consumption of 

ethanol and of biodiesel in 2020 with the assumption of EU bilateral trade liberalization with 

MERCOSUR. 

Two important points regarding the trade policy scenarios have to be emphasized. First, the size of 

the mandate is not excessive since it will require an increase in EU demand of biofuels by 70% and an 

8% increase of world production/consumption of biofuels. The limited size of the shock explains the 

magnitude of our results in the next section. Due to the potential non-linearity in our analytical 

framework (see section 5.2.3), this policy design will also explain the relatively low per unit cost (CO2 

and economic inefficiency) of such a mandate. Second, the initial ad valorem equivalent (AVE) MFN 

tariff on EU imports that we use, about 50%, appears to be an upper bound to more recent estimates 

(25%-30%). 15 Combined with the high Armington trade elasticity assumed for this product to 

represent a more homogeneous good, the effects of trade liberalization will be very strong, and may 

be overestimated. 

4.4 Sensitivity Analysis Design 

Assessing the impact of biofuel policies and the ILUC coefficients – the focus of this study – is quite 

challenging due to a lot of uncertainties. We can group them into two categories: mandate policy 

targets and varying parameter settings. We assess the robustness of our central case results by 

performing sensitivity analysis on these different dimensions. A third set of sensitivity analyses 

regarding modeling assumptions is performed on two issues and reported in relevant annexes: the 

modeling of fertilizers (Annex IV) and the interaction between pasture and crop lands (Annex VI). 

4.4.1 Mandate Policy Targets 

The overall size of the biofuels policies should matter in quantifying the economic and environmental 

impact of the policy. Due to decreasing marginal productivity, we expect that applying the same 

marginal change on a low or high level of biofuel demand and supply can play a very different role. 

The goal of this analysis is to check if (average and marginal) ILUC is constant or increasing with the 

total demand for biofuels. 

15 Please note that the estimation of the EU AVE on ethanol is complicated by two main difficulities: (i) 

identification of the relevant unit value on imports, and (2) identification of the tariff line actually used by 

Member States to import ethanol for biofuel production. 

45


Since the overall ambition of the EU mandate is an important question, we look at different values 

for the mandate: 4.6%, 5.6%, 6.6%, 7.6% and 8.6%, equivalent to 14.5 Mtoe, 17.8 Mtoe, 20.7 Mtoe, 

23.9 Mtoe and 27 Mtoe of biofuels consumption, respectively. 

4.4.2 Parameter Uncertainties 

It is important to underline that the values of some key parameters in the model are still subject to 

considerable uncertainty. It is therefore important to assess the role of alternative values in 

determining the robustness of the results. 

Land and fertilizer substitution – Due to uncertainty about the values of elasticity of substitution 

between land and fertilizers, sensitivity analysis (is done by looking at the impact of using twice the 

land/fertilizer substitution elasticity in the base case. 16 Increasing the elasticity should help the 

farmers to intensify their production more easily and will limit the pressure for new lands. 

In addition, in Annex IV, we also analyze the consequences of alternative modeling of fertilizers. 

Land substitution – Due to uncertainty about the value of the elasticity of land substitution across 

agricultural production, i.e. how easily land can be shifted from one crop to another, we investigate 

two cases: 

Elasticity of land substitution between crops are doubled; 

Elasticity of land substitution between crops and pasture are doubled. 

The last section of Annex VI provides a discussion of the role of the interaction between croplands 

and pasture in our modeling and describes three variations on how pasture land area is affected by 

increased demand for livestock. In the simulations in this report, we use the mode P=1 wherein 

increased demand for livestock could lead to intensification in some regions, thereby affecting the 

amount of land that is substituted between the livestock and crop sectors. 

Land use extension – Due to uncertainty about the value of elasticity of the land extension supply 

curve, i.e. how new land are converted to agricultural uses when the rental price of land increases, 

we conduct sensitivity analysis by varying the value of the land extension elasticity. Our main 

estimates are based on Barr, et al. (2010) for the US and Brazil and on the OECD. Current values 

assume much more flexibility in Brazil and a land extension elasticity in Brazil that is 5 times higher 

than in the US or in the EU. We look at two specific scenarios: 

16 The basic value has been calibrated based on detailed elasticity information extracted from the IMPACT 

model (Rosegrant et al. 2008) 

46


We increase the land extension elasticity in Indonesia and Malaysia to reach the level for 

Brazil; 

We reduce by half the land extension elasticity in Brazil (which could be the case if Brazil 

manages to enforce its preservation program). 

Other parameters that may be critical to te overall assessment of the emissions effects of the biofuel 

mandates are: the choice of direct emissions savings and the coefficients of land use extensions. 

Since different set of values are available and are based on different methodological choices, we 

discuss them in Annex IX. 

Technology Pathway – In the assessment of the direct GHG emissions from different biofuel 

feedstocks used by major biofuels producers, we rely on a set of direct emissions coefficients that are 

sourced from the EU RED Directive, or from the literature. The values are employed in the central 

scenario. These values, as well as the results of a sensitivity analysis on these values are discussed in 

Annex VIII. 

It is important to keep in mind that alternative technology pathways are used in an ad-hoc method 

(per unit coefficient) and do not lead to a modification of the sectoral technology used in the model. 

We expect that the better the technology (higher reduction coefficients) the better the net CO2 

balance effect. 

47


5 Results and Discussion 

In this section, we present the results of the central scenario along with alternative trade policy 

scenarios, focusing first on the potential impact on production and trade under these policies and 

then on the land use and environmental impact in terms of GHG emissions from direct and indirect 

land use changes. Included in this assessment of environmental impact is the calculation of marginal 

crop-specific ILUC change, which is an important focus of this study. The final sub-section presents 

the results of several sensitivity analyses that are designed to assess the robustness of the results to 

changes in the mandate policy and some parameter values. The full set of results indicators 

calculated for the scenarios are available in the Detailed_scenario_results.xlsx. 

5.1 Production and Trade Impact of Trade Scenarios 

In this section we examine the impact of two policy scenarios: 

First, the European mandate scenario seeks to achieve the EU policy objective of at least 5.6% 

biofuels consumption in transport fuels in 2020 by imposing that bio/fossil fuel mix on all fuels sold in 

the EU. In that case, the consumer bears most of the cost of any fuel price increases at the pump. It is 

compared to the baseline situation where no mandate is implemented. The mandate is implemented 

progressively and in a linear fashion from 2010 to 2020. It is applied on each type of biofuel and no 

blending over 5.6% is allowed for biofuels in either gasoline or diesel. No change in trade policies are 

considered (scenario MEU_BAU). 

Second, the trade liberalization scenario consists of reaching the same objective through a more 

market-based approach, by lowering the consumer price of biofuels in order to stimulate 

consumption. This is achieved, in a first scenario, by the full liberalization of biofuels sectors (scenario 

MEU_FT). A second scenario consists in a liberalization of biofuels trade between MERCOSUR 

countries and the EU (scenario MEU_MCS). We do not present in the report the detailed figures for 

the EU-Mercosur scenario since it leads to result very similar to the multilateral liberalization. 

We evaluate the effects of these policy scenarios on several key elements - biofuel production, 

biofuel imports, crop production, agricultural value-added, variation of land use by sector, variation 

of total land use, variation of the intensification index for cultivation ($ of fertilizer used by ha), direct 

emissions reduction related to biofuels, and indirect emissions related to indirect land use change 

effect. 

48


5.1.1 Biofuel Production and Imports 

Table 3 illustrates the impact of the various scenarios on biofuel production. The two first columns in 

Table 3 provide the level of ethanol production in 2008 and in 2020 in the baseline (without policy 

shocks – column Ref). The next columns give the level and variation of production in 2020 implied by 

the two scenarios with variation being a comparison with the baseline. The same table organization 

is kept throughout all the report unless indicated otherwise. 

Table 3 Level and variation of biofuels production (Mio toe and %) 

REF MEU_BAU MEU_FT 

Lev Lev Var Lev Var 

Biodiesel Brazil 0.36 0.37 1.81% 0.37 2.92% 

Biodiesel China 0.23 0.23 -0.72% 0.23 -0.76% 

Biodiesel EU27 8.15 9.04 10.92% 9.07 11.27% 

Biodiesel IndoMalay 3.58 3.65 2.06% 3.65 2.07% 

Biodiesel LAC 0.45 0.48 5.91% 0.48 6.10% 

Biodiesel RoOECD 3.24 3.24 -0.01% 3.24 0.12% 

Biodiesel USA 3.46 3.45 -0.18% 3.46 -0.03% 

Biodiesel World 19.46 20.45 5.08% 20.49 5.30% 

Ethanol Brazil 28.51 32.78 14.97% 34.36 20.50% 

Ethanol CAMCarib 7.25 7.45 2.64% 7.19 -0.89% 

Ethanol China 10.81 10.83 0.18% 10.83 0.16% 

Ethanol EU27 0.84 2.17 156.89% 0.44 -48.23% 

Ethanol LAC 0.69 0.69 0.95% 0.70 2.21% 

Ethanol RoOECD 5.66 5.78 2.03% 5.84 3.03% 

Ethanol RoW 1.51 1.50 -0.54% 1.50 -0.49% 

Ethanol USA 29.10 29.57 1.64% 29.72 2.14% 

Ethanol World 84.38 90.77 7.58% 90.57 7.34% 


The mandate scenarios and trade liberalization scenario have very contrasting effects on biofuel 

production in the European Union. In 2020 ethanol production increases by 157% in the EU under an 

EU mandate scenario, while the competition coming from increased imports in a trade liberalization 

scenario would mean a decrease by -48% in case of full liberalization scenario. The removal of tariffs 

on ethanol would be followed by a surge in European imports of this product (they are multiplied by 

6.8 by 2020 – see Table 4) under trade liberalization scenario. As previously mentioned, since the 

baseline tariff may be overestimated (by a factor of 1.5), the effects of trade liberalization simulated 

here may also be overstated. 

As can be expected, the European mandate increases overseas production of ethanol by less than 

when it is coupled with trade liberalization. The greatest impact are seen in the two largest 

producers, the US and Brazil. In particular, Brazilian ethanol production is increased by 5.8 Mios toe 

49


(+20%)in 2020 under the trade liberalization scenario, while it is increased by +4.3 Mios toe (15%) 

under a European mandate. Effects on US production are more limited US (+2.14% with trade 

liberalization). US exports to the EU do not increase significantly (they remain a tiny fraction of the 

market) but they need to replace displaced Brazil exports. However, the free trade scenario leads to 

a strong preference erosion for the Central America and Caribbean region (-83%). 

Table 4. Level and Variation of EU Biofuel Imports, by Origin (Mio toe and %) by 2020 



Biodiesel Brazil 0.00 0.00 6.21% 0.00 5.49% 

Biodiesel China 0.00 0.00 14.45% 0.00 14.59% 

Biodiesel IndoMalay 0.44 0.51 15.29% 0.51 15.46% 

Biodiesel LAC 0.19 0.22 15.69% 0.22 16.04% 

Biodiesel RoOECD 0.00 0.00 12.92% 0.00 82.07% 

Biodiesel USA 0.00 0.00 11.78% 0.00 12.10% 

Biodiesel World 0.64 0.74 15.40% 0.74 15.79% 

Ethanol Brazil 0.92 5.53 502.82% 7.56 724.32% 

Ethanol CAMCarib 0.04 0.27 517.35% 0.01 -83.48% 

Ethanol USA 0.00 0.01 546.96% 0.00 111.89% 

Ethanol World 0.96 5.82 503.58% 7.57 685.98% 


Figure 3 shows EU production of biofuels in 2020 broken down by feedstock crops. The ranking 

among feedstocks by share of productionin 2008 is not modified since the impact of trade 

liberalization for the biodiesel sector is weak and the effects of the mandate are very limited. We see 

only a slight expansion of the share of palm oil in EU biodiesel production 17 and a contraction of the 

share of rapeseed oil. It shows that palm oil is marginally more competitive and with a larger 

mandate (and a stronger demand of biodiesel), we can expect a larger use of palm oil. This is also 

true for soya (from 32% to 33%). It is important to keep in mind that with the antidumping and 

countervailing duties applied in the baseline, the significant share of US soya-based biodiesel was 

already eliminated in the baseline. 

For the ethanol sectors, the evolution of the feedstock structure of EU production is stronger. When 

the demand for EU ethanol is high (no trade liberalization), most of the production expansion will be 

based on sugar beet (from 41% to 45% of EU ethanol production). Symmetrically, with trade 

liberalization, this feedstock will be marginally the most affected ( from 41% to 37%). 

17 This is in addition to the increase in biodiesel imports. 

50


Figure 3 Structure of EU Biofuels Production by Feedstock (2020) 

Biodiesel 

Share of EU production (MToe) by 

feedstock 

100% 

90% 

80% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

0.38 0.45 0.45 

2.58 2.96 2.98 

4.33 4.58 4.60 

0.86 1.04 1.04 


Sunflower 

Soybeans 

Rapeseed 

PalmFruit 

Ethanol 

Share of EU production (Mtoe) by 

feestock 

100% 

90% 

80% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

0.43 0.99 

0.24 

0.35 

0.98 

0.16 

0.07 0.20 0.04 


Wheat 

Sugar_cb 

Maize 


5.1.2 Agricultural Production 

These various policy scenarios have significant impact on crop production, particularly on feedstocks 

needed for the production of ethanol and biodiesel. This is particularly true for rapeseed and sugar 

cane-sugar beet. For example, while the production of sugar cane-sugar beet is increased under the 

MEU_BAU scenario (+3.8% in 2020 with +9.7% for Brazil –sugar cane, see Table 5, and +9.3% for the 

EU –sugar beet, see Figure 4), this increase is much more significant in the case of trade liberalization 

(+4.9% under the MEU_FT scenario with +15% for Brazil –sugar cane, and a decrease of -2.4% for the 

EU –sugar beet). 

51


Table 5. Main Changes in Crop Production (non EU27) in 2020, 1000t 

Crops Region REF MEU_BAU MEU_FT 


Sugar_cb Brazil 913385 1001556.15 9.65% 1045492.08 14.46% 

Rapeseed CIS 571 583.00 2.06% 583.42 2.13% 

PalmFruit Brazil 3117 3196.06 2.53% 3181.86 2.07% 

Rapeseed Brazil 151 153.15 1.59% 152.85 1.39% 

Rapeseed SSA 108 108.87 1.10% 108.89 1.12% 

Sunflower Brazil 153 155.23 1.24% 154.91 1.03% 

Rapeseed RoOECD 13848 13969.92 0.88% 13975.74 0.92% 

Soybeans RoOECD 3999 4020.98 0.54% 4025.62 0.66% 

Sunflower USA 2142 2155.86 0.64% 2156.20 0.65% 

Soybeans CIS 1129 1134.41 0.46% 1135.71 0.58% 

Soybeans LAC 77981 78349.47 0.47% 78428.70 0.57% 

Sunflower LAC 5883 5916.54 0.57% 5916.34 0.57% 

Rapeseed LAC 141 142.09 0.52% 142.10 0.53% 

OthCrop Brazil 9090 9034.08 -0.61% 9002.90 -0.96% 

Wheat IndoMalay 1 0.55 -5.92% 0.55 -6.81% 


These policy scenarios have a substantial impact on the European production of agricultural crops 

(Figure 4). As a result of the development of ethanol and biodiesel, the European production of crops 

used in these processes of production is increased in 2020: rapeseeds, sugar beet, wheat, maize, 

soybeans and sunflower. 

The production of various agricultural crops competes for common scarce productive resources (like 

land). On the one hand the production of agricultural commodities for non-food purposes can have 

negative consequences on other agricultural commodities through increased price of this common 

resource (this effect should be limited by the presence of co-products in the analysis). On the other, 

demand for food is inelastic and there should be some substitution effects in demand that could 

positively affect the production of other agricultural crops. Production of other crops (rice, vegetable 

and fruit) can be negatively affected but the phenomenon is limited. 

52


Figure 4 Variation of EU Crop Production - 2020 - (volume and percentage) 

14,000.00 

12,000.00 

10,000.00 

8,000.00 

MEU_BAU 

MEU_FT 

1000 tons 

6,000.00 

4,000.00 

2,000.00 

0.00 

-2,000.00 

-4,000.00 

10.00% 

8.00% 

6.00% 

4.00% 

2.00% 

MEU_BAU 

MEU_FT 

0.00% 

-2.00% 

-4.00% 


Figure 5 illustrates how agricultural value-added could be affected by these different scenarios. The 

potential impact of both policies on agricultural value-added is positive in almost all 

countries/regions throughout the world, in particular in the three countries/regions shown on Figure 

5: Brazil, Indonesia and Malaysia, the EU and the US. These policies create more activity in the 

53


agricultural sector and the impact is worldwide. While the mandate is more positive for European 

agricultural value-added than for Brazil and the US, the impact is larger for the US and Brazil. 

Figure 5 Variation of agricultural value-added in 2020 (%) 

1.20% 

1.00% 

0.80% 

MEU_BAU 

MEU_FT 

0.60% 

0.40% 

0.20% 

0.00% 


These gains in agricultural value-added have to be compared with the cost to consumers (consumers 

are negatively affected in the EU) in order to derive a net economic benefit/loss. This is done through 

the calculation of welfare effects of European policies not only for the EU but also for other 

countries/regions as shown in Table 6. The two policies have minimal effects on other 

countries/regions welfare, except for Brazil which benefits from significant improvement in their 

terms of trade thanks to their exporting status of oilseeds for biodiesel and sugar cane. As far as the 

European Union is concerned both policies are neutral: in that sense the increase in agricultural 

added value observed on Figure 5, is offset by negative impact of both policies on consumers’ surplus 

and public receipts. 

54


Table 6. Real Income Impact of European Biofuel Policies, 2020 (Variation / Baseline) 



Brazil 856 857 0.06% 857 0.08% 

CAMCarib 444 444 -0.01% 444 -0.02% 

China 4593 4592 0.00% 4592 -0.01% 

CIS 1093 1091 -0.18% 1091 -0.17% 

EU27 15182 15184 0.01% 15182 0.00% 

IndoMalay 564 564 -0.02% 564 -0.03% 

LAC 1605 1604 -0.05% 1604 -0.06% 

RoOECD 8590 8589 -0.01% 8588 -0.01% 

RoW 5639 5633 -0.11% 5633 -0.11% 

SSA 912 911 -0.12% 911 -0.12% 

USA 15219 15218 0.00% 15218 -0.01% 

World 54697 54687 -0.02% 54684 -0.02% 


5.1.3 Fuel and/or Feed? 

As mentioned earlier the production of biofuels also produces several by-products for which there is 

current or potential demand: Dried Distillers Grains with Solubles (DDGS) obtained from the 

production of ethanol and which is used as animal feed, and oilcakes (animal feeds) from biodiesel 

production. When accounting for by-products, biofuels development should lead to less pressure on 

food markets and in particular on markets for animals feeds. The increased availability of these byproducts 

should have beneficial side effects in other areas of agriculture. A biofuel mandate could 

potentially lead to a positive impact on livestock production in terms of reduced prices for animal 

feed. 

The model used in this analysis includes by-products and illustrates how the development of biofuels 

production can clearly contribute to the consumption of biofuels by-products in cattle and “other 

animal” sectors. Price of meals will decrease by 0.9% to 1%, with the strongest reduction in rapeseed 

cakes. In the DDGS market, the expansion in supply will lead to more substantial price changes (as 

much as -45% for beet pulp in Europe) in the scenario without trade liberalization. This strong result 

is related to the strong bias of the mandate towards ethanol production and the fact that the initial 

DDGS market is very small. Since DDGS in the EU only goes to the domestic market in our model, and 

since new trade flows cannot be generated in our framework, all the initial DDGS production is linked 

55


to biofuel ethanol plants. 18 At the opposite end, when trade liberalization is implemented, EU 

ethanol production, as well as co-products production, is sharply reduced. Since sugarcane ethanol is 

not associated with a by-product in our model, the market is depleted and prices go up. With weak 

substitution effects, the meal prices will decrease less (changes reduced by one-tenth). 

The augmentation of consumption of co-products is driven by more availability of DDGS and oilcakes, 

of which prices are reduced thanks to the EU mandate. 

As illustrated in Figure 6, this is beneficial for the value-added in livestock sectors particularly in the 

European Union where the reduction of prices of these intermediate commodities are more 

significant than elsewhere: the value-added in the cattle sector will increase by almost 0.08% while 

the one for the “Other Animals” sector will be augmented by 0.07%. The results are also positive for 

value-added in the same sectors of the US. Globally the value-added in the cattle sector throughout 

the world is augment by 0.04% (0.03% as far as the “Other animal” sector is concerned). In Brazil, on 

the other hand, the livestock sector will suffer from land competition with the different crops (-0.07% 

of pasture land, see Table 7) and a rising price of soya and other feedstocks . 

Figure 6 Variation of value-added in livestock sectors in 2020 (%) – MEU_BAU scenario 

0.10% 

0.05% 

0.00% 

-0.05% 

Brazil USA EU27 World 

Cattle 

Other Animals 

-0.10% 

-0.15% 


18 It will be interesting to change the elasticity of substitution between DDGS and other energy feed to see if 

the strong results remain. 

56


5.2 Land Use Effects 

5.2.1 Land use 

Changes in crop production, particularly due to the increased demand for feedstock crops used as 

inputs in biofuels, will have different implications on the expected patterns of land use under the 

mandates and trade liberalization scenarios. 

Table 7 indicates the variation in land use by type of land which could be expected from these policy 

scenarios. The amount of cropland is significantly affected in Brazil (+0.54% without trade 

liberalization, +0.77% with trade liberalization, see Figure 7). This result is due to the combination of 

the demand for ethanol (sugar cane) and oilseeds (soya) and the high elasticity of land extension for 

this country. However, due to the AEZ level modeling of land extension, it appears that primary 

forest are not the main source (see Figure 8 and Table 7) of new land for sugar cane production but 

Savannah/Grassland (South East of Brazil). The other regions that are mostly affected are the EU, the 

CIS region, the rest of Latin America and Indonesia-Malaysia. However, since land extension is more 

difficult in these regions (lower elasticity of land extension), the effect is limited. 

Globally the mandate increases cropland use by 0.07% in 2020 and by 0.08% under the trade 

liberalization scenario, with slightly more encroachment into areas reserved for forest. The land use 

changes under the two policy scenarios have implications on CO2 emissions and these are discussed 

in the next section. 

57


Figure 7 Cropland Extension by Region, 2020, Km2 

Brazil CAMCarib China CIS EU27 IndoMalay LAC RoOECD RoW SSA USA 

Mandate with trade liberalization 

EU27: 460 

Brazil 6,866 

Mandate without trade liberalization 

Brazil: 4813 

EU27: 780 

0 2,000 4,000 6,000 8,000 10,000 12,000 

Km2 


Figure 8 Source of Cropland Extension by Type of Land 19 

Brazil 

Forest 

managed 

1% 

Forest 

primary 

15% 

Other 

12% 

EU27 

Pasture 

0% 

Grassland 

0% 

Forest 

managed 

34% 

Savanah 


58% 

Pasture 

14% 

Other 

66% 

Forest 

primary 

0% 


19 These results are based on estimates of past behavior on deforestation in Brazil and we do not consider new preservation 

policies in the central scenario. 

58


Table 7. Variation of Total Land Used (thousands of km²) 

2020 2020 2020 2020 2020 



Cropland Brazil 888.60 893.41 0.54% 895.46 0.77% 

Forest_total Brazil 4391.84 4391.05 -0.02% 4390.78 -0.02% 

Pasture Brazil 1371.17 1370.49 -0.05% 1370.21 -0.07% 

SavnGrasslnd Brazil 1838.39 1835.61 -0.15% 1834.35 -0.22% 

Cropland China 1421.29 1421.37 0.01% 1421.37 0.01% 

Forest_total China 2112.52 2112.45 0.00% 2112.45 0.00% 

Pasture China 1083.30 1083.30 0.00% 1083.30 0.00% 

SavnGrasslnd China 1927.67 1927.67 0.00% 1927.67 0.00% 

Cropland EU27 1004.03 1004.81 0.08% 1004.49 0.05% 

Forest_total EU27 1449.27 1449.00 -0.02% 1449.11 -0.01% 

Pasture EU27 617.18 617.17 0.00% 617.18 0.00% 

SavnGrasslnd EU27 205.20 205.20 0.00% 205.20 0.00% 

Cropland IndoMalay 344.41 344.55 0.04% 344.55 0.04% 

Forest_total IndoMalay 867.13 867.04 -0.01% 867.04 -0.01% 

Pasture IndoMalay 34.05 34.02 -0.08% 34.02 -0.08% 

SavnGrasslnd IndoMalay 138.54 138.54 0.00% 138.54 0.00% 

Cropland LAC 397.51 397.91 0.10% 397.92 0.10% 

Forest_total LAC 3294.18 3294.07 0.00% 3294.07 0.00% 

Pasture LAC 794.01 794.07 0.01% 794.07 0.01% 

SavnGrasslnd LAC 2213.70 2213.70 0.00% 2213.70 0.00% 

Cropland World 12425.91 12434.11 0.07% 12435.66 0.08% 

Forest_total World 37704.94 37703.17 0.00% 37703.05 0.00% 

Pasture World 10870.45 10869.46 -0.01% 10869.26 -0.01% 

SavnGrasslnd World 29860.28 29857.50 -0.01% 29856.25 -0.01% 


Note: The land category “Other” is not displayed on the table. 

An interesting question which is related to the expansion of cropland is the relative decomposition of 

production increase between yield changes and extensive land use. Table 8 provides such a 

decomposition at the world level for each crop. For instance, in the pure mandate case, the world 

increase of 0.91% of rapeseed production is achieved by increasing land by 0.54% and by increased 

use of new capital and labour per Ha (0.34%); intensification of fertilizer used plays only a minor role. 

At the other hand, we see that for wheat the production increase is achieved completely by 

intensification, through increased use of fertilizers and through factor intensification. 

59


Table 8 Decomposition of production increase 

MEU_BAU 

Yield Yield Land 

use 

Factors Fertiliser 

Change 

increase 

MEU_FT 

Total Yield Yield Land 

use 

Producti Factors Fertiliser 

Change 

on increase 

increase 

Total 

Producti 

on 

Increase 

Rapeseed 0.32% 0.04% 0.54% 0.90% 0.34% 0.02% 0.61% 0.97% 

PalmFruit 0.10% 0.21% 0.31% 0.10% 0.20% 0.30% 

Maize 0.04% 0.03% 0.01% 0.08% 0.03% 0.03% -0.01% 0.05% 

OthCrop 0.01% 0.00% 0.00% 0.01% 0.02% 0.02% -0.01% 0.03% 

OthOilSds 0.01% 0.01% -0.03% -0.01% 0.01% 0.02% -0.03% 0.00% 

Rice 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 

Soybeans 0.04% 0.06% 0.12% 0.22% 0.05% 0.07% 0.15% 0.27% 

Sugar_cb 0.66% 0.54% 2.67% 3.87% 0.62% 0.37% 3.98% 4.97% 

Sunflower 0.11% -0.10% 0.37% 0.38% 0.11% -0.10% 0.39% 0.40% 

VegFruits 0.00% 0.05% -0.06% -0.01% 0.00% 0.05% -0.06% -0.01% 

Wheat 0.06% 0.05% 0.00% 0.11% 0.00% 0.04% -0.09% -0.05% 


5.2.2 Emissions 

As displayed in Table 9, the sum of land use related emissions implied by the European mandate is 

107 million tons of CO2 equivalent in 2020 without trade liberalization and 118 million with 

elimination of MFN duties on biodiesel and ethanol. Even without trade liberalization, most of the 

emissions effects (between 50% and 60% of world emissions) are concentrated in Brazil where these 

are driven by demand for sugar and soybeans. However, we see that emissions related to 

deforestation represent just a share (between half and one third) of Brazilian emissions. Modeling 

the land extension at the AEZ level shows that forest is less impacted than other biotopes (grassland) 

due to the extension of sugar protection. Without trade liberalization the EU is the second region in 

terms of direct emissions (nearly 10.63 Mios tCO2eq). Trade liberalization allows the EU to cut its 

direct emissions by 40% but the CIS and Brazil will emit much more. Taking peatlands into account 

plays a minor role in the broad picture (up to 1.1% in the case were largest emissions figures are 

used). But if we compare these additional figures to the other CO2 emissions of Indonesia and 

Malaysia, we see that these figures can add 25% to overall emissions of this region, acknowledging 

the fact that it remains a minor supplier for the EU (less than 10% of EU biodiesel consumption when 

we add biodiesel imports and palm oil imports) and that the mandate target implies limit increase in 

biodiesel consumption. 

60


Table 9. Indirect land use emissions related to biofuels in 2020 

(Mios tCO2eq - extra emissions are positive values) 

Forest 

Biomass 

change 

5.6% EU Mandate 5.6% EU Mandate + Full trade 

liberalization on biofuels 

Organic Total land Forest Organic 

Carbon in use Biomass Carbon in 

Mineral Soil emissions change Mineral Soil 

Total land 

use emissions 

Brazil 23.97 33.33 57.30 28.50 46.02 74.52 

CAMCarib 0.52 0.52 0.22 0.22 

China 1.57 0.65 2.22 1.43 0.60 2.03 

CIS 3.18 5.08 8.26 2.91 4.52 7.43 

EU27 3.03 7.60 10.63 1.80 4.50 6.30 

IndoMalay 3.39 1.53 4.92 3.38 1.53 4.90 

LAC 2.63 3.58 6.21 2.71 3.70 6.41 

RoOECD 1.08 2.47 3.55 0.87 2.34 3.22 

RoW 1.20 0.94 2.14 0.88 0.71 1.59 

SSA 1.49 4.50 5.99 1.36 4.04 5.41 

USA 1.88 2.89 4.76 2.24 3.47 5.71 

World 43.41 63.09 107.50 46.07 71.66 117.74 

Additional MtCo2 emissions from peatlands IPCC method 0.17 

Values are indentical in both scenarios at 0.01 MtCO2eq Couwenberg(2009): 1.38 


As shown in Table 10, the sum of direct emissions reductions 20 generated by the substitution of 

fossile fuel by biofuels and implied by a European liberalization of trade in ethanol and biodiesel is 

slightly higher: -21 million tons of CO2 equivalent in 2020 under the trade liberalization scenario 

instead of-18 Mios. This result is driven by the increased use of sugar cane ethanol that is the most 

efficient feedstock. The net emissions balance (land use emissions minus direct emission savings) is 

positive and slightly larger under the liberalization case than under the pure mandate scenario. Even 

if the liberalization leads to more emissions through indirect land use effects, using efficient 

imported biofuels delivers a net missions reduction in a 20 year period. 

20 Each MJ of fossil fuel is assumed to generate 25gr of carbon, i.e. about 92 gr. of CO2. 

61


Table 10 Emissions balance. Annualized figures. CO2 Mto2 eq. 

MEU_BAU 

MEU_FTA 

Direct 

emissions 

Land use 

change 

Total 

emissions 

Direct 

emissions 

Land use 

change 

Total 

emissions 

Brazil -0.05 2.87 2.82 -0.06 3.73 3.67 

CAMCarib -0.32 0.03 -0.29 0.24 0.01 0.25 

China -0.02 0.11 0.09 -0.02 0.10 0.08 

CIS 0.00 0.41 0.41 0.00 0.37 0.37 

EU27 -18.36 0.53 -17.83 -21.24 0.31 -20.93 

IndoMalay -0.01 0.25 0.24 -0.01 0.25 0.24 

LAC 0.01 0.31 0.32 0.01 0.32 0.33 

RoOECD 0.12 0.18 0.30 0.21 0.16 0.37 

RoW 0.02 0.11 0.13 0.02 0.08 0.10 

SSA 0.00 0.30 0.30 0.00 0.27 0.27 

USA 0.45 0.24 0.69 0.72 0.29 1.01 

World -18.17 5.33 -12.84 -20.11 5.89 -14.22 


Note: Land use emissions column is based on Table 9 figures divided by 20 (years). 

The emissions credit is attributed to the country that consumes the biofuel. 

Additional peat lands emissions are not included in this table. 

Table 11 displays the carbon balance sheet of the 5.6% mandate under our different scenarios. The 

upper part of the table displays the total carbon release (from forest biomass and soil contents) due 

to the change in land use during the 2008-2020 period following the implementation of the mandate. 

The lower part shows average ILUC effect computed with our model equal to the sum of carbon 

release from forest biomass and soil carbon content. All annual coefficients take the stock value of 

the upper table and divides them by 20 years and divided by the increase in EU consumption of 

biofuels. The average ILUC computed here is between 17.7 gCO2eq/Mj (no trade liberalization) and 

19.5 gCO2eq/Mj (with trade liberalization).The net emission balance on a 20-year period is about - 

42.82gCO2/MJ if the mandate is not associated with an open trade policy and slightly more under 

trade liberalization (-46.93 gCO2/MJ). These coefficients are average values since they are based on 

the full mandate increase (from 3.3% to 5.6%) and takes into consideration all the direct and indirect 

effects in the CGE framework in terms of income and substitution effects. But they do not include 

CO2 variations not related directly to the biofuel policies (such as the income effect on the steel 

industry). 

62


Table 11. Carbon balance sheet 

2020 2020 2020 


Total carbon release from forest biomass 

(MtCO2eq) 43.41 46.07 

Total carbon release from organic carbon in 

mineral soil (MtCO2eq) 63.09 71.66 

EU Consumption of biofuel in 2020 (million GJ) 443 743 746 

Annual carbon release from forest biomass 

(gCO2eq/MJ) 7.23 7.61 

Annual carbon release from organic carbon in 

mineral soil (gCO2eq/MJ) 10.50 11.84 

Annual direct savings (gCO2/MJ) -60.55 -66.38 

Total emission balance on a 20 years period 

(gCO2/MJ) -42.82 -46.93 


5.2.3 Crop specific ILUC 

Applying the method described in Annex VII, we can also compute the marginal ILUC coefficient for 

each crop. In this case, we investigate the marginal effect of the 5.6% mandate by increasing the 

demand for biofuel in the EU27 by a marginal amount of 1 million GJ in the 2020 (about 0.1% of the 

EU consumption level in 2020) situation and allowing the corresponding increase in biofuel (domestic 

or imported) production to come from one feedstock only. We compute the marginal effect for each 

feedstock at the end of the mandate in 2020. Table 12 displays the coefficient of emissions from land 

use changes for the eight feedstocks, for ethanol – without constraint on the feedstocks - and 

biodiesel. Figures are provided with and without the peatland effects. Concerning the later, we use a 

simple average of the IPCC and Couwenberg coefficients. 

Results show that sugarcane and sugarbeet, with the lowest marginal ILUC, are the most efficient 

feedstocks in terms of land use under the mandate scenario. The average ethanol coefficients from 

these two feedstocks are between 16 and 19 gCO2/Mj with a life cycle of 20 years. For wheat and 

sugar beet, under trade liberalization the ILUC effect increased. Since the EU will always outsource is 

supply of sugar cane ethanol in Brazil, the trade liberalization scenario has a very limited effect on 

the sugar cane coefficient. 

Concerning biodiesel, even if peat land emissions are considered, palm oil is the most efficient 

feedstock, although still at a level three times above the emission levels for sugar cane ethanol. Palm 

oil appears as an efficient feedstock and can compete with crops for two reasons: it produces coproducts, 

even in limited quantity and has a very high oil yield (up to six times the rapeseed yield by 

hectare). The average biodiesel coefficients (between 54gCO2/Mj and 58gCO2/Mj) are between 

63


rapeseed oil and the soybean oil. The latter is the most costly biodiesel in terms of ILUC since the 

soya market puts a lot of pressure on land extension in Brazil. 

Table 12 Marginal Indirect Land Use emissions, gCO2/MJ per annum. 20 years life cycle. 

MEU_BAU 

MEU_FT 

Without With Peatland Without Peatland With Peatland 

Peatland effects 

effect 

effect 

effect 

Ethanol 17.74 17.74 19.16 19.18 

Ethanol SugarBeet 16.07 16.08 65.48 65.47 

Ethanol SugarCane 17.78 17.78 18.86 18.86 

Ethanol Maize 54.11 54.12 79.10 79.15 

Ethanol Wheat 37.26 37.27 16.04 16.12 

Biodiesel 58.67 59.78 54.69 55.76 

Palm Oil 46.40 50.13 44.63 48.31 

Rapeseed Oil 53.01 53.68 50.60 51.24 

Soybean Oil 74.51 75.40 67.01 67.86 

Sunflower Oil 59.87 60.53 56.27 56.89 


Note:The marginal coefficient is computed in 2020 after the implementation of the 5.6% mandate. 

Compared to the average ILUC coefficients reported in Table 11, the figures in Table 12 are slightly 

different. We can provide two explanations. First, we are dealing with marginal coefficients that are 

expected to be above the average due to the decreasing marginal productivity embedded in the 

model (see next section). Second, as previously discussed, the mandate is mainly driven by an 

increased consumption of ethanol. As shown in the production figures, this ethanol will be produced 

from sugar cane (imports) and sugar beet, the most efficient feedstock in terms of land use. 

The marginal ILUC effects reported in Table 12 combine with direct emissions reductions to generate 

the net emissions balance reported in Table 13. Sugar cane, Sugar beet and Wheat ethanol will 

generate marginal net emissions savings (negative emissions) under both the 5.6% mandate and the 

trade liberalization scenario, with the strongest effect for Sugar cane. For biodiesel, only palm oil will 

generate emission savings. 21 

21 Under the central assumption here that palm oil direct savings coefficient is 61%. 

64


Table 13 Marginal Net Emissions by Feedstock. gCO2/Mj. 20 years life cycle. 

MEU_BAU 

MEU_FT 

Without Peatland With Peatland Without Peatland With Peatland 

effects 

effect 

effect 

effect 

Ethanol -49.69 -49.68 -53.55 -53.53 

Ethanol Sugar Beet -35.86 -35.85 21.84 21.83 

Ethanol SugarCane -53.95 -53.95 -55.53 -55.53 

Ethanol Maize 3.64 3.65 62.82 62.87 

Ethanol Wheat -7.00 -6.99 -5.02 -4.95 

Biodiesel 5.95 7.06 3.63 4.70 

Palm Oil -21.98 -18.25 -22.43 -18.76 

Rapeseed Oil 8.76 9.42 7.42 8.06 

Soybean Oil 24.07 24.96 18.95 19.80 

Sunflower Oil 8.73 9.38 7.74 8.37 


Note: Negative figures represent an emission reduction, positive values represent an emission increase. 

5.3 Sensitivity Analysis 

This section discusses two aspect of the sensitive analysis done in this study: 

The policy target; 

The value of key parameters. 

On the later issue, we only study alternative cases but a richer and systematic analysis should be 

performed in future research. 

5.3.1 Alternative Mandate Targets 

We compute the average ILUC of the mandate for five levels of mandatory blending in the EU: 4.6%, 

5.6%, 6.6%, 7.6% and 8.6% for the two main trade scenarios: status quo (Figure 9) and trade 

liberalization (Figure 10). 

As expected, the direct emission saving coefficient is reduced as the level of the mandate increases. 

Greater pressure for biofuel production from a higher target results in increasing use of less efficient 

feedstock. Similarly, starting with trade liberalization and a low mandate, the EU will import primarily 

sugar cane ethanol and with the increasing pressure on this feedstock, domestic sources of ethanol 

will become more attractive and the biofuel mix will become less efficient in terms of direct savings. 

65


Figure 9 Indirect land use emissions and direct savings for different mandate levels, No change in trade policy 

40.00 

4.60% 5.60% 6.60% 7.60% 8.60% 

20.00 

0.00 

-20.00 

-40.00 

-60.00 

-80.00 

Annual carbon release (gCO2eq/MJ) 

Annual direct savings (gCO2/MJ) 

Total emission balance on a 20 years period (gCO2/MJ) 

T 



Figure 10 Indirect land use emissions and direct savings for different mandate levels, Free trade 

scenario 

40.00 

4.60% 5.60% 6.60% 7.60% 8.60% 

20.00 

0.00 

-20.00 

-40.00 

-60.00 

-80.00 

-100.00 

Annual carbon release (gCO2eq/MJ) 

Annual direct savings (gCO2/MJ) 

Total emission balance on a 20 years period (gCO2/MJ) 



Concerning the ILUC emissions, we see a net increase of the adverse effects of the biofuel demands 

on land use as the level of the mandate increases. A 4.6% mandate could be achieved without 

66


noticeable land use impact, however any level above this point starts to generate emissions. Moving 

from 4.6 to 6.6 % will increase sharply the average emissions to reach 25gCo2/Mj. A 8.6% mandate 

without trade liberalization will cut by nearly half of the emissions savings under the 4.6% mandate. 

However, the total emissions balance remains positive for all the level of the mandate considered 

here. 

A key issue in this research is the question of whether the non-linear ILUC is just a feature of the 

model or whether it also reflects an underlying reality. First, the evolution of the size of the mandate 

leads to an evolution in the biofuel mix: no additional biodiesel is needed at 4.6% when about 5Mtoe 

of biodiesel is required by a 8.6% mandate. Since biodiesel is less emissions friendly, the average 

effect deteriorates. Second, nonlinearity of the ILUC effect can be expected from the modeling 

framework. Several mechanisms contribute to this effect: 

The capacity to substitute one type of land for another: it is represented by the concavity of 

the CET function in the land use module. The marginal productivity of one hectare moving 

from one sector to another is declining quickly with the low elasticity used. The first unit of 

land planted to barley can be transformed “easily” to wheat for instance, but this marginal 

transformation ratio is deteriorating. From the modeling point of view, the CET framework is 

not totally satisfactory but it remains the mainstream approach in the literature. However, 

how can we explain in the reality that farmers continue to have diversified productions, even 

if the price of one commodity dominates the other. Even when the wheat price is high, not 

all land in Europe is not shifted to wheat. There are many possible reasons for this: desire of 

diversification from farmers, real differences in land quality for the different crops, short 

term perception vs long term perception etc. Overall, they will lead to the same 

consequences: if farmers shift “some” units of land to the expanding crops easily, they will 

not do it in a linear way. They will stop converting eventually, and if they want to produce 

more of one crop, they will go for “new” land, while keeping their other production at a 

certain level. It means that substitution is non linear and that there is more pressure on new 

land with the increase in magnitude of demand from biofuels. A similar mechanism applies to 

pasture and forest that is converted to cropland. There is limited substitution (and non 

linearity due to the CET effect). It represents the fact that (a) pasture and forestry land 

converted to cropland have decreasing marginal productivity, (b) there are institutional 

factors that could hinder the conversion of these lands to cropland. 

The rigidity of other sectors to reduce part of their own consumption of feedstocks. The 

capacity of other sectors, and final consumers, to reduce their consumption level of 

feedstocks is also non linear (and represented by CES function). If they can initially forego a 

67


few units easily (e.g. Palm oil by cosmetic industry), their marginal propensity to do so 

declines quickly (=their marginal cost to do it increase). In a symmetric way, the absorption 

capacity for co-products by the livestock sector is disputable. Is it linear or not? In the model, 

it is not. But it seems also that in the “real” word, people argue about the limit in DDGS, or 

meals (at least one type of meal) in the animal feed. 

The saturation effect on fertilizers. 

The below-average productivity assumed for new units of land. 

Every model is an abstraction of reality but should, at the same time, represent the essential features 

and behavior of that reality as correctly as possible. The non-linear features in this model are widely 

used in most biofuels models and indeed in most (agro-)economic models. There is sound economic 

rationale behind these behavioral assumptions. Abandoning decreasing returns would go against 

economic logic and common sense. On the other hand, it is difficult to estimate how strong these 

decreasing returns effects should be. The available empirical evidence is limited and often very 

different estimates for key parameters are available. There are two options here: extensive 

sensitivity analysis on key parameters (which we do below) and collecting more robust empirical 

evidence. The latter is outside the scope of this research project and may take many years to 

complete. 

5.3.2 Land substitution 

Both sensitivity analyses (doubling the elasticity of substitution between crops, and alternatively, the 

elasticity of substitution between cropland and pasture) have very similar results. Emissions are 

reduced by 10% on average. Marginal ILUC is reduced by 30% since this parameter plays a key role in 

defining the marginal productivity profile for the crops. 

5.3.3 Land extension 

If we apply Brazil's land extension elasticity to Indonesia and Malaysia, i.e. 0.10 instead of 0.05, the 

ILUC effects will be stronger in this region. Emissions increase by about 4 millions of CO2eq and the 

marginal ILUC of palm oil increases by 10%, reaching the same level as for rapeseed oil. 

If land extension elasticity in Brazil is reduced by half, global ILUC emissions are reduced by one-third 

and the total emissions balance improves. Brazilian exports to the EU are not significantly affected 

since land is taken from other sectors and production becomes more intensive. 

68

69 



6 Concluding Remarks 

This section summarizes our main findings and then provides some recommendations for future 

research. 

6.1 Lessons Learned 

The main lesson learned is that ILUC does indeed have an important effect on the environmental 

sustainability of biofuels. However, the size of the additional EU 2020 mandate, under current 

assumptions regarding the future evolution of renewable energy use in road transport, is sufficiently 

small (5.6% of road transport fuels in 2020) and does not threaten the environmental viability of 

biofuels. If the underlying assumptions should change however, either because the mandated 

quantities turn out to be higher and/or because the model assumptions and parameters need to be 

revised, there is a real risk that ILUC could undermine the environmental viability of biofuels. Nonlinear 

effects, in terms of biofuels volumes and behavioural parameters, pose a risk. 

At the same time, this biofuels modeling project has demonstrated how the current limits to data 

availability create significant uncertainty regarding the outcomes predicted by these policy 

simulations. The model represents a state of the art simulation of the real world, but more data 

collection work will be required to reduce this margin of uncertainty. 

In terms of trade policy, the main result is that biofuels trade liberalization would lead to slightly 

more ILUC effects through deforestation outside the EU (especially in Brazil). But this is compensated 

by the use of a more efficient biofuel (sugar cane ethanol) that improves emissions savings and 

results in animproved CO2 emission balance. At the same time such an effect can take place only if 

we assume that the share of ethanol in total biofuel consumption can increase drastically from 19% 

to 45% by 2020. 

Effects on food prices will remain limited (maximum +0.5% in Brazil, +0.14% in Europe). Although EU 

biofuel policy has no significant real income consequences for the EU, some countries may 

experience small negative effects, particularly oil exporters (-0.11% to -0.18% of real income by 2020) 

and Sub-saharan Africa (-0.12%) due to the fall in oil prices and rise in food prices, respectively. 

Analysis of ILUC by crop indicates that ethanol, and particularly sugar-based ethanol, will generate 

the highest potential gains in terms of net emissions savings. For biodiesel, palm oil is the efficient 

feedstock in terms of CO2 emissions, even if peatland emissions are taken into account. 

From a methodological point of view, our study confirmed that yield response and land substitution 

elasticities play a critical role in our assessment. The potential non-linearity of ILUC coefficients was 

70


also demonstrated. However, our main conclusions remain robust to the sensitivity analyses 

performed at this stage. We have also confirmed the importance of having a high quality database 

with the need of linking the value and the quantity matrix to feed the model with marginal rates of 

substitution that are relevant. In terms of policy design, taking into account the biofuels mandates in 

other economies was important to limit the capacity of the EU to absorb foreign production. 

However, we have limited our analysis to a conservative case (5% mandates for China, Canada, 

Japan, Australia, New Zealand, Switzerland, Indonesia and Indonesia) and a stronger constraint may 

lead to higher ILUC impact. 

Even more important is the role of the mix between ethanol and biodiesel. Depending on the 

flexibility allowed for the ratio between the two biofuels, land use effects and trade policy effects can 

be very different. 

6.2 Suggestions for Further Research 

Based on our analysis, we can underline a few directions for future research. 

First, due to strong impact of the non linearity on our results, assessing the relevance of this behavior 

is critical. On one hand, new modeling approach should be introduced to as an alternative to the CET 

framework of land reallocation. Modeling explicit conversion costs (fixed costs) will allow explaining 

the short term low elasticity of substitution existing in the literature and will be compatible with 

stronger marginal productivity (no yield decrease) in the long run. As biofuel policies are expected to 

be long term policies, this later approach seems reasonable. At the same time, more econometric 

work is needed to estimate the behavior of EU farmers in the short and long run, in particular in the 

context of the more market-oriented CAP. Similarly, assessing the relevance of the assumption on 

decreasing marginal productivity of new land plays an important role here. 

Second, our modeling of land extension at the AEZ level allows for the consideration of different 

extension coefficients for different regions within a country. With this feature, it will be beneficial to 

have access to more detailed data for an extended set of countries (beyond Brazil). 

Third, different assumptions on the mix between biodiesel and ethanol should be studied. 

Fourth, the role of certifications, the emergence of differentiated biofuels, crops and land prices 

based on their “carbon” contents, and direct savings coefficients, should be studied to understand to 

which extent minimum requirements in the EU legislation impact the market. 

Fifth, the modeling of endogenous yield increases, based on research and development activities 

may be useful to limit the land use effects. 

71


Sixth, more critical is the need to improve the overall quality of data for the EU27. In this exercise, 

aside from introducing new sectors in the database, considerable effort was spent in correcting some 

inconsistencies and upgrading the GTAP7 database. However, the quality of the original social 

accounting matrix for the EU in the GTAP7 is very weak and some strange intersectoral linkages 

remain. Moving to the latest GTAP7.1 (recently released in mid-February 2010) that includes updated 

EU SAMs based on the JRC AgroSams and benefiting from the CAPRI input/outputs information 

appears to be a strong requirement to provide a accurate analysis for the European Union, 

particularly in looking at domestic policies. 

Seventh, a higher level of geographical disaggregation is needed to gain a better understanding of 

land use effects, e.g. having Canada and Australia in one region leads to an important loss of 

information in terms of production allocation and elasticity of supply, but also of the carbon content 

of different biotopes. 

72


7 ANNEXES 

Annex I. Construction of the Global Biofuels Database 

External data for 2004 on production, trade, tariffs and processing costs of the new sectors, 

especially for ethanol and biodiesel, for use in splitting these sectors from GTAP sectors were 

compiled from published sources, FAO stats and from the BACI databases. The primary feedstock 

crops used in the production of liquid biofuels in the major producing countries were identified from 

available literature. The input-output relationships in each biofuels producing country in the GTAP 

database were then examined to determine the feedstock processing sector from which the new 

ethanol and biodiesel sectors could be extracted. Since the global database is comprised of national 

social accounting matrices (SAMs) which are from different years of data, some of which reflect 

outdated agricultural production relationships, the global database was adjusted using agricultural 

input-output relationships developed from FAO data 22 . 

The database has been developed on a mix 2004 and 2007 data to ensure enough maturity in the 

biofuels sector (especially trade pattern) 

Ethanol 

Data on ethanol production for 2004 and 2007, in millions of gallons, were obtained from industry 

statistics provided by the Renewable Fuels Association for annual ethanol production by country. 23 

The data covers 33 individual countries plus a sum for “other countries”. Producer costs structure are 

extracted from OECD (2008) from which data on ethanol processing costs for the major ethanol 

producers (USA, Brazil, EU) were compiled. Bilateral trade for ethanol byproduct in 2004 and 2007 

was obtained from the reconciled BACI trade database which is developed and maintained at CEPII. 

Depending on the country, the ethanol sector was carved out either from the sugar (SGR) sector, the 

other food products (OFD) sector, or the chemicals, rubber and plastics (CRP) sector and then 

aggregated to create one ethanol sector. Ethanol producers were first classified according to the 

primary feedstock crops used in production. The input-output accounts in the GTAP database were 

then examined for each ethanol producer to determine which processing sector used a large 

proportion of the feedstock as intermediate input. This is then the processing sector that is split to 

create the ethanol sector in that country. For example, a large share of sugarcane production in 

22 The food and agricultural input-output database is documented in Peterson (2008). 

23 See: http://www.ethanolrfa.org/industry/statistics/#EIO citing F.O. Licht. Renewable Fuels Association, Homegrown for 

the Homeland: Industry Outlook 2005, (Washington, DC: 2005), p. 14. 

73


Brazil goes to an established sugar ethanol processing sector, which is incorporated in GTAP’s 

chemicals, rubber and plastic (CRP) sector in the Brazilian I-O table. Thus CRP is the sector that was 

split in Brazil to extract the sugar ethanol sector. However, similar analysis indicated that it was the 

sugar processing (SGR) sector that should be split in other sugar ethanol producing countries in Latin 

America. Production of grain-based ethanol in the United States, Canada and in the European Union 

was introduced in the data by splitting the other food products (OFD) sector where wheat and cereal 

grain processing takes place. 

Total consumption of ethanol in each region was computed from the data on production, total 

exports and total imports. Ethanol was assumed to go directly to final household consumption and 

not as an intermediate input into production. Production cost data in terms of the share of 

feedstock, energy and other processing costs were used to construct technology matrices for 

ethanol. These vary by country depending on the primary feedstock used in production. 

In details, each feedstock is the only agricultural inputs of a sub ethanol sectors. Each ethanol sectors 

will produce ethanol and and a coproduct (DDGS) except the sugar cane sector. They all share the 

same techonology (intermediate consumptions, labor) except the sugar cane sector that is less 

energy intensive (cogeneration). 

All the sub-ethanol sectors sell their liquid ethanol to a supra ethanol sector that collects the 

different varieties and provide its output to final consumers, intermediate consumption for the road 

transportation sector and to export markets. 

The international trade of ethanol is classified in the Harmonized System (HS) under HS6 codes 

220710 and 220720 which cover undenatured and denatured ethyl alcohol, respectively. Since it is 

difficult from trade information to know the exact use of ethanol (agrifood, industry or biofuel), we 

prefer to rely on trade figures from F.O. Litch. Although ethanol production from different feedstocks 

is introduced by splitting the appropriate food processing sectors (SGR, OFD, CRP), as guided by the 

input-output relationships for each region, ethanol trade is actually classified under trade of the 

GTAP beverages and tobacco (B_T) sector. It is the B_T sector that we split to take bilateral ethanol 

trade and tariff information into account. 

Concerning the EU tariff on ethanol, we assume an average ad valorem equivalent of 50%. However, 

the effective AVE is difficult to compute since tariffs on the two types of ethanol are significantly 

different and the mix difficult to define. In addition, some Member States are not applying the 

specific tariffs of 220710 and 220720 but a lower one considering ethanol for biofuel as a nonagricultural, 

chemical input. 

74


Biodiesel 

Data on biodiesel production in the European Union, in million tons, were obtained from published 

statistics of the European Biodiesel Board. 24 Biodiesel production data for non-EU countries for 2004 

was estimated based on 2007 production data for these countries, obtained from F.O. Licht 25 , 

deflated using 2004-2007 biodiesel production average growth rate for the EU. The volume data 

were converted to US$ millions using 2004 price data. Information on biodiesel processing costs was 

obtained from the OECD (2006). The international trade of biodiesel is classified in under the HS 3824 

position, mainly under 382490. Once again this product includes non fuel-related imports that make 

difficult any direct use. Therefore, we combine HS6 trade flows from BACI, 8- or 10- digit trade flows 

from the US and the EU trade data and rescale flows to match F.O. Licht estimates. 

The biodiesel industry is created from the CRP sector in GTAP and relevant feedstocks are extracted 

from the OFD and CRP sectors depending on the initial IO links. The technology of the sector (share 

costs) is based on OECD (2008) report without any significant difference across countries, except for 

the nature of the feedstock (type of vegetal oil) used. 

Maize 

The most important feedstock crops for biofuel production have to be treated separately in the 

database in order to more accurately assess the impact of biofuels expansion on feedstock 

production, prices and on land use. Wheat and sugarcane\sugar beet are both separate sectors in the 

GTAP database. Maize (corn), however, is classified under the GTAP cereal grains sector which 

include crops that are not used as feedstock in biofuels production. The GTAP cereal grains (GRO) 

sector was split to create the maize (MAIZ) and other cereal grains (OGRO) sectors. Maize production 

volume and price data for 2004, as well as production data for other cereals (barley, buckwheat, 

canary seeds, fonio, millet, mixed grains, oats, and cereal grains, nes) were compiled from FAO 

Production Statistics. 26 This allowed us to compute the shares of maize production to total cereal 

grains production in each country. Similarly, bilateral trade data from the BACI trade database for 

maize and for the GTAP GRO sector allowed us to compute trade shares for maize trade to total GRO 

trade for each bilateral trade flow. We then used the production shares information and trade shares 

information to split the GRO sector into MAIZ and OGRO. We assume that the production technology 

for MAIZ and OGRO in each country are the same as those used for the original sector, GRO. 

24 Available online at: http://www.ebb-eu. org/stats.php. 

25 As cited in OECD (2008). 

26 Available online at: http://faostat.fao.org/site/567/default.aspx. 

75


Oilseed crops (Palm nut, Rapeseed, Soybeans, Sunflower Seed) 

For oilseeds, we compile 2004 production volume and prices data from FAO Production Statistics for 

the oilseed crops that are significant feedstocks for biodiesel production (palm nut, rapeseed, 

soybeans, sunflower seed) as well as for other oilseed crops. Bilateral trade data for oilseeds used in 

biodiesel, as well for the GTAP OSD sector, were obtained from the BACI trade database. 

The different oilseeds are extracted from the OSD sector proportionally to their production value. No 

technology differences are assumed across them. In several cases, the OSD sector in GTAP was too 

small to accommodate production value estimated based on FAO statistics. In this case, we extract 

resources from the OCR (other crops) sector. 

Vegetable Oils (Palm oil, Rapeseed oil, Soybeans oil , Sunflower Seed Oil) 

For vegetable oils, we compile 2004 production volume and prices data from FAO Production 

Statistics for the vegetable oils that are used for biodiesel production (palm oil, rapeseed oil, soybean 

oil, sunflower seed oil) as well as for other oilseed crops. Bilateral trade data for oilseeds used in 

biodiesel, as well for the GTAP OSD sector, were obtained from the BACI trade database. In addition 

to the oils value, we add the co-products value in each subsector. Each subsector technology is 

defined on the relevant crushing technology where only one oilseed is used to produce one type of 

vegetal oil. 

Fertilizer 

Fertilizers are part of the large CRP sector in GTAP. A separate treatment of fertilizers is necessary to 

more adequately assess the implications of biofuels expansion on the interactions between fertilizers 

and land in crop production. The production values for 2004 for nitrogen, phosphate and potash 

fertilizers were obtained from production and prices data from the FAO Resource Statistics and from 

published data. 27 Bilateral trade data for fertilizers and for the GTAP CRP sector were obtained from 

the BACI database. Tariff data were obtained from the 2004 MAcMap database. The fertilizer 

production values and trade shares information were used to split the CRP sector into FERT and 

CRPN. We adapt an average production technology for fertilizers based on the detailed US inputoutput 

table and we assume that fertilizers are used only as an intermediate input in the crop 

production sectors. 

27 FAO fertilizer production data available online at: http://faostat.fao.org/site/575/default.aspx. Price data obtained were 

from: http://www.farmdoc.uiuc.edu/manage/newsletters/fefo08_13/fefo08_13.html 

76


Transport Fuel 

Fuels used for transport are part of GTAP’s petroleum and coal sector (P_C). A separate treatment of 

transport fuels is necessary to provide a better assessment of the likely substitution between 

transport biofuels and transport fuels from fossil fuels. Data on the value of consumption of fossil 

fuels 28 was used along with trade data to obtain the value of transport fuel production by country. 

Bilateral trade data and tariffs for transport fuel were obtained from the BACI and MAcMap 

databases, respectively. The transport fuel production values and trade shares information were 

used to split the P_C sector into TP_C and OP_C. We assume that the production technologies for 

TP_C and OP_C in each country are the same as those for the original sector, P_C. However, we 

assume that in contrast to OP_C, TP_C is the main fuel product comprising 90 percent of fuels used 

as intermediate input in the GTAP transport sectors (land, water and air transport) and in final 

household demand. TP_C and OP_C are equally split as fuel inputs used in the production of all other 

sectors. 

28 From national fuel consumption data reported in (Metschies) International Fuel Prices 2005, 4th edition, available at: 

http://www.international-fuel-prices.com. 

77


Annex II. Modeling Energy and Agricultural Processes of Production 

The MIRAGE model has been expanded to address its shortcomings in the energy sector and thus 

better adapt it to the specific needs of the study. It has been undertaken following a literature 

review. This review reveals the existence of two main approaches to energy modeling in the 

literature. 

The “top-down” approach focuses on the modeling of macroeconomic activity and international 

trade and derives energy demand from the activity implied by this modeling. Burniaux and Truong 

(2002) for example develop an energy version of the GTAP model (the GTAP-E model) and use it to 

study the impact of alternative implementations of the Kyoto Protocol on welfare and terms of trade 

in eight regions of the world. 

A bottom-up approach places a lot of emphasis on the technical description of the energy sector and 

provides a more realistic and detailed modeling of energy efficiency. It selects the most efficient 

process of energy production corresponding to a certain level of energy demand. For example the 

MEGABARE model (ABARE, 1996) makes use of the technology bundle approach which introduces 

substitutability between different technologies (for example between the electric arc furnace and 

the basic oxygen furnace in the steel industry) while the use of a specific technology implies a 

Leontief combination of primary factors and intermediate consumption. 

Although this kind of approach is much more difficult to implement on a large scale, it provides very 

interesting elements. For example the substitutability of capital and energy depends on whether the 

model is used in a short or long term perspective. Following an energy price increase, in the short 

term energy and capital are complementary while in the long term a new technology could be 

adopted which utilizes more capital and less energy. Attention needs to be paid to this aspect. Finally 

it is possible to envisage combining the two approaches. The CETM model for example (Rutherford et 

al., 1997) manages to combine the top-down and bottom-up approach. In this model, a partial 

equilibrium model of the energy sector is developed and linked to a general equilibrium model 

through energy price and quantity variables. 

The bottom-up approach is obviously much more realistic but at the same time it is very demanding 

in terms of both data and behavioral parameters. In addition, it has been shown that the top-down 

approach provides a better assessment of economic agents’ actual responses to changes in prices. As 

this project focuses on the potential impact of biofuel mandates on world prices, exports and imports 

of energy and agricultural commodities and worldwide changes in land use, a top-down approach 

appears to be much more suitable for the purpose of this study. 

78


The GTAP-E model is a typical example of the top-down approach to modeling (Figure 11). The model 

introduces complementarity between intermediate consumption and a composite of Value-Added 

and Energy. It is worth noting that intermediate consumption does not include energy inputs 

(gas/oil/coal/electricity/petroleum products), although it includes energy feedstock. 

The details of the Value-Added and Energy composite are represented in Figure 12Figure 11. This 

modeling approach has four main advantages. Firstly inside the energy composite, the demands for 

each source of energy (electricity/coal/gas/oil/petroleum products) can have different degrees of 

substitutability. In particular demand for gas, oil and petroleum products are relatively substitutable 

while demand of each of these three energy sources is only moderately substitutable with coal and 

electricity. Secondly in the standard GTAP model, as well as in the standard MIRAGE model capital is 

as substitutable with energy as skilled labor due to the inclusion of all energy inputs in the 

intermediate consumption branch of the nesting. In the GTAP-E model the inclusion of energy inputs 

in the Value Added branch of the nesting allows for the differentiation of substitutabilities. Thirdly 

this representation can account for the fact that investment in capital may reduce the demand for 

energy and that the intensity of this relation can vary by sector. 

Figure 11. Structure of production in the GTAP-E model 

Fourthly this representation of productive process can take into account both a short-term 

complementarity between capital and energy and a long-term substitutability. Both the GTAP and 

the MIRAGE models are based on the ‘Putty-Clay hypothesis’ which holds that old capital is sector- 

79


specific while new capital is mobile. Thus following an increase in energy price the substitution 

between capital and energy is rather limited, as in the short term most of the capital is sector 

specific. However, in the long run, if the price shock is permanent, the degree of substitution is much 

larger. Thus the GTAP-E model takes into account both the rigidity in energy use in the short term 

and its flexibility in the long term. While the GTAP-E model represents a major progression in terms 

of energy modeling we do think that it is not fully satisfactory in this case, for several reasons. 

Firstly a key issue of the debate around the development of a biofuels sector and its impact on food 

prices and CO2 emissions is what the literature calls the ‘indirect land use effect’. In other words 

because the allocation of land to the production of agricultural feedstock for non-food purpose 

decreases food supply, it exerts pressure on agricultural prices. This has a tendency to encourage an 

increase in land supply, either from forest or livestock utilization and this change in itself contributes 

to increased CO2 emission. One decisive element in this mechanism is how increased agricultural 

prices translate into increases in land supply. In fact, faced with higher demand farmers can either 

chose a more extensive production process (increased land supply under a constant yield) or a more 

intensive production process (increased yield under a constant land supply). The modeling of 

agricultural processes has to take this mechanism into account. This is the reason why we adopt a 

new nesting, as illustrated in Figure 13. 

In agricultural sectors, the output is a Leontief combination of a “modified Value Added” and a 

“Modified Intermediate Consumption”. We use the term ‘modified’ as from the Value Added side it 

incorporates all primary factors, plus the energy products, plus other products like fertilizers and 

animal feedstock. From the intermediate consumption side it does not incorporate all commodities 

used as intermediate consumption in the production process. This “Modified Value Added” is a 

combination of two composites taking into account the traditional MIRAGE assumptions on the 

elasticity of substitution, which is 1.1 in this case. The first one is a composite of land and either 

animal feedstock in livestock sectors or fertilizers in crops sectors. It enables the key issue of choice 

between intensive and extensive production processes to be tackled. The elasticity of substitution for 

this CES function varies between 0.1 and 2 according to the GTAP database, except for Northern 

countries for which the default elasticity is fixed to 0.1. 

80


Figure 12. Structure of the Capital & Energy Composite in the GTAP-E model 

The other composite is a combination of the standard MIRAGE approach and the GTAP-E approach: 

It incorporates a capital-energy composite according to which investment in capital can 

reduce the demand for energy; 

As only new capital is mobile, the degree of substitutability between capital and energy is 

greater in the long term; 

In Figure 13, under the Capital-energy composite we incorporate the nesting illustrated in 

Figure 12 which incorporates different degrees of substitutability between 

coal/oil/gas/electricity/petroleum products. 

Skilled labor and the capital-energy composite are rather complementary while both can be 

substituted for unskilled labor. 

81


Figure 13. Structure of the Production Process in Agricultural Sectors in the Revised MIRAGE Model 

The paper by Burniaux and Truong (2002) was the inspiration for the elasticities of substitution of the 

different CES nesting levels described above. Between energy and electricity, it is set at 1.1, between 

energy and coal it is 0.5, and between fuel oil and gas it is 1.1. Based on estimates from Okagawa and 

Ban (2008) - (EUKLEMS estimates), the elasticity of substitution between capital and energy is 0.2 in 

Industry, 0.3 in services and 0.03 in agriculture. 

Finally it is worth noting that a distinctive feature of this new version of MIRAGE is in the grouping of 

intermediate consumptions into agricultural inputs/ industrial inputs/services inputs. This introduces 

greater substitutability within sectors, for example substitution is higher between industrial inputs 

(substitution elasticity of 0.6), than between industrial and services inputs (substitution elasticity of 

0.1). At the lowest level of demand for each intermediate, firms can compare prices of domestic and 

foreign inputs and as far as foreign inputs are concerned, the prices of inputs coming from different 

regions. In non-agricultural sectors demand for energy exhibits specific features which are 

incorporated as follows: 

In transportation sectors (Road transport and Air and Sea Transport) the demand for fuel 

which is a CES composite of fossil fuel, ethanol and biodiesel, is rigidified. The modified Value 

82


Added is a CES composite with very low substitution elasticity (0.1) between the usual 

composite (unskilled labor and a second composite which is a CES of skilled labor and a 

capital and energy composite) and fuel which is a CES composite with high elasticity of 

substitution (1.5) of ethanol, biodiesel and fossil fuel. 

In sectors which produce petroleum products, intermediate consumption of oil has been 

rigidified. The modified intermediate consumption is a CES composite (with low elasticity, 

0.1) of a composite of agricultural commodities, a composite of industrial products, a 

composite of services and a composite of energy products which is a CES function (with low 

elasticity) of oil, fuel (composite of ethanol, biodiesel, and fossil fuel with high elasticity, 1.5) 

and of petroleum products other than fossil fuel. The share of oil in this last composite is by 

far the biggest one. This implies that when demand for petroleum products increases, 

demand for oil increases by nearly as much. 

In the gas distribution sector the demand for gas has been rigidified. It has been introduced 

at the first level under the “modified intermediate consumption” composite, at the same 

level as agricultural inputs, industrial inputs and services inputs. This CES composite is 

introduced with a very low elasticity of substitution (0.1). 

In all other industrial sectors we keep the production process illustrated in Figure 4, except 

that there is no land composite and that fuel is introduced in the intermediate consumption 

of industrial products. 

83


Annex III. Final Consumer Energy Demand 

Introduction of a new CES level for energy demand 

Because a LES-CES calibration is more efficient for respecting income elasticity values rather than 

price elasticity ones, it appeared relevant to better set the demand function in order to reflect the 

low elasticity of energy demand to prices. That is why we introduced a third level in the demand 

structure by setting an additional LES-CES function at the first level. The overall demand structure, as 

shown in Figure 14, is therefore: 

A first LES-CES for energy treatment: note that in this first stage, income elasticities for this 

function will be assumed to be one, i.e. minimum shares will be set to zero, and the function 

will follow a CES behavior. 

A second LES-CES function for all other goods. This function is calibrated thanks to a specific 

program that has been adjusted in order to take into account the presence of the first LES- 

CES. 

A CES function in order to represent highly substitutable goods. 

Figure 14. Demand Structure Adapted for Final Energy Consumption 

The direct price elasticity of fuel for transportation is calibrated at - 0.45 to reproduce the right 

evolution of the EU fuel demand for transportation. It corresponds to an intermediate value in the 

literature. 

84

85 



Annex IV. Fertilizer Modeling 

A logistic function for modeling fertilizer effect 

Modeling fertilizers is a delicate task since a simple CES assumption cannot be used to represent the 

impact of fertilizers on crop yield. Indeed, increasing fertilizer use could allow an increase in yields in 

the short run. However, some saturation can occur and some countries cannot get higher yield 

through fertilizers because of an already intensive use of them (Kumar and Goh, 2000). 

We choose here to represent yield reaction to fertilizer as a logistic function. The most general 

logistic functional form would be probably the most appropriate to describe how yield reacts 

because it can be very precisely calibrated on biophysical data. The general form of such a function is 

the following: 

where f is the level of fertilizer input per ha, ymin the potential mimimum yield attainable (bottom 

asymptote), ymax is the maximum yield attainable (top asymptote), y0 is the yield where the 

maximum efficiency is reached (inflexion point), and a is a parameter giving the maximum efficiency 

level. 

However, in a CGE framework, this representation is quite complex to implement. Indeed, this 

function is not convex and therefore does not guarantee the uniqueness of a solution. Second, this 

function is delicate to calibrate because it incorporates many coefficients which require biophysical 

information that are not available for every region. As a consequence, we decided to use a simplified 

yield representation of this function. In order to ensure that the convexity is preserved, we assume 

y0 > ymax-ymin\2. A set of available functions are displayed in Figure 15. 

These functions therefore allow for the modeling of different levels of fertilizer saturation, 

and different levels of response to an increase in fertilizer levels. 

86


Figure 15. Possible concave yield functional forms (ymax = 5) 


Detailed parameters of the function are available upon request. 

In a comparison of our logistic approach and a more traditional CES function between land and 

fertilizer wherein the CES elasticity was calibrated to be comparable with the logistic elasticity at the 

initial point, it appears that the differences are generally minimal (less than 4% on overall ILUC). 

87


Annex V. Modeling of Co-Products of Ethanol and Biodiesel 

On the supply side, meals are produced by the vegetal oil sectors and we calibrate quantity and value 

based on a representative crushing equation for each sector. Yields are assumed to be identical 

across countries and do not change overtime. They are oilseed specific. No by-products (glycerol) of 

biodiesel is considered. 

For the ethanol sectors, DDGS are introduced for all sectors except the Sugar cane based industry. 

For the latter, we only assume that bagasse will generate an income of 6% of the production cost but 

the market is not represented explicitly. 

The substitution patterns between the different feeds are different depending on their nutritional 

content. Oil cakes are appreciated for their protein content (Table 14) and used as a food 

complement to ordinary rations of cereals and DDGS, for which the caloric content is more relevant. 

We therefore introduced two substitution degrees, based on different expressions of feed volume: 

Oil cakes: the first level of substitution describes substitution between oil cakes on the basis 

of their protein content. In order to ensure a consistent substitution, the different values of 

cakes where converted into protein volume, using the shares displayed inTable 15Error! 

Reference source not found.. The default value for elasticity of substitution used at this level 

is 5 which implies a very high substitution. 

Table 14 Protein Content of Oil Cakes used for the Modeling 

Protein content 

per ton 

Rapeseed cake 38% 

Soybean cake 45% 

Palm kernel cake 20% 

Sunflower cake 39% 


Feed, grains and DDGS input: the second level of substitution includes the aggregate of oil 

cakes in substitution with other types of feed and grains and with DDGS. At this level, all 

inputs are expressed in their energy content (see Table 15 showing energy content in 

metabolizable energy, taken from Board on Agriculture and Renewable Resources, 

Commission on Natural Resources, National Research Council (1982)). For oil cakes, an 

average energy content is computed from the initial composition of oil cakes for each 

country and livestock sector. 

Table 15 Energy Content of Feed for Livestock - Metabolizable Energy 

88


Feed Livestock Mcal/t Note 

Rice Cattle 2.42 rice bran - ruminant 

Rice OthAnim 2.59 poultry 2.11 - swine 3.07 

Wheat Cattle 3.08 wheat grain - ruminant 

Wheat OthAnim 3.13 poultry 3.02 - swine 3.25 

Maize Cattle 3.03 grain - ruminant 

Maize OthAnim 3.34 poultry 3.38 - swine 3.3 

VegFruits cattle 0.74 potato, tubers, fresh 

VegFruits OthAnim 0.76 poultry 0.71 - swine 0.82 

OthCrop Cattle 2.9 barley, grain - ruminant 

OthCrop OthAnim 2.51 poultry 2.51 - swine 2.91 

Rapeseed Cattle 0.33 fresh, early bloom 

Rapeseed OthAnim 0.29 Derived from meal value 

Soybeans Cattle 0.64 fresh, dough stage 

Soybeans OthAnim 0.54 Derived from meal value 

Sunflower Cattle 1.36 Sunflower, seed meal not hulled - ruminant 

Sunflower OthAnim 1.68 poultry 1.54 swine 1.81 

SoybnCake Cattle 2.94 Soy meal 0.44 - Ruminant 

SoybnCake OthAnim 2.52 Poultry 2.22 - Swine 2.82 

RpSdCake Cattle 2.66 Rapeseed meal prepressed - Ruminant 

RpSdCake OthAnim 2.3 

Extrapolated from Rapeseed summer values Poultry 2 

- Swine 2.61 

PalmKCake Cattle 3.1 239 kcal / MJ; source: FAO* 

PalmKCake OthAnim 2.5 Extrapolated - should not show in the data 

SunflowerCakel Cattle 2.27 Sunflower meal withou hulls, sol ext - Ruminant 

SunflowerCakel OthAnim 2.36 Poultry 2.08 - Swine 2.65 

http://www.fao.org/ag/AGP/agpc/doc/Proceedings/manado/chap25.htm 

89


Annex VI. Modeling Land Use Expansion 

The mechanism of land use expansion in the revised MIRAGE is based on theoretical foundation that 

is supported by the literature on this issue, but at the same time was designed to be simple enough 

for modeling purposes. The representation explained in this Annex has been introduced in some 

previous works (Bouet et al., 2007 and Valin et al., 2008). This note explains the mechanism in play in 

as much detail as possible. 

1 – Modeling land use expansion: a normative approach 

The first important idea is that this representation of land use is based on the principle that an 

increase in the price of land used for economic activity leads to conversion of new land. Since 

MIRAGE is an economic model, agents are assumed to follow an optimization behaviour. Therefore, 

the rationale of agents in the model is completely different from the rationale presented in Fargione 

et al. (2008) where the assessment is conducted by assuming that a producer arbitrarily plants 

his\her crops on a new area of land, the type of which remains to be determined. As in the case of 

most CGE models that rely on neoclassical assumptions, a producer in MIRAGE only reacts to prices 

and no other rationality constraint is taken into account. Land use conversion is consequently driven 

by price changes.It is also important to consider that, from the econometric point of view, the 

relationship between deforestation and cropland expansion is not yet fully understood. These 

phenomena are quite complex, and most of them depend on the combination of various factors 

which includes prices and others. Furthermore, due to the lack of robust estimates, field specialists 

and geographic economists are very reluctant to propose aggregated elasticities of prices variations 

with respect to land expansion variation. Some scientists also stress that deforestation is impossible 

to model (most studies about land use expansion concerns deforestation for understandable 

reasons, but of course, this seems to be applicable to other). Geist and Lambin (2001) provide a very 

good insight on this complex issue. 

With the background given above, , a few assumptions were made for this analysis: 

- Strong evidence relying on geographical analysis (even if it does not guarantee any causality 

linkage), supports the fact that international markets and price incentives affect land use 

decisions (see Morton et al., 2006 for geographical analysis, Ghimire et al., 2001). And it is 

straightforward to infer that there is a positive correlation between land expansion and the 

price level. 

- The elasticities of land expansion are usually lower than the elasticities of land use 

substitution, 

90


- Furthermore, if yield increases are capped and demand is rigid, deforestation will occur to 

furnish the corresponding supply whatever the value of the elasticity. But we do not know at 

what price. 

However, we do not know the magnitude of the elasticities of land expansion. There are no robust 

estimates from the econometric literature because of the complexity of the linkage and the highly 

fragmented data available for land use in deforested regions, the lack of a continuous time series on 

local prices, and more importantly, land rent, when they exist. More importantly, if we assume for 

each region such an elasticity, we do not know the variation of this elasticity across regions and we 

do not know its sensitivity to specific crop prices. For example, how much does deforestation in 

Indonesia react to price of palm oil in comparison to deforestation in the Amazon with respect to 

price of beef or soybeans? 

One can therefore understand the difficulty of the task of estimating indirect land use change of 

biofuels. Linking crop price changes to land use changes is a much more complex exercise than the 

assessment of the contribution of biofuels to the2008 food price crisis (wherein no land expansion is 

considered since it is a short term phenomena). And yet several quantitative analyses of the food 

price crisis produced a wide range of estimates. A practical way to address such an issue is as follows: 

- We implement in the model with the mechanisms we know, i.e. the positive correlation 

between prices and land use expansion; 

- We base our elasticities on working assumptions, respecting the constraints stated above 

(lower than substitution elasticities but high enough to support the fact that cropland and 

other managed land expansion is driven in part by demand for land products) 

- We perform sensitivity analysis around these values. Values close to the substitution value 

will mean that producers are indifferent between expanding their production by replacing 

their production and using new land. A very low elasticity indicates that the producer will not 

expand much (protected areas of natural land). A land expansion elasticity higher than the 

substitution elasticities will mean that there is little competition for managed land because 

producers can expand at little cost in new areas. 

- We choose to adopt a neutral normative assumption concerning elasticities across regions 

and crops, which means that we assume that each producer, whatever his production type 

or his region, reacts the same way to a price change. 

Even if this approach is weak in terms of support of econometric evidence, it corresponds to the 

most heuristic representation that we can incorporate in an economic model to represent this 

complex phenomenon. 

91


2 –Land Use Substitution 

The details of this mechanism has been documented above. What is however important to keep in 

mind is that a distinction is made between two types of land: managed land, which has an economic 

return, and unmanaged land which is represented without any economic value. 

Managed land includes in the default mode (mode P=0, P standing for “Pasture”): 

- Cropland (cultivated land including permanent crops land and set aside land). 

- Pastureland 

- Managed forest 

These different types of land are substitutes for each other. They are represented in the model in the 

form of economic rental values and the representative land owner can choose to allocate the landproductivity 

(homogenous to land rent values at initial year and defined as land surface adjusted by a 

productivity index) between land use with different substitution levels. 

When demand for a crop increases, prices for the crop go up, and more land is allocated to this crop. 

This land is taken from other uses (pasture and managed forest) with respect to the respective prices 

of these two other categories. In the standard specifications, the price of pasture land is directly 

affected by the demand for cattle products (beef meat and dairy). Forest prices are affected by the 

demand for raw wood products. The magnitude of substitution follows the Constant Elasticity of 

Transformation (CET) specification: 

where L 1 and L 2 are hectares-productivity associated with two different land uses and PL 1 and 

PL 2 are their respective prices. A is a calibration constant and σ is the elasticity of 

transformation. 

If the elasticity of transformation is high, the possibility for land replacement within managed land 

will allow for low prices for the increased demand for crops and aggregated cropland price will not 

increase significantly. But if transformation possibilities inside managed land are smaller (for 

instance, simultaneous demand for competing products on the land market; a very homogenous use 

of the managed land; or very small elasticity of transformation), then cropland prices will rise in 

response to the increased demand. Land use expansion will occur in response to the price increase. 

3 - Land Use Extension 

The mechanism for land use expansion in each region and each AEZ can be represented with the 

simple equation below: 

92


Where 

is managed land expansion into unmanaged land in region r and AEZ z: this land is 

allocated to cropland 

is the exogenous land evolution trend in AEZ z and region r based on 

historical data 

is the average price of managed land for region r and AEZ z, 

is the reference price of managed land in the baseline for the region r 

is an elasticity of land expansion 

is the area of land available for rain-fed crops in region r and AEZ z and not already in use 

This relation has the following properties: 

- In the initial year, MANAGED_LANDZ ini = MANAGED_LANDZ Exo t and therefore LANDEXT = 0 

- In dynamic evolution, land expansion corresponds to the exogenous trend based on 

historical trends. 

- Around the initial point, LANDEXTZ is small in the exponent; therefore, land expansion 

elasticity equals 

- When price of cropland increases, LANDEXTZ increases and MANAGED_LAND expands. In 

this framework, only demand of new land for crops is considered. Therefore, it is the price of 

cropland that determines the expansion and the associated natural land uptake is attributed 

to cropland. 

- When LANDEXTZ increases, becomes smaller and the elasticity of land 

expansion is reduced by this factor. This means that price increases need to be more and 

more important to allow expansion, reflecting the fact that land expansion becomes harder 

when as more available land is used up. If this elasticity gets close to zero, land expansion 

becomes indeed impossible. 

Implicitly, this equation defines what other studies have referred to as a “land supply curve”. Land 

supply curves are often calibrated on physical values (such as productivity displayed in Figure 17). 

93


However, this does not really increase their robustness because the most significant indicator is the 

expansion elasticity at the starting point, which depends more on behavioral factors than on 

biophysical factors (even if biophysical factors can explain a part of the behavior). 

In the revised MIRAGE model, the default value for land expansion has been set at the level of 

substitution value between managed forest and cropland-pasture aggregate in the substitution tree 

(between 0.05 and 0.1 varying by region). However, sensitivity analyses are critical on account of the 

uncertainty on this parameter. 

4 – A Database on Land Available at the AEZ Level 

In order to use a proxy for land available for rain fed crops at the AEZ level, we computed our own 

estimates by decomposing IIASA databases following the procedure outlined below: 

1) Each region is associated with a reference macro region which has similar geophysical 

characteristics. It is then assumed that available land distribution ratio across LGP will be 

close. 

2) The land distribution ratio of the LGP are distributed across AEZ (it means it is distributed 

across climatic zones). For this the key of distribution is a geometric mean of cropland and 

total land. 

3) The land distribution ratio obtained are applied to the land available in the country. 

4) The land available obtained is compared to land under cultivation at the AEZ x country level. 

When land available is less than cropland area, three cases are considered: 

a. If the total of land available is less than the total cropland for the aggregate region, 

then cropland is considered fixed and no expansion will be possible in the region. 

b. If the total of land available – cropland is positive and twice larger for the sum of the 

positive terms than the sum of the negative terms, then one redistributes the 

negative terms, i.e. one considers that AEZs where there is less land available than 

cropland are computation biases. The gap is then redistributed across regions where 

land available is higher than cropland. The key used for AEZ distribution is land 

available – cropland. 

c. If the total of land available – cropland is positive but less than twice larger for the 

sum of the positive terms than the sum of the negative terms, then one consider that 

the data available does not allow a correct distribution of available land and no 

redistribution is done. Land expansion is enabled but only for AEZs where land 

available – cropland > 0. 

94


5) Once all land available is distributed across AEZ and larger than cropland, a last step is to 

check that this land_available does not exceed AEZ area of land with soil (i.e. total productive 

land > land available for crop). For AEZs where this condition is not respected, the extra land 

available is distributed among other AEZs using the land distribution ratio as a key of 

distribution. 

Therefore, the database obtained respects the following criteria: 

- All land available in regions summed across AEZ matches national data from IIASA on land 

available for crops; 

- In each AEZ, land available is equal or greater than cropland. If equal, no expansion is 

considered in the AEZ (and no decrease of cropland). 

- In each AEZ, land available is less than the total quantity of productive land. 

- Available land distribution across AEZ follows the distribution of the macro region mapped 

with the region considered. 

Applied to the aggregation of 10 regions, the distribution is displayed in 

Table 16 and 

95


Figure 16. 

Table 16 Share of Land Available for Rainfed Crop Cultivation Computed for the MIRAGE Model (km²) 

RoOECD China RoW IndoMalay USA LAC Brazil CAMCarib EU27 World 

AEZ1 297 1,077 1,374 

AEZ2 45,883 97,840 2,227 145,950 

AEZ3 19,502 998,449 19,659 100,983 1,138,593 

AEZ4 30,267 16 115,237 12,250 68,796 462,927 11,377 700,870 

AEZ5 45,407 284 32,734 158,354 848,911 5,285 1,090,975 

AEZ6 81,214 579,105 1,544,057 67,731 2,272,107 

AEZ7 5,833 298 1,242 7,373 

AEZ8 68,838 16,227 157,856 207,758 73,444 8,574 532,697 

AEZ9 731 709,365 108,956 87,212 33,521 939,785 

AEZ10 46,740 283 24,478 80,139 2,675 39,079 193,394 

AEZ11 80,196 6,450 50,422 65,656 608 36,979 240,311 

AEZ12 42,983 1,827 35,634 150,950 173,797 1,399 13,814 420,404 

AEZ13 28 104 8,605 8,737 

AEZ14 2,471 415,007 91,014 8,379 6,103 522,974 

AEZ15 2,973 1,547,775 42,379 6,221 1,599,348 

AEZ16 3,215 2,145 1,830 738 4,699 12,627 

AEZ17 541 323 864 

AEZ18 1,056 1,056 

TOTAL 470,944 34,132 4,078,761 44,984 536,901 1,305,897 3,133,958 85,792 138,070 9,829,439 

96


Figure 16 Land Available for Rainfed Cultivation in Unmanaged Land Area (in km²) 

AEZ1 AEZ2 AEZ3 AEZ4 AEZ5 AEZ6 AEZ7 AEZ8 AEZ9 AEZ10 AEZ11 AEZ12 AEZ13 AEZ14 AEZ15 AEZ16 AEZ17 AEZ18 

EU27 

CAMCarib 

Brazil 

LAC 

USA 

IndoMalay 

RoW 

China 

RoOECD 

0 500,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 3,500,000 4,000,000 4,500,000 

Source: Computed from IIASA databases to obtain AEZ distribution and using symmetric assumptions on the 

share of available land under managed pasture and forest and the share of land not under management. 

5 - Marginal Productivity of New Land from Expansion 

The variable LANDEXTZ is not a land-productivity as in the CET structure. That is why it is necessary to 

attribute a productivity factor to the new land converted to make it homogenous with the land 

already in use. A first approach was to multiply the area of land by the marginal productivity of land 

with respect to mean land productivity. Figure 17 shows the distribution curve that is used in the 

model in order to compute the marginal yield to apply. An index of average yield for cropland is 

computed by integrating the curve between the origin and the yellow dot and dividing by the x-axis 

value of the yellow dot. The marginal yield for expansion is then obtained by dividing the marginal 

productivity of managed land by the average productivity of cropland (this indicator is referred to as 

“yield elasticity to land expansion” in the GTAP/CARB study). 

However, we have relied on a much simpler approach in the final study. We assume that marginal 

land productivity in all regions is half the existing average productivity and will not change. This ratio 

is increased to 75% for Brazil. It is important to keep in mind that this assumption remains strong and 

recent research seems to show that recent marginal land extension were taking place on land with at 

least average level yields. 

97


Figure 17. Example of productivity distribution profile for the USA. 

Note : Y axis is a relative index of potential productivity for a 0.5 x 0.5 degree grid cell in the IMAGE 

model. X axis represents the productive land (cultivation potential > 0) and is normalized from 0 to 

1. Black dots (thick line) represent the initial data of the distribution, sorted from the highest value 

to the lowest value, on a 0.5 x 0.5 degree grid cell basis. The thin line represents the interpolation 

curve defined as an 11 th degree polynomial function, and interpolation points are represented with 

black cross. The yellow circle represents the marginal position of arable land use expansion, under 

the assumption that the most productive land is used for cropland. The red point represents the 

marginal position of agricultural land expansion (cropland, pasture and managed forest) under the 

assumption that the most productive land is used for this category. When managed land expand, 

we consider that the marginal value to consider is the latter. 

6 - Allocation of Land Expansion Between other Uses in the Model 

Once land expansion is computed in the model, the difficult task of allocating it between the 

different types of unmanaged land remains. In the revised MIRAGE model, because we rely primarily 

on FAO data, only three different types of unmanaged land are distinguished: 

- Primary forests 

- Savannah and Grassland: this category is mixed with Pastureland into the reference 

“Meadows and Pastures” under FAO nomenclature. With the Monfreda-Ramankutty-Foley 

(2007) database that we use to distinguish the AEZ in managed land, we can disentangle 

these categories, assuming that Pastureland is associated with an economic use, whereas 

Grassland and Savannah are not. 

98


- Other land (shrubland, mountains, deserts, urbanized areas). 

We then allocate the expansion following a coefficient for each land use type. This coefficient 

corresponds to the proportion of the land use type which is converted to cropland when 1 ha of 

cropland expansion occurs. 

We use coefficients from the Winrock database (EPA RIA, Feb 2010) for countries for which this data 

is available. These coefficients are estimated by remote sensing analysis and are supposed to 

specifically correspond to the effect of cropland expansion. For Brazil, these coefficients are AEZ 

specific and thus allows us to accurately reproduce the heterogeneity of expansion distribution 

between AEZs. For other regions, we compute the distribution at the AEZ level with the national 

distribution keys and we eventually adjust using cross entropy if some land use types are not 

available in a specific AEZ. Therefore, the national distribution is conserved whatever the specific 

repartition at the AEZ level. 

It should be noted that in some regions managed land expansion can be a managed land retraction. If 

so, we use the same coefficient to allocate the new land between land use, except for primary forest 

that cannot be recovered by afforestation in that case. Primary forest is therefore replaced by 

plantation forest. 

7 - Pasture and Managed Forest Retroaction 

Representation of cropland expansion into other land uses differ a lot across models depending on 

the transformation possibilities between cropland, pasture and forest land. In computable general 

equilibrium models (like GTAP used for CARB), the representation of land rent for cattle and forest is 

such that demand for these new sectors affects land use. But in many partial equilibrium models that 

do not represent demand for these types of good, (for instance the FAPRI model used by EPA for 

countries other than the US 29 , AGLINK or other models without representation of cattle land), this 

feedback effect is not represented. This is an important issue since the effect of the pasture sector on 

land use can be a large source of uncertainty in results, as long as new demand for cattle is 

associated with new demand for land (which seems to be the case in some areas of the Brazil 

deforestation frontier). For example, some income effect in large and poor areas like Africa can have 

a significant land use effect via a drop in demand for meat following an increase in food prices due to 

biofuels. 

29 The FASOM model used in the EPA assessment of biofuel carbon emissions and compute the ILUC effect 

represent US cattle and US forest. It can therefore represent the effect of land requirements of these sectors. 

99


In order to test the influence of the retroaction of these sectors to biofuel policies, we considered 

several variations in the modeling to better control the possible assumptions: 

- The first mode (P=0) is the GTAP assumption, where all pasture land is allocated to the 

production function of cattle. All pasture land is assumed to be used efficiently so that 

increased demand for cattle products will require an expansion of pasture land. This 

assumption is clearly not realistic for some regions, where cattle intensification is possible. 

- One variant (P=1), which is used in our central scenario, relaxes the P=0 assumption by 

allowing for cattle intensification using an intensification index. At the present time, this 

index is computed in a very simple way: it only corresponds to the number of cattle heads 

(expressed by bovine equivalent, using weight of animals as an indicator of their feed intake) 

by hectare (see 

Table 17). This indicator could be refined to take into account the 

heterogeneity of productivity of grassland, which however cannot be done easily with a nonspatially 

explicit model. From this index of cattle density, we impose a level above which no 

intensification is possible. For countries where no intensification is possible, we attribute all 

pasture to the cattle production function. But for countries where cattle density is below the 

cap, we attribute only a share of the total pasture, which corresponds to the area on which 

the cattle would reach the intensification limit value. Because only a share of pastureland is 

related to production, this design lowers the effect of new demand of cattle. 

Table 17 Number of cattle head (bovine eq) per square kilometers for main regions 

Region 

Cattle head eq per km2 

Rest of OECD countries 31 

China 53 

Rest of World 35 

Indonesia & Malaysia 577 

South Asia 790 

USA 44 

Other Latin America countries 60 

Brazil 118 

Central America and Carribeans 109 

EU27 168 

Source: FAOSTAT (2009) 

- A second variant (P=2) is closer to the assumption in some partial equilibrium models. We 

assume that intensification is possible for cattle (and also for forest), and we remove these 

land types from the substitution tree. This means that there is no retroaction from 

100


pastureland or from forest land on cropland in the model. Technically, this is done by 

assuming that these sectors do not remunerate land but instead remunerate a fixed natural 

resource that is not substitutable with land. Doing so, substitution can only occur within 

cropland, between crop types. In this design, “managed land” area is reduced to cropland 

and expansion occurs in more land types than before. It can expand in: 

o 

o 

o 

o 

o 

Pastureland 

Managed forest 

Primary forests 

Savannah and Grassland 

Other land (shrubland, mountains, deserts, urbanized areas). 

The share of pastureland and managed land affected by land use demand from cropland is no 

longer distributed endogenously with respect to demand of cattle and wood but 

exogenously, using fixed coefficients (more likely, Winrock coefficients). 

All these mechanisms allow us to explore the different dimensions of potential impact of biofuel 

policies on land use change. In turn, computing land use change allows us to compute the associated 

GHG emissions. However, the detailed description of all these different linkages is done mainly for 

explanatory purpose because of all uncertainties on the addressed phenomena, as already discussed 

in the introduction of this annex. 

101


Annex VII. Measurement of Marginal Indirect Land Use Change 

The indirect land use change effects from the use of different biofuel feedstock to produce an 

additional 10 6 GJ of biofuels in the EU is computed in terms of CO2 emissions from the equilibrium 

state reached under the mandate in 2020. Marginal ILUC are computed on a selection of different 

scenarios for 8 different biofuel feedstock: 

- Wheat 

- Corn 

- Sugar beet 

- Sugar cane 

- Rapeseed oil 

- Soybean oil 

- Palm oil 

- Sunflower oil 

The computation starts from the equilibrium state reached under the mandate in 2020. A small shock 

of an extra incorporation commitment of 10 6 GJ is applied to the EU mandate of the level selected 

(4.6%, 5.6%, 6.6%, 7.6%, or 8.6%). For this shock, the level of intermediate consumption of all 

feedstock, except the one studied, is fixed for biofuel production in all regions. The extra demand of 

EU for biofuel is consequently met by an extra production of biofuel with this feedstock only. This 

production can be supplied domestically or come from other regions if some production capacities 

exist in these other regions. This mechanism is illustrated in Figure 18. 

In addition, the demand of regions other than EU for biofuel is maintained constant during the shock 

to ensure that at constant production volume a country does not divert its exports and domestic 

oriented production of biofuel, used with other feedstock, to exports to the EU. Similarly, trade in 

biofuel to non-EU markets are considered unchanged during the marginal shock. Consequently, the 

supply of biofuels across the world only varies by the extra use of the selected feedstock and this 

extra production is sent to the EU for incorporation in transportation fuel. This modeling enables the 

computation of the land use change effects related to the marginal shock on feedstock. 

102


Exports 

EU Biofuel demand: x% mandate in 2020 

+ 10 6 GJ extra demand in feedstock 2 

Foreign 

consumption 

Domestic biofuel 

production 

Foreign biofuel 

production 

Feedstock 

1 

Feedstock 

2 

Feedstock 

3 

Feedstock 

1 

Feedstock 

2 

Feedstock 

3 

Domestic land 

Foreign land 

Figure 18 Modeling of a Marginal ILUC Shock 

Land use change emissions, expressed as gCO2/MJ and gCO2/t of biofuel, are computed from the 

land use change in the model using IPCC Tier 1 methodology. Two types of emissions are considered: 

- Emissions from biomass lost by deforestation: when an area of forest is converted into 

cropland or grassland, the carbon content above ground and below ground is considered 

released into the atmosphere. These emissions are accounted for as a stock variation and as 

an annual loss on a period of amortization of twenty years (no discounting coefficient is 

applied). 

- Emissions from release of carbon in mineral soil: cultivation of new land under several 

management practices is considered releasing carbon on an annual basis for a period of 

twenty years. This carbon release is accounted for on an annual basis. 

This modeling enables the comparison of the indirect land use effect with direct effects, which can be 

measured with a detailed description of sector specificities. Land use change effects are also 

computed by the model. The indicators which are computed are: 

1) Feedstock saving per annum - Prod EU (gCO2eq / MJ and kgCO2eq / t) 

Emissions Prod EU (biofuel) = Production variation (biofuel) * EU Emission factor (biofuel) 

2) Feedstock saving per annum - Conso EU (gCO2eq / MJ and kgCO2eq / t) 

These emissions correspond to savings from the extra world production consumed in the EU. 

It is therefore computed as: 

Emissions Conso EU (biofuel) 

103


= EU production for domestic demand (biofuel)* EU emission factor (biofuel) 

+ Imports (biofuel) * Exporter emission factor (biofuel). 

3) Feedstock saving per annum - Conso World (gCO2eq / MJ and kgCO2eq / t) 

This indicator provides the total carbon savings for the feedstock selected at the world level, 

as a consequence of the EU increase in demand. It incorporates the values from 3) but also 

takes into account change in consumption of other countries affected by the EU mandate. It 

is simply computed as: 

Emissions Conso World (biofuel) = Sum Regions [Production region (biofuel) * Region emission 

factor (biofuel)] 

4) Carbon payback time from 2020 (Conso EU) 

Carbon payback time is computed in reference to the second direct emission indicator (2 = 

Conso EU). This period of time is computed as: 

Carbon payback = Land use change initial emissions (1) 

/ Annual emissions savings - Conso EU (3) 

The coefficients of direct GHG emissions reduction used for different biofuels feedstock in different 

regions are given in the next section. 

104


Annex VIII: The Role of Technology Pathway 

This study uses coefficients of direct GHG emissions reduction for different biofuels feedstock in 

different regions, as reported in Table 18. The Set 1 values are employed in the model and the Set 2 

values are considered for sensitivity analysis. 

These values have no direct impact in our modeling exercise since they are only used in an ad-hoc 

manner to compute the net emissions effects. They have no influence in the outcome of the 

simulations. Their choice is highly debatable since they should refer to future technological paths and 

different methods of estimation of the direct saving effects have been discussed in the literature. We 

show in this annex the consequences of alternative values on the net emissions computations. 

Final users of this research report can easily use alternative values for direct savings and combine 

them with our ILUC computations to determine final net values to ensure their compatibility with 

policy targets. An important debate is to determine if we should consider technological pathways 

that do not match the minimum requirements of the EU legislation. The answer is not 

straightforward since in each country we can have a mix of heterogeneous production processes with 

different levels of energy intensity. Even if the EU manages to enforce specific standards for the 

biofuels sold in its market, substitution can occur: “clean” producers will shift their production to the 

EU market, and may collect a price premium, and the other producers will supply other markets. Due 

to this potential substitution effects, the EU demand of high standard biofuels may still lead to the 

expansion of low energy efficient suppliers, leading to contrasted effects on the environment. 

However, to which extent this mechanism will take place is unclear. Our model assumes only one 

average technology in each country. Future research will demonstrate if we see dual markets for 

biofuels emerge (high standard vs low standard) and how the sector reacts to certification processes, 

and for the later, if they are enforceable. 

105


Table 18 Reduction of CO2 associated with different feedstock – Values used in calculations 

Feedstock Set 1 Set 2 Source (Set 1) 

Wheat (EU) -45% -53% EU Dir (2009) 

Wheat (Other) -32% -50% EU Dir (2009) 

Maize (EU) -56% -56% EU Dir (2009) 

Maize (USA)* -46% -69% EPA (2009) 

Maize (Other)** -29% -29% FAO (2008) 

Sugar Beet -61% -61% EU Dir (2009) 

Sugar Cane -71% -71% EU Dir (2009) 

Soya -40% -50% EU Dir (2009) 

Rapeseed -45% -50% EU Dir (2009) 

Palm Oil -36% -62% EU Dir (2009) 

Sunflower -58% -58% EU Dir (2009) 

Sources: European Council, (2009). Directive of the European Parliament and of the Council on the promotion of the use of 

energy from renewable sources, EPA assessment, JEC estimates (substitution method). 

* EPA (2009) Draft Regulatory Impact Analysis: Changes to Renewable Fuel Standard Program 

** FAO (2008),The State of Food and Agriculture 

Sensitivity analysis: Alternative CO2 direct savings figures 

As previously discussed, looking at alternative direct savings coefficients has different 

interpretations. On one hand, we can consider that our capacity to measure efficiently these 

coefficients today is delicate, in particular if we consider if we assume technologies implemented in 

2020. On the other hand, it also represents how the country-level mix of different technologies for a 

biofuel will evolve with time. Since we rely on average coefficient in each country, looking for higher 

saving coefficients in absolute level will represent the increase of the share of energy-efficient 

producers (plants powered by gas or cogeneration) and the decrease of less efficient producers (e.g. 

coal powered plants). Using the set 2 instead of the set 1 does not change the main picture. Direct 

savings are improved slightly but the main difference is between the two trade scenarios. Indeed, 

trade liberalization leads to a decline in EU ethanol production, in particular wheat based ethanol 

and increase the share of sugar cane ethanol in EU consumption. Since the set 2 increases direct 

savings of wheat ethanol, the gap between the two scenarios is slightly reduced. Finally, the set 2 

improves the net emissions of the palm oil and make it the most attractive vegetal oil (under a 

median assumption concerning peat land emissions). 

106


Annex IX: The Role of Land Extension Coefficients 

The choice of extension coefficients plays a critical role in the CO2 emissions related to the ILUC. 

Indeed, they distribute the increase (or decrease) of agricultural land over the different ecosystems. 

With each ecosystem being associated with different CO2 contents, the distribution of the land 

extension across them defines the CO2 emissions related to the ILUC. Put differently, for the same 

amount of “new” land requested by agriculture, the emissions outcome may vary largely just due to 

the value of these coefficients. In this report, we use coefficients computed by Winrock International 

for the US EPA as reported in Table 19. 

Table 19 Land Extension Coefficients 

ForestManaged ForestPrimary Other Pasture Savannah & 


Argentina 16.4% 0.0% 24.7% 35.6% 23.3% 

Brazil 0.5% 16.3% 11.2% 23.5% 48.5% 

CAMCarib 0.0% 30.4% 10.7% 16.1% 42.9% 

Canada 1.4% 7.8% 42.5% 32.2% 16.1% 

China 5.6% 2.2% 27.3% 39.0% 26.0% 

CIS 3.7% 5.6% 33.3% 30.7% 26.7% 

EU27 8.4% 0.4% 23.5% 36.8% 30.9% 

IndoMalay 3.2% 51.7% 7.0% 7.0% 31.0% 

LAC 17.8% 10.8% 14.3% 23.4% 33.8% 

Oceania 9.0% 0.0% 32.6% 36.0% 22.5% 

RoOECD 14.6% 0.0% 18.8% 20.8% 45.8% 

RoW 3.4% 3.7% 36.9% 39.3% 16.7% 

SEasia 1.1% 20.4% 21.5% 23.1% 33.8% 

SouthAfrica 1.1% 5.1% 28.4% 43.2% 22.2% 

SouthAsia 12.7% 0.0% 32.4% 31.0% 23.9% 

SSA 0.1% 13.0% 16.7% 28.6% 41.7% 

USA 5.4% 2.5% 21.1% 47.4% 23.7% 

Source: EPA (2010) based on Winrock International computations 

Note: For Brazil, the model used AEZ specific coefficients. Figures in the table are simple average of 

the AEZ values. 

107


Annex X. Biofuels Policies 

EU Biofuel Policies 

The European Biofuel policy is quite complex because it is driven not only by the Biofuel Directive, 

but also by others directives and regulations related to Energy, Fuels Quality, Agriculture and Trade 

Policies. 

The Biofuel Directive 30 introduces some constraints on the substitution requirements of fossil fuels 

by biofuels. The main goal of this policy is to reduce greenhouse gas emissions particularly in 

transportation and to lessen dependence on fossil fuels by diversifying energy sources, especially 

towards environmentally friendly technologies. For this purpose, this Directive prescribes several 

mandates for biofuel blending with current fuels at different dates. The first objective was, for each 

EU member to have a 2% market share for biofuels in 2005, then 5.75% in 2010. With the recent 

Renewable Energy Directive, a target of (at least) 10% in 2020 was added. 

In order to help EU members with the implementation of the previous Directive, the Energy Tax 

Directive authorises the EU countries to introduce some tax reductions and exemptions for 

biofuels. 31 The application of both directives differs from one EU country to another. Austria, 

Belgium, Germany and Luxembourg have obtained the best results in response to the targets of the 

Biofuels Directive. They have reached a 2.5 to 2.75 % market share for biofuels. Moreover, other 

developing EU members have also attained the 2005-target: Slovenia (2.5% in 2006), Latvia (2.75% in 

2006), Greece (2% in 2005 and 2006) and the Czech Republic (3.7% in 2005 and 1.78% in 2006). 

However, some other EU members have not yet fulfilled their biofuel commitments, despite various 

incentives (e.g. United Kingdom, Malta, Cyprus, etc.). For instance, the United Kingdom, although it 

has not applied any energy tax reduction/exemption, has favouring production subsidies and capital 

grants for biofuel projects. Austria, Germany and France have all taken similar approaches, reducing 

or exempting biofuel production from taxes imposed on mineral oils, depending on the biofuel type 

(e.g. ethanol or biodiesel) and the level of blending (i.e. Austria exempts 100% tax for pure biodiesel 

but only slightly reduces this tax for 5%-ethanol gasoline). 

The Common Agricultural Policy also plays an important role in encouraging biofuels production. 

Since the 2003-CAP Reform, the supply of energy crops has benefited from direct payments and 

decoupled support without any set-aside obligation and without any loss of income support. 

Moreover, these energy crops also benefit from a premium over the price received by producers and 

30 Directive 2003/30/EC of 8 May 2003 concerning biofuel promotion for transport use. 

31 Energy Tax Directive 2003/96/EC of 27 October 2003. 

108


following the Common Market Organisation regulation, sugar beet production for ethanol is 

exempted from production quotas. 

EU trade policies also affect domestic biofuel production as well as reducing export opportunities 

and production incentives for foreign biofuel producers (e.g. USA, Brazil, Indonesia, Malaysia, etc.). 

The Most Favourite Nation (MFN) duty for biodiesel is 6.5%, while for ethanol tariff barriers are 

higher (€19.2 /hectolitre for the HS6 code 220710 and €10.2 / hectolitre for the code 220720). Even 

if tariffs for biodiesel were to be reduced, trade would still have to face more restrictive non-tariff 

barriers (NTBs) in the form of quality and environmental standards, which already mostly affect 

developing country exporters. 

Nevertheless, some European partners already benefit from a duty-free access for biofuels under the 

Everything But Arms Initiative, the Cotonou Agreement, the Euro-Med Agreements and the 

Generalised System of Preferences Plus. Many ethanol exporters, such as Guatemala, South Africa 

and Zimbabwe, use this free access opportunity. However, most ethanol imports come from Brazil 

and Pakistan under the ordinary European GSP without any preference for either since 2006. 

Concerning European biofuel exports, the EU has a preferential access for ethanol in Norway through 

tariff-rate quotas (i.e. 164 thousand hectolitres for the code 220710 and 14.34 thousand hectolitres 

for 220720). 

Trade liberalisation for biofuels is a contentious issue in the multilateral negotiation of the Doha 

Round (being relevant both to discussions on agricultural trade liberalisation and trade and 

environment) as well as in the bi-lateral negotiations between the EU and the Mercosur countries. 

Clearly key countries, products and interests are common to both. 

Brazilian Biofuel Policies 

Ethanol policies have been implemented in Brazil since the mid-70s and today blending obligations 

for ethanol are up to 20-25% for gasoline. More recently, Brazil has introduced biodiesel blending 

targets of 2% in 2008 and 5% in 2013, similar to the EU’s. 

In order to reach these obligations, Brazilian federal and state governments grant tax 

reductions/exemptions. The level of advantage varies on the basis of the size of the agro-producers 

and the level of development of each Brazilian region. 

The Common External Tariff (CET) of Mercosur also protects domestic biofuel production, with 

ethanol duties of 20% and biodiesel 14%. These tariffs could be eliminated or significantly reduced 

under the Doha and/or the EU-Mercosur negotiations. Furthermore, no non-tariff barriers constrain 

Brazilian imports of biofuels (e.g. no TRQ on biofuels in Mercosur). 

109


Another important explanatory factor in the growth of the ethanol sector in Brazil is the role of 

foreign investment. Most recent investments come from Europe and the United States. They not 

only concern distillation plants but also sugar cane production. The competitive prices of raw 

materials and the high level of integration in the process explain the lower costs for ethanol 

production in Brazil and the motivation of the foreign investors. 

US Biofuel Policies 

In the USA, as in Brazil, Biofuels policies date back to the 70s. However they are as complex as those 

of the EU because fiscal incentives and mandates vary from one state to another and differ from 

federal ones. The Energy Tax Act of 1978 introduces tax exemption and subsidies for the blending of 

ethanol in gasoline. In contrast, biodiesel subsides are more recent, they were introduced in 1998 

with the Conservation Reauthorization Act. 

Concerning mandates on biofuels consumption, they were instigated by the Energy Policy Act of 

2005 at the federal level, although obligations for biofuel use existed at the state level (e.g. 

Minnesota introduced a mandate on biofuels before the federal government, which it increased to 

20% in 2013). This 2005 Act sets the objective of the purchasing of 4 billion gallons of biofuels in 2006 

and 7.5 billion gallons in 2012. 

The current biofuels policies in the USA consist of three main tools output-linked measures, support 

for input factors and consumption subsidies. Tariffs and mandates benefit biofuels producers 

through price support. Tariffs on ethanol (24% in equivalent ad valorem) are higher than biodiesel 

(1% in equivalent ad valorem) which limit imports especially from Brazil. Moreover, producers 

benefit from tax credits based on biofuels blend into fuels. The Volumetric Ethanol Excise Tax Credit 

(VEETC) and the Volumetric Biodiesel Excise Tax Credit (VBETC) provide the single largest subsidies to 

biofuels, although there are additional subsidies linked to biofuel outputs. 32 

Investments in biofuels also receive financial support from the government, as a kind of capital 

subsidies. Support is also provided for labor and land used in biofuel production in some states (e.g. 

Washington). Input subsidies are another important element in biofuel support in the USA. US 

ethanol production overwhelmingly uses corn which is one of the most heavily subsidized crops in 

the country. In contrast, soybeans, which are the main feedstock used for biodiesel production in the 

USA, are not very subsidized in the USA, which means that prices are not inflated and production is 

less attractive for farmers. Finally, indirect biofuel consumption is also supported by the federal 

32 E.g. a federal small producer tax credit - equivalent to a 10% tax credit per gallon on the first 15 million 

gallons produced -, blenders’ credits, supplier tax refunds and other subsidies at the state level 

110


government through investment in infrastructure for transport, storage and distribution (Koplow, 

2006; Koplow, 2007). 

Modeling Biofuel Policies 

In order to calibrate the model and to run the different simulation scenarios for the European biofuel 

policies, we need to build a “policy” data set and to identify some technical requirements to be 

incorporated into the model. 

Obligations in Substitution Requirements for Biofuels 

The EU members are required to report to the Commission on their implementation of biofuel 

policies. Considering the development disparities, the implementation of these policies is largely 

developed in larger countries such as France, Germany or Austria and not in small countries such as 

Malta and Cyprus. However, the EU mandate for the share of biofuel in fossil fuel according to their 

energy content is compulsory for all countries. Only 5 of 27 EU members have reached the 5% target 

for 2005. One year later this number had doubled and today it shows a positive trend. 

Using the national reports to the European Commission relating each country’s biofuel policies, we 

have built a new database that contains the real percentage of biofuels in fuels according to their 

energy content (from 2003 to 2006) and the national and European targets for the years up to 2020. 

For a better use of this database, we differentiate between biodiesel and ethanol with details of how 

much these percentages in energy content terms represent in percentage of the final product (by 

volume), in order to have better information for the model calibration. 

Table 20 shows detailed information about the past application of the biofuels mandates for the 

European Union since 2003 and the prospective application and targets up to 2020. 

111


Table 20 Biofuel Use and Mandates in the European countries (% of energy content) 

Countries GTAP 2003 2004 2005 2006 2007 2008 2009 2010 

2020 

Consumption 

weight 

use EU target use (*) EU target use (*) EU mandate 

Austria 0,015 0,06 1,28 2,5 2 2,5 4,3 5,75 5,75 5,75 5,75 10 10 

Belgium 0,028 0 1 2 2 2,75 3,5 4,25 5 5,75 5,75 10 10 

Bulgaria 0,003 0 0 0 2 0 0 0 0 0 5,75 0 10 

Cyprus 0,001 0 0,5 1 2 1 1 2 2 2 5,75 10 10 

Czeck Republic 0,012 1,12 2,41 3,7 2 1,78 1,63 2,45 2,71 3,27 5,75 10 10 

Germany 0,2 1,18 2,54 3,9 2 2 5,75 7,15 7,88 8,6 5,75 10 10 

Danemark 0,009 0,17 0,24 0,27 2 0,39 1,73 3,07 4,41 5,75 5,75 10 10 

Spain 0,079 0,76 1,38 2 2 3,34 4,62 6 7,26 8,66 5,75 10 10 

Estonia 0,001 0 0,1 0,2 2 2 2,13 2,25 2,38 2,5 5,75 10 10 

Finland 0,005 0,1 0,1 0,1 2 1,75 3,37 5,05 6,63 8,35 5,75 10 10 

France 0,176 0,68 1,34 2 2 1,75 3,5 5,75 6,25 7 5,75 10 10 

Great Britain 0,16 0,03 0,3 0,3 2 0,73 1,15 2 2,8 3,5 5,75 10 10 

Greece 0,012 0 0,35 0,7 2 2,5 3 4 5 5,75 5,75 10 10 

Hungary 0,012 0 0,4 0,6 2 1,63 2,66 3,69 4,72 5,75 5,75 10 10 

Ireland 0,006 0 0,03 0,06 2 1,14 1,75 2,24 4,18 6,12 5,75 10 10 

Italy 0,129 0,5 0,75 1 2 2 2 3 4 5 5,75 10 10 

Lituania 0,002 0 1 2 2 2,75 3,5 4,25 5 5,75 5,75 10 10 

Luxembourg 0,002 0 0 0 2 2,75 2,75 2,75 2,75 5,75 5,75 10 10 

Latvia 0,002 0,21 1,11 2 2 2,75 3,5 4,25 5 5,75 5,75 10 10 

Malta 0 0 0,15 0,3 2 1,92 3,5 5,15 6,7 8,38 5,75 10 10 

Netherdlands 0,027 0,03 1,02 2 2 2 2 2,94 3,88 5,75 5,75 10 10 

Poland 0,03 0,49 0,5 0,5 2 1,5 2,3 3,16 4,03 5,75 5,75 10 10 

Portugal 0,022 0 1 2 2 2 3 5,75 5,75 5,75 5,75 10 10 

Romania 0,006 0 0 0 2 0 0 0 0 0 5,75 0 10 

Slovakia 0,006 0,14 1,07 2 2 2,5 3,2 4 4,9 5,75 5,75 10 10 

Slovenia 0,003 0 0 0 2 1,2 2 3 4 5 5,75 10 10 

Sweden 0,019 1,33 2,17 3 2 3,55 4,1 4,65 5,2 5,75 5,75 10 10 

EU27 0,54 1,2 1,81 2 1,8 3,18 4,47 5,21 6,05 5,75 9,59 10 

Source: Source: Cepii's calculations based on European Commission - National Reports 

Notes: (*) calculated based on national targets and mandates 

National incorporation rates will need to be aggregated at the EU27 level in order to be used with 

the model aggregation, first in the baseline up to 2007 and then in scenarios up to 2020. 

In the baseline scenario, we also need to take into account the mandates for biofuel blending in 

other important countries, such as Brazil and the United States. According to the IEA databases and 

ACG (2005), Brazilian bio-ethanol consumption ratio between 2005 and 2010 should increase and, 

according to the forecast, lead to about a 40% increase in production. Today Brazil blends between 

20-25% of bio-ethanol with gasoline. Since 2005, the Brazilian government has been trying to repeat 

their ethanol policy with biodiesel and new mandatory targets for biodiesel blending have been set 

for 2008, increasing up to 2013 (see Table 21). For the United States, some mandatory incorporation 

has also been ruled out. 

112


Table 21. Current official targets on share of biofuel in total road-fuel consumption 

Countries Official Targets Year Products 

India 5% In near future Biofuels 

Japan 500 million litres 2010 

China 15% 2020 total renewable fuels 

Thailand 2% 2010 Biofuels 

Brazil 20-25% 2006 Ethanol 

40% increase in production 2005-2010 Ethanol 

2% 2008 biodiesel 

5% 2013 biodiesel 

Indonesia 2% of total fuels 2010 biodiesel (palm oil) 

5% of total fuels 2025 biodiesel (palm oil) 

Malaysia 5% In near future biodiesel (palm oil) 

USA 2.78% 2006 Ethanol 

Canada 3.5% 2010 Ethanol 

Source: IEA database; ACG (2005); USDA Brazil report (2007); IFQC Biofuels Center (2006). 

The modeling of the mandates requires firstly splitting the petroleum and coal product sector (p_c) 

from the GTAP database into the petroleum and coal sectors (p_c_fuels and p_c_others). This aspect 

was essential to introducing the consumption obligations for fuels in the transport sector. 

Secondly, even if the biofuel demand without mandate is calibrated assuming a CES function, the 

introduction of the mandate implies removing substitution possibilities and using a Leontief 

structure. As a consequence, for our different mandate scenarios, we will impose fixed shares 

between each biofuel and fossil-fuels. Moreover, we interpret the blending requirement for each 

biofuel in a different way, which means that the mandate is global but rather there is a specific 

mandate modeling by biofuel type. 

Tax Incentives for Biofuels 

The implementation of biofuel consumption mandates is coupled with other support measures. 

Each European country can chose their policy tools independently in order to facilitate and 

encourage biofuel consumption. More specifically, each European member state can grant tax 

reductions/exemptions on biofuel production or consumption in order to reach the European 

mandatory consumption target. However, there is no prescription for implementing these tax 

incentives (e.g. type of biofuels, blending level, taxes, investment grants, etc.), and each member 

state can design its own policy in line with its tax system and the national context. This discretionary 

implementation of tax incentives makes it harder to represent the total biofuel support at the 

European level. 

113


National reports of each Member State detail national taxes/subsidies support at the production 

and consumption level and these can be used to build a database for implementing the baseline and 

different scenarios in our model. Once the EU database on different biofuels support measures was 

completed, we calculated an equivalent ad valorem tax/subsidy on consumption, to get an estimate 

of the effect of support measures at the European level. 

Table 22. Diesel and Biodiesel excise taxes in the European Union ($/liter). 

GTAP 

Consumption 

weight 2005 2006 2007 2005 2006 2007 2005 2006 2007 

AUT 0,015 0,371 0,368 0,368 0,397 0,409 0,403 -0,025 -0,041 -0,035 

BEL 0,028 0,407 0,423 0,407 0,000 0,459 0,202 0,407 -0,036 0,204 

BGR 0,003 

CYP 0,001 

CZE 0,012 0,411 0,435 0,435 0,384 0,409 0,411 0,027 0,026 0,025 

DEU 0,200 0,583 0,583 0,583 0,583 0,508 0,000 0,000 0,074 0,583 

DNK 0,009 0,502 0,501 0,501 0,037 0,434 0,440 0,464 0,067 0,061 

ESP 0,079 0,365 0,365 0,374 0,335 0,335 0,335 0,030 0,030 0,040 

EST 0,001 

FIN 0,005 0,396 0,396 0,396 0,000 0,000 0,396 0,396 0,396 0,000 

FRA 0,176 0,517 0,517 0,531 0,409 0,409 0,310 0,108 0,108 0,221 

GBR 0,160 0,304 0,304 0,342 0,000 0,000 0,322 0,304 0,304 0,020 

GRC 0,012 0,859 0,857 0,879 0,397 0,360 0,358 0,462 0,497 0,521 

HUN 0,012 0,409 0,399 0,399 0,422 0,422 0,422 -0,012 -0,022 -0,022 

IRL 0,006 0,456 0,456 0,456 0,459 0,459 0,456 -0,002 -0,002 0,000 

ITA 0,129 0,506 0,512 0,516 0,471 0,512 0,474 0,035 0,000 0,042 

LTU 0,002 

LUX 0,002 0,339 0,345 0,360 0,057 0,062 0,000 0,281 0,283 0,360 

LVA 0,002 

MLT 0,000 

Diesel tax 2004 

dollar/liter 

Biodiesel tax 

exemption 2004 

dollar/liter 

Biodiesel tax 2004 

dollar/liter 

NLD 0,027 0,453 0,453 0,460 0,000 0,384 0,378 0,453 0,068 0,082 

POL 0,030 0,320 0,362 0,363 0,000 0,310 0,322 0,320 0,052 0,041 

PRT 0,022 0,386 0,389 0,451 0,000 0,000 0,000 0,386 0,389 0,451 

ROM 0,006 

SKV 0,006 0,466 0,482 0,482 0,434 0,434 0,476 0,032 0,048 0,006 

SLK 0,003 

SWE 0,019 0,472 0,491 0,498 0,459 0,484 0,484 0,013 0,007 0,015 

EU27 0,439 0,442 0,454 0,311 0,338 0,260 0,128 0,105 0,193 

Source: CEPII's calculations based on European Environment Agency, OECD (for diesel tax) and Biofuels at what cost? EU, 

IISD (for biofuels tax exemptions). 

Table 23. Gasoline and Ethanol excise taxes in the European Union ($/liter). 

114


GTAP 

Consumption 

weight 2005 2006 2007 2005 2006 2007 2005 2006 2007 

AUT 0,015 0,517 0,517 0,517 0,533 0,533 0,552 -0,016 -0,016 -0,035 

BEL 0,028 0,682 0,734 0,734 0,000 0,732 0,438 0,682 0,002 0,296 

BGR 0,003 

CYP 0,001 

CZE 0,012 0,489 0,518 0,518 0,372 0,037 0,037 0,117 0,481 0,481 

DEU 0,200 0,812 0,812 0,812 0,806 0,806 0,123 0,006 0,006 0,690 

DNK 0,009 0,672 0,677 0,677 0,347 0,347 0,000 0,325 0,330 0,677 

ESP 0,079 0,491 0,491 0,491 0,459 0,459 0,461 0,032 0,032 0,030 

EST 0,001 

FIN 0,005 0,729 0,729 0,729 0,000 0,000 0,000 0,729 0,729 0,729 

FRA 0,176 0,730 0,730 0,753 0,471 0,471 0,409 0,259 0,259 0,343 

GBR 0,160 0,367 0,367 0,410 0,000 0,000 0,000 0,367 0,367 0,410 

GRC 0,012 0,859 0,857 0,879 0,347 0,347 0,358 0,512 0,510 0,521 

HUN 0,012 0,498 0,486 0,486 0,508 0,508 0,513 -0,011 -0,022 -0,027 

IRL 0,006 0,549 0,549 0,549 0,546 0,546 0,549 0,004 0,004 0,000 

ITA 0,129 0,686 0,699 0,699 0,000 0,000 0,000 0,686 0,699 0,699 

LTU 0,002 

LUX 0,002 0,548 0,548 0,573 0,097 0,099 0,000 0,451 0,449 0,573 

LVA 0,002 

MLT 0,000 

Gasoline tax 2004 

dollar/liter 

Ethanol tax 

exemption 2004 

dollar/liter 

Ethanol tax 2004 

dollar/liter 

NLD 0,027 0,826 0,828 0,842 0,000 0,620 0,626 0,826 0,208 0,216 

POL 0,030 0,429 0,444 0,525 0,000 0,459 0,484 0,429 -0,015 0,041 

PRT 0,022 0,670 0,692 0,723 0,000 0,000 0,000 0,670 0,692 0,723 

ROM 0,006 

SKV 0,006 0,498 0,516 0,516 0,459 0,459 0,461 0,039 0,057 0,055 

SLK 0,003 

SWE 0,019 0,660 0,668 0,678 0,682 0,682 0,657 -0,022 -0,014 0,021 

EU27 0,605 0,610 0,625 0,325 0,372 0,214 0,280 0,238 0,411 

Source: CEPII's calculations based on European Environment Agency, OECD (for fuel tax) and Biofuels at what cost? EU, IISD 

(for biofuels tax exemptions). 

In the European Union other incentives do exist, but since excise tax exemption represents more 

than 60% of biofuel fiscal policy incentive we will run our scenarios based on this consumption 

tax/subsidy. 

In the case of Brazil, there are many different consumption and production incentives. Production 

incentives for oilseed production include tax reductions and exemptions, especially federal taxes 

whose reduction level depends on the agriculture type and on the production regions (e.g. only 

subsistence agriculture from the North are exempted from federal taxes, while large agricultural 

producers from the South only benefit from a 32% tax reduction). Each biofuel project also benefits 

115


from loan assistance and there are also some tax reductions at the industrial level. Brazilian states 

also apply different tax incentives on consumption (e.g. 12% tax for biofuels and between 12-17% 

for fossil-fuels). There are also price control policies for biofuels as well as other policies to motivate 

the use of flex-fuel vehicles. Since it is important to introduce Brazilian supports into the baseline, it 

will be necessary to have a national measure taking into account these differences across states. 

Agricultural Policy 

Since the 2003 CAP reform, decoupled policies have been applied to EU energy crops without any 

loss of income and without the initial restrictions due to set-aside obligations. Moreover, the 

production of energy crops benefits from a premium of €45 / hectare with a maximum of 1.5 million 

hectares. Biofuels production in the European Union is also encouraged by the special provision 

included in the CAP for agricultural inputs. 

Concerning sugar beet for ethanol production, the CAP exempts this part of the supply from 

production quotas. This last policy is part of the last Common Market Organisation sugar reform. 

Production quota exemptions for sugar and premiums on energy crops have to be taken into account 

in modeling EU biofuel support. The sectoral split between energy crops and food crops could be 

important to implementing these policies. 

Focus on some Biofuels policies considered in the Baseline scenario 

For the baseline scenario we introduce the current biofuel policies in the EU27, the USA and Brazil 

into the model. These countries mandate a target blend ratio for the percentage of biofuels, which 

should be incorporated into fossil fuels. In order to reach their objectives these countries 

simultaneously implement various fiscal aids and grants, which are incorporated into the model. 

In the EU27, policy in this area is decided at Member State level. Biofuel blend targets are therefore 

compulsory for some countries, but not all. Today, only nine of the twenty-seven European countries 

have set a mandatory requirement for biofuel blend ratios. They couple these obligations with fiscal 

incentives, which also vary from one country to another. Most of them involve total or partial 

reductions in excise-tax on biofuel blended transport fuels or tax-free biofuel quotas. Others also 

include output or input subsidies, the latter supported by the Common Agricultural Policy (CAP). 

Finally, there are some countries that provide investment grants to biofuel development projects, 

such as flex-fuels cars or biofuel distribution infrastructure. 

The heterogeneity in the European biofuels’ policy makes it difficult to simulate scenarios at the EU 

level. For that reason we have introduced some assumptions into the simulations. In the case of the 

baseline scenario, we have introduced the EU targets for biofuel use (at least 2% in 2005 and 3.3% in 

116


2008). At the country level, some countries, but by no means all, have reached the 2005 target. We 

construct our baseline scenario for the level of biofuels blending with fossil fuels on the basis of the 

mean (consumption weighted) development in blending shares at the EU27 level. 

We modeled the excise-tax reduction by calculating the mean (consumption weighted) values for 

each year since 2004 at the EU27 level. For instance, in 2004 the average excise-tax credit was $0.578 

per liter of biodiesel and $0.634 per liter of ethanol. In 2007 the tax credit for biodiesel was slightly 

lower ($0.544 per liter) while that for ethanol was slightly higher ($0.649 per liter) (Kutas et. al, 

2007). For model calibration and for the baseline scenario we use the tax excise credit data from the 

existing literature because the values are very incomplete for the moment and in addition they are 

lower than those in other key papers. Although, as indicated above, there are several other more 

marginal policy measures which impact on the biofuel market and which could have been considered 

(energy crop payment, set-aside payment and market price support), we only model the excise-tax 

credit because it represent more than 60% of the total effective support for biofuels provided in the 

EU. The CAP is also modeled, but without taking into account certain detailed policies related to 

biofuels (e.g. the “no production quota” for sugar beet). Other key policies including biofuel trade 

protection are also considered and the mandate mechanism is explicitly modeled. 

In the USA, both a federal mandate and state-level targets or mandates for biofuel blends exist. The 

federal objective is that 15.2 billion liters (equivalent to 2.78% of gasoline consumption) should be 

consumed in 2006 and 28.4 billion liters (equivalent to 5.2%) by 2012. At state level, these objectives 

may vary. For instance, Iowa State has set a target of 10% by 2009 and 25% by 2020. This is one of 

the highest targets in the USA, where targets do not generally exceed 20%. According to 

AgraFNP(2008) the ethanol industry is lobbying for a higher level of blending - up to 12 or 13%. 

However, so far levels have remained lower, so we only introduce a mandate of 10% for biofuel 

blending in the baseline. 

Subsidies are an important policy tool. Since 2004 the federal and state governments replaced fueltax 

exemption for biofuels with volumetric subsidies or/and consumption mandates. At the federal 

level the volumetric excise-tax credit for ethanol is $0.135 per liter and for biodiesel it is $0.264 33 . 

Direct production subsidies are also significant. There is a federal small producer tax credit of $0.026 

per liter and subsidies to support biofuel production of $0.05 per liter provided at the state level. 

Although there are other indirect support measures related to agricultural inputs or capital grants, 

33 Volumetric biodiesel excise-tax credit distinguishes two different products and thus subsidies: biodiesel 

derived from waste oil, which benefits from 0.132 US dollar per liter and biodiesel derived from agricultural 

fats and oils which receives 0.264 US dollars per liter. In our baseline scenario, we assume the second case 

since we do not have detailed information to model second generation biofuels. 

117


we only consider the above policies support in the baseline scenario, since they together represent 

more than 65% of total biofuels support in the USA ( Koplow, 2006; Koplow, 2007). 

The third country we consider in the baseline is Brazil, where we also introduce detailed information 

about mandates and fiscal aids in the model. Historically Brazil has imposed a mandate for ethanol 

consumption, which presently varies between 20 and 25% depending on the ethanol price. The 

government officially launched the Biodiesel Program in 2004 and in 2005 the new law (LEI N°11097) 

authorized the voluntary blending of biodiesel with petrol diesel for the first 3 years, moving towards 

a mandatory target of 2% for biodiesel blending by 2008 and 5% by 2013 (Methanol Institute et. al, 

2006). 

Mandates in Brazil are therefore differentiated by biofuel type although our modeling does not 

include this distinction. In our baseline scenario however, we take the Brazilian ethanol mandate as 

representing the biofuel mandate. This is a realistic simplification given the predominance of ethanol 

(in the matrix, the biodiesel sector is currently almost nonexistent in Brazil). In modeling the fiscal 

support to biofuels, the excise tax reduction is the most significant element. For ethanol the excise 

tax levied is 67% lower than that applied to gasoline. Decomposing the ethanol excise tax credit by 

source we find that, in 2007, the federal element was $0.135 per liter and the Sao Paulo state part 

$0.224 per liter. The excise tax reduction for biodiesel was fairly stable over the 2004-2007 period. 

Initially it was $0.0973 per liter while at the time of writing it has increased slightly to $0.0992 per 

liter. Other tax exemptions linked to the type of feedstock and the feedstock origin also exist, but 

they are minor compared to the excise-tax credit (Jank et.al, 2007; FAO, 2008b). 

118


8 REFERENCES 

ABARE (1996), “The MEGABARE Model: Interim Documentation”, Canberra 

Abbott P., C. Hurt, and W. Tyner (2008). "What's Driving Food Prices? " Farm Foundation Issue Report. 

Abbott, P., C. Hurt, and W. Tyner (2009). "What's Driving Food Prices? March 2009 Update," Farm Foundation 

Issue Report. 

AgraFNP (2008), ‘Biofuels Brazil’, November 18, issue N°41. 

Banse, M, van Meijl, H and G. Woltjer. (2008) "The Impact of First and Second Generation Biofuels on Global 

Agricultural Production, Trade and Land Use", GTAP Conference Paper, June 2008. 

https://www.gtap.agecon.purdue.edu/resources/download/3693.pdf 

Barr, K., B. Babcock, M. Carriquiry, A. Nasser, and L. Harfuch (2010). "Agricultural Land Elasticities in the United 

States and Brazil", Center for Agricultural and Rural Development, Working Paper 10-WP 505, February 2010. 

Bchir, H., Y. Decreux, J-L Guerin, and S. Jean (2002). "MIRAGE, a Computable General Equilibrium Model for 

Trade Policy Analysis," CEPII Working Paper No 2002-17. France. December 2002. Available at : 

http://www.cepii.fr/anglaisgraph/workpap/pdf/2002/wp02-17.pdf 

Birur, Dileep & Hertel, Thomas & Tyner, Wally, 2008. "Impact of Biofuel Production on World Agricultural 

Markets: A Computable General Equilibrium Analysis," GTAP Working Papers 2413, Center for Global Trade 

Analysis, Department of Agricultural Economics, Purdue University. 

Board on Agriculture and Renewable Resources, Commission on Natural Resources, National Research Council 

(1982). United States-Canadian tables of feed composition : nutritional data for United States and Canadian 

feeds, Subcommittee on Feed Composition, Committee on Animal Nutrition, Washington, DC. 

Boeters, S., P. Veenendaal, N. van Leeuwen, and H. Rojas-Romagoza. (2008). "The Potential for Biofuels 

Alongside the EU-ETS," CPB Netherlands Bureau for Economic Analysis. Paper presented at the GTAP 

Conference in Helsinki, June 2008. See: https://www.gtap.agecon.purdue.edu/resources/download/3871.pdf 

Bouet, A., L. Curran, Dimaranan, B., Ramos, M.P., and H. Valin,(2008). “Biofuels: Global Trade and 

Environmental Impact Study”. Report for DG Trade. ATLASS Consortium. 

Bouët, A., Y. Decreux, L. Fontagné, S. Jean, and D. Laborde (2008). 

World. Review of International Economics. 16(5), 850-863. 

Assessing Applied Protection Across the 

Britz, W. and T. Hertel (2009). "Impacts of EU Biofuels Directives on Global Markets and EU Environmental 

Quality: An Integrated PE, Global CGE analysis," Agric. Ecosyst. Environ. (2009), 

doi:10.1016/j.agee.2009.11.003. 

Burniaux, J. M. and T.P. Truong (2002). "GTAP-E: An Energy-Environment Version of the GTAP Model," GTAP 

Technical Paper No 16, Center for Global Trade Analysis, Purdue University, West Lafayette, IN. 

Commission of the European Communities (CEC) (2008). Proposal for a Directive of the European Parliament 

and of the Council on the Promotion of the Use of Energy from Renewable Sources. Commission of the 

European Communities, Brussels. 

Commission of the European Communities (CEC) (2007). The Impact of a Minimum 10% Obligation for Biofuel 

Use in the EU-27 in 2020 on Agricultural Markets. Note to File, DG Agriculture, 30/04/07. 

Commission of the European Communities (CEC) (2006). Renewable Energy Roadmap. COM(2006) 848, 

Commission of the European Communities, Brussels. 

Commission of the European Communities (CEC) (2003). Directive 2003/30/EC of the European Parliament 

and of the Council on the Promotion of the Use of Biofuels or other Renewable Fuels for Transport. Official 

Journal L 123, 17.5.2003, Brussels. 

Couwenberg, J. (2009). "Emission Factors for Managed Peat Soils: An Analysis of IPCC Default Values," 

Wetlands International, Produced for the UNFCCC meetings in Bonn, June 2009. 

119


Croezen, H. and F. Brouwer. (2008). Estimating Indirect Land Use Impacts from By-Products Utilization - A 

Report to the Renewable Fuels Agency. Published on www.renewablefuelsagency.org, CE Delft, Delft, 

Netherlands. 

Crutzen, P. J., A.R. Mosier, K. A. Smith, and W. Winiwarter. (2007). N20 release from Agro-Biofuel Production 

Negates Global Warming Reductions by Replacing Fossil Fuels. Atmospheric Chemistry and Physics 

Discussions, 7, p. 11191-11205. 

Decreux, Y. and H. Valin. (2007). MIRAGE : Updated Version of the Model for Trade Policy Analysis: Focus on 

Agriculture and Dynamics.’ CEPII Working Paper, N°2007-15, October 2007. [Online] Available at : 

http://www.cepii.fr/anglaisgraph/workpap/pdf/2007/wp07-15.pdf 

De Santi, G. (2008). Biofuels in the European Context: Facts and Uncertainties. Joint Research Centre (JRC), 

Petten, the Netherlands. 

Diao, X., E. Diaz-Bonilla, S. Robinson, and D. Orden. (2005). Tell Me Where It Hurts, An’ I’ll Tell You Who To 

Call: Industrialized Countries’ Agricultural Policies And Developing Countries. IFPRI Discussion Paper 84. 

Washington, D.C.: International Food Policy Research Institute. 

Dixon, P.B., S. Osborne & M.T. Rimmer (2007). The Economy-wide Effects in the United States of Replacing 

Crude Petroleum with Biomass. Energy and Environment 18(6): 709-722. 

DOE (2008). DOE Actively Engaged in Investigating the Role of Biofuels in Greenhouse Gas Emissions from 

Indirect Land Use Change. see: http://www1.eere.energy.gov/biomass/pdfs/obp_science_response_web.pdf 

Eickhout, B. (2008). “The Local And Global Consequences of the EU Renewable Directive for Biofuels.”, 

Netherlands Environmental Assessment Agency, at: http://www.rivm.nl/bibliotheek/rapporten/500143001.pdf 

EPA. (2009). “Draft Regulatory Impact Analysis: Changes to Renewable Fuel Standard Program.” 

European Biodiesel Board. Website: http://www.ebb-eu.org/stats.php 

European Council (2009). “Directive 2009/30/Ec of the European Parliament and of The Council.” 23 April 2009. 

[Online] Available from: http://www.r-e-a.net/document-library/thirdparty/rea-and-fqddocuments/FQD_090605_Directive_200930EC_OJ.pdf 

European Environment Agency (2004). Transport Biofuels: Exploring Links with the Energy and Agricultural 

Sectors. EEA briefing 04, European Environment Agency, Copenhagen. 

Fargione, F, J. Hill, D. Tilman, S. Polasky, and P. Hawthorne (2008), “Land Clearing and the Biofuel Carbon Debt. 

Science 319. 

Farrell, A. E., R.J. Plevin, B.T. Turner, A.D. Jones, M. O’Hare M. and D.M. Kammen (2006). “Ethanol Can 

Contribute to Energy and Environmental Goals.” Science, 311, 27/01/06. 

F.O. Licht. Renewable Fuels Association (2005) “Homegrown for the Homeland: Industry Outlook 2005”, 

(Washington, DC: 2005), p. 14. 

F.O. Licht’s (2008). “World Ethanol and Biofuel Report.” Various Issues. *Online+ Available from: 

http://www.agra-net.com/portal2/puboptions.jsp?Option=archive&pubid=ag072 

Food and Agricultural Organization of the United Nations (2008a). Climate Change and Food Security – A 

Framework Document. Food and Agriculture Organization, Rome. 

Food and Agricultural Organization of the United Nations (FAO) Fertilizer Production Data. [Online] Available 

from: http://faostat.fao.org/site/575/default.aspx. 

Food and Agricultural Organization of the United Nations (FAO) Fertilizer Price Data. [Online] Available from: 

http://www.farmdoc.uiuc.edu/manage/newsletters/fefo08_13/fefo08_13.html 

Food and Agricultural Organization of the United Nations (FAO) Production Statistics: [Online] Available from: 

http://faostat.fao.org/site/339/default.aspx. 

Food and Agricultural Organization of the United Nations (FAO) Supply and Utilisation Accounts (SUA) . [Online] 

Available from:: http://faostat.fao.org/site/354/default.aspx. 

Food and Agricultural Organization of the United Nations (FAO) (2008b). “The State Of Food and Agriculture. 

Biofuels: Prospects, Risks and Opportunities.” FAO-UN, Rome, pp.138. 

120


Gaulier, G. et al. (2008). "BACI: A World Database of International Trade Analysis at the Product-level", CEPII 

Working Paper, 2008. 

Geist H. J., and E.F. Lambin (2001). “What Drives Tropical Deforestation? A Meta-Analysis of Proximate and 

Underlying Causes of Deforestation Based on Subnational Case Study Evidence.” LUCC Report Series; 4. 

Gerber, N., M. van Eckert, and T. Breuer (2009). "Biofuels and Food Prices: A Review of Recent and Projected 

Impacts," RIO 9 - World Climate & Energy Event, Rio de Janeiro, Brazil, 17-19 March 2009. 

Ghimire, K. B., L. Solon, and L. Barraclough (2001). “Agricultural Expansion and Tropical Deforestation: Poverty, 

International Trade and Land Us.”. United Nations. Research Institute for Social Development. 

Gnansounou, E. and L. Panichelli. (2008). "Background document," Workshop on Biofuels and Land Use 

Change, São Paulo, Brazil, 20-21 November 2008. 

Gurgel, A., J.M. Reilly and S. Paltsev (2007).”Potential Land Use Implications of a Global Biofuels Industry.” 

Journal of Agricultural & Food Industrial Organization, 5. 

Hertel, T., W. Tyner, and D.K. Birur (2008). “Biofuels For All? Understanding The Global Impacts Of 

Multinational Mandates.” GTAP Conference Paper, Helsinki, June 2008. 

Hertel, T., S. Rose, and R. Tol. (2008). “Land Use in Computable General Equilibrium Models: An Overview.” 

GTAP Technical Paper. 

Horridge, J.M. (2005). SplitCom: “Programs to Disaggregate a GTAP Sector.” Centre of Policy Studies, Monash 

University, Melbourne, Australia. http://www.monash.edu.au/policy/SplitCom.htm 

International Energy Agency (2009). “World Energy Outlook – 2009.” 

International Energy Agency (2004). “Biofuels For Transport – An International Perspective.” 

InternationalEnergy Agency, Paris. 

International Energy Agency (2008). “Energy Technology Perspectives.” 2008. International Energy Agency, see: 

http://www.iea.org/textbase/npsum//ETP2008SUM.pdf 

International Monetary Fund. (April 2009). World Economic Outlook. Washington DC. 

Jank, M., G. Kutas, L. do Amaral and A. Nassar (2007). “EU and US Policies on Biofuels: Potential Impacts on 

Developing Countries.” The German Marshal Fund of the United States, Washington, DC . 

Kampman, B, F. Brouwer and B. Schepers. (2008). “Agricultural Land Availability and Demand In 2020, Report to 

the Renewable Fuels Agency.” published on www.renewablefuelsagency.org, CE Delft, elft;, The Netherlands. 

Koplow, D. (2007). “Biofuels: At What Cost? Government Support for Ethanol And Biodiesel in the United 

States: 2007 Update,” Technical report, International Institute for Sustainable Development. 

Koplow, D. (2006). “Biofuels: At What Cost? Government Support for Ethanol and Biodiesel in the United 

States, Technical Report.” International Institute for Sustainable Development, 45. 

Kretschmer, B. and S. Peterson (2008). "Integrating Bioenergy into Computable General Equilibrium Models -- A 

Survey," Kiel Working Paper No. 1473. 

Kumar, K., and K. M. Goh (2000). "Crop Residues and Management Practices: Effects on Soil Quality, Soil 

Nitrogen, Crop Yield and Nitrogen Recovery." Advances in Agronomy. Ed. Donald Sparks. San Diego, CA: 

Academic P, 2000. 198-279. 

Kutas, G., C. Lindberg and R. Steenblik (2007). “Biofuels: at What Cost? Government Support for Ethanol and 

Biodiesel in the European Union.” Technical report, International Institute for Sustainable Development. 

Ludena, C., T. Hertel, P. Preckel, K. Foster, and A. Nin Pratt (2006). ‘Productivity Growth and Convergence in 

Crop, Ruminant and Non-Ruminant Production: Measurement and Forecasts.’ GTAP Working Paper No. 35. 

[Online] Available from: https://www.gtap.agecon.purdue.edu/resources/download/2900.pdf. 

Metschies, Gerhard, (2005). ‘ International Fuel Prices. ‘4th Edition. Federal Ministry of Economic 

Development, Eschborn, Germany 2005. 

Methanol Institute &International Fuel Quality Center (2006). “A Biodiesel Primer: Market & Public Policy 

Developments, Quality, Standards & Handling.” April, pp.31. 

121


Mitchell, D. (2008). “A Note on Rising Food Prices.” World Bank, Washington DC. 

Monfreda, C., N. Ramankutty and T. Hertel. (2007), 'Global Agricultural Land Use Data for Climate Change 

Analysis', GTAP Technical Paper. 

Mortimer, N. D, Ashley, A, Evans, A, Hunter, A. T, and V.L. Shaw. (2008). “Support for the Review of Indirect 

Effects of Biofuels.” Northe energy, Stocksfield, UK, published on www.renewablefuelsagency.org 

Morton, D. C., R.S. DeFries, Y.E. Shimabukuro, L.O. Anderson, E. Arai, F. del Bon Espirito-Santo, R. Freitas, and J. 

Morisette (2006). “Cropland Expansion Changes Deforestation Dynamics in The Southern Brazilian 

Amazon.”PNAS, Vol. 103 no. 39 September 26, 2006 14637-1464 

Narayanan G., B. and T. L. Walmsley, Eds. (2008). Global Trade, Assistance and Protection: The GTAP 7 

Database. Center for Global Trade Analysis, Purdue University. 

https://www.gtap.agecon.purdue.edu/databases/v7/v7_doco.asp 

OECD/FAO (2008) OECD/FAO Agricultural Outlook 2008-2017, http://www.agri-outlook.org 

OECD (2008). Economic Assessment of Biofuel Support Policies, Directorate for Trade and Agriculture, OECD 

Paris, at: http://www.oecd.org/dataoecd/19/62/41007840.pdf 

OECD (2006). “Agricultural Market Impacts of Future Growth in the Production of Biofuels, Working Party on 

Agricultural Policies and Markets.” OECD Paris. 

Okagawa, A. and K. Ban (2008). "Estimation of Substitution Elasticities for CGE Models," Discussion Papers in 

Economics and Business from Osaka University No 08-16, Graduate School of Economics and Osaka School of 

International Public Policy (OSIPP). 

Peterson, E. (2008). "GTAP 7 Data Base Documentation Chapter 8.A: Food and Agricultural Data Base." Center 

for Global Trade Analysis. [Online]. Available from : 

https://www.gtap.agecon.purdue.edu/resources/download/4178.pdf 

Rajagopal, D. and D. Zilberman. (2007). "Review of Environmental, Economic and Policy Aspects of Biofuels," 

World Bank Policy Research Working Paper No. 4341. 

Ramankutty, N., J.A. Foley and N. Olejniczak (2002). 'People on the Land: Changes in Global Population and 

Croplands during the 20th Century', Ambio 31(3), 251--257. 

Reilly, J. and S. Paltsev (2007). "Biomass Energy and Competition for Land," MIT Joint Program on the Science 

and Policy of Global Change, Report No.145. Available at: 

http://web.mit.edu/globalchange/www/MITJPSPGC_Rpt145.pdf 

Renewable Fuels Agency (2008). “The Gallagher Review of the Indirect Effects of Biofuels Production.” 

Renewable Fuels Agency, London. At renewablefuelsagency.org. 

Rosegrant, M. (2008). “Biofuels and Grain Prices – Impacts and Policy Responses.” Testimony for the U.S. 

Senate Committee on Homeland Security and Governmental Affairs, May 7, 2008. 

Rosegrant, M., C. Ringler, S. Msangi, T.B. Sulser, T.Zhu and S. Cline (2008). “International Model for Policy 

Analysis of Agricultural Commodities andTrade: Model Description.” International Food Policy Research 

Institute. 

Rutherford, Thomas F., W. David Montgomery and Paul M. Bernstein (1997)."CETM: A Dynamic General 

Equilibrium Model of Global Energy Markets, Carbon Dioxide Emissions and International Trade," February 

1997. 

Seale, J., A. Regmi, and J.A. Bernstein. (2003). “International Evidence on Food Consumption Patterns.” 

Technical Bulletin No. TB1904. 

Searchinger, T., R. Heimlich, R.A. Houghton, F. Dong, A. Elobeid, J. Fabiosa, S. Tokgoz, D. Hayes, and T.H. Yu. 

(2008). “Use of US Croplands for Biofuels Increases Greenhouse Gases Through Emissions from Land Use 

Change.” Science Express, February 7th 2008. 

Sheeran, J. (2008). “Testimony to The European Parliament Development Committee on Behalf of the World 

Food Programme.” Brussels, 6/3/08. 

122


Taheripour, F., D.K. Birur, T.W. Hertel, and W.E. Tyner. (2008). “Biofuels and Their By-Products: Global 

Economic and Environmental Implications.” GTAP Conference paper, Helsinki, June 2008. 

Taheripour, F., D.K. Birur, T.W. Hertel, and W.E. Tyner. (2007). “Introducing Liquid Biofuels into the GTAP 

Database.” GTAP Research Memorandum. Center for Global Trade Analysis, Purdue University. 

Tipper, R., C. Hutchison, and M. Brander (2009). A Practical Approach for Policies to Address GHG Emissions 

from Indirect Land Use Chane Associated with Biofuels. Ecometrica Technical Paper TP-080212-A. 

United Nations (2007). “Biofuels Issues in The New Legislation on the Promotionof Renewable Energy.” 

Contribution from the United Nations1 (UNEP, FAO, UNDP, UN-HABITAT, UNIDO and WHO) to the consultation, 

http://ec.europa.eu/energy/res/consultation/doc/2007_06_04_biofuels/non_og/un_en.pdf 

United States Government Accounting Office (2009). Biofuels: Potential Effects and Challenges of Required 

Increases in Production and Use. US GAO Report to Congressional Requesters, August 2009. 

Valin, H., B. Dimaranan, and A. Bouet (2009). “Biofuels in the World Markets: CGE Assessment of 

Environmental Costs Related to Land Use Changes.” GTAP Conference Paper, XIIth Conference on Global 

Economic Analysis. 

Von Braun, J. (2008). “Biofuels, International Food Prices and the Poor, Testimony to the United States Senate 

Committee on Energy And Natural Resources.” June 12, 2008, Washington DC. 

Wetlands International website: http://www.wetlands.org/ 

Witzke, P. et al. (2008). “Modeling of Energy-Crops in Agricultural Sector Models – A Review of Existing 

Methodologies,” JRC Scientific and Technical Reports. 

Worldwatch Institute (2006). ‘Biofuels for Transportation: Global Potential and Implications for Sustainable 

Agriculture and Energy in the 21st Century’. Washington, DC.: Worldwatch Institute. 

.Zah, R., H. Boni, M. Gauch, R. Hischier, M. Lehman, and P. Wager. (2008). “Life Cycle Assessment of Energy 

Products: Environmental Assessment of Biofuels.” Technical report, EMPA – Materials Science & Technology, 

Federal Office for Energy (BFE), Bern. 

123

ANNEX A.3


2 April 2010 

VIA ELECTRONIC MAIL 

Mr. Bertin Martens 


B-1049 Brussels, Belgium 

E: sg-acc-doc@ec.europa.eu 

bertin.martens@ec.europa.eu 

RE: 

Application Requesting Access to Documents Relating to the “Global Trade 

and Environmental Impact of the EU Biofuels Mandate” Report 

Dear Mr. Martens: 

On behalf of Transport & Environment, ClientEarth, European Environmental Bureau, and BirdLife 

International (collectively “Applicants”), we submit this application for access to documents under 

the Public Access Regulation 1 and the Aarhus Regulation. 2 

We request all documents in the possession, custody, or control of the Commission that relate to the 

report “Global Trade and Environmental Impact of the EU Biofuels Mandate." Under the Public 

Access Regulation, “documents” means “any content whatever its medium... concerning a matter 

relating to the policies, activities and decisions falling within the [Commission’s] sphere of 

responsibility.” 3 This includes, but is not limited to, external and internal communications, internal 

files, memorandums, sound recordings, visual recordings, surveys, workshop minutes, notes, draft 

reports, studies, emails, or portions thereof. Therefore, specifically, we request all documents 

mentioning, discussing, analyzing or describing the following: 

drafts of the above-identified report, “Global Trade and Environmental Impact 

of the EU Biofuels Mandate” by the International Food Policy Research Institute, 

which was finalised on 25 March 2010, including those documents analysing the 

7% scenario and its associated impacts on land use change; and 

all communications, including emails from the Directorate-Generals and third 

parties, that resulted in the decision to settle on the currently estimated volume 

of 5.6% of road transport fuels in 2020 to meet the 10% renewable energy 

mandate by 2020. 

This request is intended to secure all documents created after 15 October 2009 for the abovementioned 

report, which is the date that Applicants submitted the previous application for similar 

documents. It is our view that all pre-15 October 2009 documents relating to the report were 

requested in our 15 October 2009 application. Nevertheless, to the extent the Commission has 

determined that certain pre-15 October 2009 documents—as that term is expansively defined—do 

not to fall within the purview of our 15 October 2009 application, such as emails from DGs and 

1 Regulation (EC) No 1049/2001 of the European Parliament and of the Council of 30 May 2001 regarding public access to 

European Parliament, Council and Commission documents (also referred to herein as the “Public Access Regulation”). 

2 Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 September 2006 on the application of the 

provisions of the Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in 

Environmental Matters to Community institutions and bodies (also referred to herein as the “Aarhus Regulation”). 

3 Public Access Regulation, Article 3(1) and Recitals 10-11; see also Aarhus Regulation, Article 2(1)(d) and Recital 8. 

1 | P a g e


industry, internal correspondence, or any other document, we request that those documents be 

made available too. 4 

In the instance that the Commission feels that certain documents are in Applicants’ possession, 

please contact us to discuss and we will outline the documents in our possession or provide a log of 

the documents received to date. Please note that this invitation to confer does not constitute any 

waiver of our right to a timely response within the statutorily prescribed time-limits. 5 

If the Commission is legally required to redact any of the documents identified above, please inform 

us of the required redactions, outline the justifications for such redaction, identify the period of time 

the redaction is justified, and then produce copies of the documents as redacted. 6 If the Commission 

decides to withhold any responsive documents, we request a log that describes such documents and 

the basis for your determination that such documents fall under an exception in the Public Access 

Regulation. 7 We further request that, in responding to this application, the Commission comply with 

all relevant time-limits set forth in the Public Access Regulation and its internal rules. 8 

To the extent possible, we prefer the documents in electronic format. All documents should be sent 

to Nuša Urbančič at 26 Rue d’Edimbourg, 3 rd Floor, Mundo-B, B-1050 Brussels, Belgium. Please 

produce the documents on a rolling basis. At no point should the Commission’s search for—or 

deliberations concerning—certain documents delay the production of others that the Commission 

has already retrieved and elected to produce. 

If you have any questions or concerns, do not hesitate to contact Tim Grabiel of ClientEarth at +33 

(0)6 32 76 77 04 and tgrabiel@clientearth.org or, in the alternative, Nuša Urbančič of Transport & 

Environment at +32 (0)2 893 0846 and nusa.urbancic@transportenvironment.org. Thank you for 

your prompt attention to this matter. 

Sincerely, 

Nuša Urbančič 

Transport & Environment 

Policy Officer 

Tim Grabiel 

ClientEarth 

Staff Lawyer 

Ariel Brunner 

BirdLife International – European Division 

Head of European Union Policy 

Pieter de Pous 

European Environmental Bureau 

Senior Policy Officer 

4 It is our position that the 15 October 2009 application—as drafted and submitted—includes within its scope a request for all 

emails and other documents from the Directorate-Generals related to the above-identified report. Therefore, we make this 

request without any prejudice to our rights and claims in the legal proceedings before the General Court of the European Union. 

5 Public Access Regulation, Articles 6-8. 

6 Public Access Regulation, Article 4(6) and 4(7); see also Aarhus Regulation, Article 6. 

7 See Public Access Regulation, Article 4(6), Article 4(7), and Article 11(1); see also Case T-2/03, Verein für 

Konsumenteninformation v Commission of the European Communities (2005), paragraph 73. 

8 See, e.g., Public Access Regulation, Articles 6-7; Commission Decision of 5 December 2001 amending its rules of procedure 

(notified under document number c(2001) 3714) 2001/937/EC, ECSC, Euratom, Annex, Article 3. 

2 | P a g e

ANNEX A.4

Tim Grabiel 


From: 

Sent: 

To: 

Subject: 

TRADE-ACCES-DOCU M ENTS@ec. europa. eu 

Tuesday, April27,2010 3:50 PM 

tg rabiel@cl ientearth. org 

Request for Access to Documents - Gestdem no. 2010/1595 

Dear Mr Grabiel, 

Thank you for your email of 0210412010 registered on 06/0412010 applying for a copy of documents in 

accordance with Regulation (EC) N" 10491200I regarding public access to European Parliament, Council 

and Commission documents. 

Your application is currently being dealt with. However, in view of the number of documents applied for, 

we have to extend the prescribed period by another of 15 working days before you receive a reply. We 

apologize for this delay. 

Yours sincerely, 

Christine Moumal 


DG Trade 0.1 - Policy Coordination

ANNEX A.5


8 June 2010 

VIA ELECTRONIC MAIL 

The Secretary-General 

c/o Ms. Catherine Day 



E-mail: sg-acc-doc@ec.europa.eu 

RE: 

Confirmatory Application for Reconsideration of Denial of Application 

Requesting Access to Documents Containing Environmental Information 

Dear Secretary-General: 

ClientEarth, Transport & Environment, European Environmental Bureau, and BirdLife International 

(collectively “Applicants”) submit this confirmatory application for reconsideration of the statutory 

denial of our application—submitted on 2 April 2010 and registered on 6 April 2010—requesting 

access to environmental documents. The original application requested documents on the “Global 

Trade and Environmental Impact of the EU Biofuels Mandate” study by the International Food Policy 

Research Institute and related communications. Those documents provide information necessary for 

meaningful public participation that, if not released, will effectively foreclose the public’s ability to 

engage in the decision-making process on this important issue. Through this confirmatory 

application for reconsideration, Applicants respectfully request that the Secretary-General 

reconsider the denial and grant access to the requested documents. 

FACTUAL BACKGROUND 

In April 2009, the European Parliament and Council approved Directive 2009/28/EC on the 

promotion of the use of energy from renewable sources and amending and subsequently repealing 

Directives 2001/77/EC and 2003/30/EC (hereinafter “RED” for Renewable Energy Directive), which is 

designed to promote wind power, solar energy, hydropower, and energy from biomass. 1 RED 

requires Member States to source 20% of their energy needs from renewables by 2020. It also 

outlines a 10% target for renewables in transportation, which is expected to be met through the 

increased use of biofuels. On the same day, the European Parliament and Council approved Directive 

2009/30/EC amending Directive 98/70/EC as regards the specification of petrol, diesel and gas-oil 

and introducing a mechanism to monitor and reduce greenhouse gas emissions (hereinafter “FQD” 

for Fuel Quality Directive), which includes sustainability criteria and targets a 6% reduction in 

lifecycle greenhouse-gas (GHG) emissions from fuels consumed in the EU by 2020. 2 

During the legislative process, the Community legislature recognised that these policies could be 

less-effective than envisioned and, at times, counter-productive, adversely impacting both forests 

and climate as a result of impacts from indirect land-use change (ILUC). For that reason, the 

Community legislature included an identical provision in RED and FQD requiring the Commission to 

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 

(hereinafter “RED” for Renewable Energy Directive). 

2 Directive 2009/30/EC of the European Parliament and of the Council of 23 April 2009 amending Directive 98/70/EC as 

regards the specification of petrol, diesel and gas-oil and introducing a mechanism to monitor and reduce greenhouse gas 

emissions and amending Council Directive 1999/32/EC as regards the specification of fuel used by inland waterway vessels 

and repealing Directive 93/12/EEC (hereinafter “FQD” for Fuel Quality Directive). 

1 | P a g e


report by 31 December 2010 on the impacts of ILUC and, if appropriate, make proposals to 

incorporate those GHG emissions into the regulatory framework of the directives: 

The Commission shall, by 31 December 2010, submit a report to the European 

Parliament and to the Council reviewing the impact of indirect land-use change on 

greenhouse gas emissions and addressing ways to minimise that impact. The report 

shall, if appropriate, be accompanied by a proposal, based on the best available 

scientific evidence, containing a concrete methodology for emissions from carbon 

stock changes caused by indirect land-use changes, ensuring compliance with this 

Directive [...]. 

Such a proposal shall include the necessary safeguards to provide certainty for 

investment undertaken before that methodology is applied. With respect to 

installations that produced biofuels before the end of 2013, the application of the 

measures referred to in the first subparagraph shall not, until 31 December 2017, 

lead to biofuels produced by those installations being deemed to have failed to 

comply with the sustainability requirements of this Directive if they would otherwise 

have done so, provided that those biofuels achieve a greenhouse gas emission 

saving of at least 45 %. This shall apply to the capacities of the installations of 

biofuels at the end of 2012. 

The European Parliament and the Council shall endeavour to decide, by 31 

December 2012, on any such proposals submitted by the Commission. 3 

It is envisioned that this will take the form of amendments to the directives themselves or a separate 

legislative proposal that conforms to the statutory mandate. At the present, the Commission is 

drafting this report and considering the form of accompanying proposal. But a significant amount of 

information and communications have already been produced that must be made available to the 

public. 

On 2 April 2010, Applicants submitted a request to DG Trade (hereinafter referred to as the 

“Commission”) for access to documents under Regulation (EC) No 1049/2001 regarding public access 

to European Parliament, Council, and Commission documents. The request detailed several 

documents for disclosure: 

drafts of the above-identified report, “Global Trade and Environmental Impact of 

the EU Biofuels Mandate” by the International Food Policy Research Institute, which 

was finalised on 25 March 2010, including those documents analysing the 7% 

scenario and its associated impacts on land use change; and 

all communications, including emails from the Directorate-Generals and third 

parties, that resulted in the decision to settle on the currently estimated volume of 

5.6% of road transport fuels in 2020 to meet the 10% renewable energy mandate by 

2020. 

The application was registered on 6 April 2010. Then, on 27 April 2010, the Commission responded, 

granting itself an additional 15 working days to comply: 

3 RED, Article 19(6); FQD, Article 7d(6). 

2 | P a g e


Thank you for your email of 02/04/2010 registered on 06/04/2010 applying for a 

copy of documents in accordance with Regulation (EC) N° 1049/2001 regarding 

public access to European Parliament, Council and Commission documents. 

Your application is currently being dealt with. However, in view of the number of 

documents applied for, we have to extend the prescribed period by another of 15 

working days before you receive a reply. We apologize for this delay. 

The proffered basis for the time extension is “the number of documents applied for.” Since then, 15 

working days have expired and the Commission has not responded. With this confirmatory 

application for reconsideration, Applicants now request that the Secretary-General reverse this 

improper denial and grant access to all requested documents. 

VIOLATIONS OF REGULATIONS PROVIDING PUBLIC ACCESS TO ENVIRONMENTAL INFORMATION 

In denying the request, the Commission violates two bedrock regulations providing access to 

environmental information. The first is Regulation (EC) No 1049/2001 of the European Parliament 

and of the Council of 30 May 2001 regarding public access to European Parliament, Council and 

Commission documents (hereinafter “Public Access Regulation”), which establishes the right of 

public access to environmental documents. It ushered in a new era of accessibility and legitimacy to 

Community institutions, 4 codifying the principles of openness, transparency and democracy to 

promote legitimacy, accountability, and effectiveness in Community decision-making. It also 

reaffirmed the “right” of public access to documents. 5 

The second is Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 

September 2006 on the application of the provisions of the Aarhus Convention on Access to 

Information, Public Participation in Decision-making and Access to Justice in Environmental Matters 

to Community institutions and bodies (hereinafter “Aarhus Regulation”), which gives the public’s 

right to environmental information fuller effect when relating to environmental information in the 

possession of Community institutions. 6 The Aarhus Regulation was adopted five years after the 

Public Access Regulation, reaffirming and strengthening these principles under its first pillar, “access 

to environmental information.” 7 Together, these regulations grant to Applicants the right to the 

documents and environmental information sought. 

Under Article 7(1) of the Public Access Regulation, the Commission must handle the application 

promptly and provide a written reply in the case of denial: 

An application for access to a document shall be handled promptly. An 

acknowledgement of receipt shall be sent to the applicant. Within 15 working days 

from registration of the application, the institution shall either grant access to the 

document requested and provide access in accordance with Article 10 within that 

period or, in a written reply, state the reasons for the total or partial refusal and 


European Parliament, Council and Commission documents (hereinafter “Public Access Regulation”), Recital 3. 

5 Public Access Regulation, Recital 4. 

6 Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 September 2006 on the application of 

the provisions of the Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to 

Justice in Environmental Matters to Community institutions and bodies (hereinafter “Aarhus Regulation”). 

7 Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in 

Environmental Matters to Community institutions and bodies (hereinafter “Aarhus Convention”), Article 1. 

3 | P a g e


inform the applicant of his or her right to make a confirmatory application in 

accordance with paragraph 2 of this Article. 8 

In exceptional cases, for example in the event of an application relating to a very large number of 

documents, “the time-limit provided for... may be extended by 15 working days, provided that the 

applicant is notified in advance and that detailed reasons are given.” 9 The requirement to give 

detailed reasons is intended to force the institution—here, the Commission—to articulate the basis 

for the extension so that the applicant can ensure no abuse or maladministration. A failure to reply 

within the prescribed time-limit “shall entitle the applicant to make a confirmatory application.” 10 

Under Article 7(4), a “failure to reply within the prescribed time-limit shall entitle the applicant to 

make a confirmatory application.” The Commission has failed to reply to 2 April 2010 application. 

Therefore, this confirmatory application is proper. Without any basis upon which to challenge the 

lawfulness of the denial, Applicants are limited to discussing the general obligations under the law, 

the unlawfulness of the time extension, and the public interest in the documents. This is without 

prejudice to our right to bring additional claims of law and fact in legal proceedings, if necessary. 

I. The Commission’s Rejection of Applicants Request Violates the Law 

As the Commission is well aware, the Public Access Regulation establishes the right of public access 

to EU documents. It ushered in a new era of accessibility and legitimacy to Community institutions, 

codifying the principles of openness, transparency and democracy to promote legitimacy, 

accountability, and effectiveness in EU decision-making. It also reaffirmed the right of public access 

to documents. 11 The Aarhus Regulation gives fuller effect to the public’s right to environmental 

information when in the possession of EU institutions. The Aarhus Regulation was adopted five years 

after the Public Access Regulation, reaffirming and strengthening these principles under its first 

pillar, “access to environmental information.” 12 The right to access environmental information also 

serves as an essential condition precedent to give full effect to the Aarhus Regulation’s second pillar, 

“public participation in decision-making.” 13 

The Public Access Regulation’s two-stage administrative procedure is designed to “to ensure the 

widest possible access to documents,” “to establish rules ensuring the easiest possible exercise of 

this right,” and “to promote good administrative practice on access to documents.” 14 The prescribed 

time-limits assist in “giv*ing+ the fullest possible effect to the right of public access to documents.” 15 

In order to ensure the “right of access is fully respected,” the Public Access Regulation also provides 

the possibility of court proceedings or complaints to the Ombudsman. 16 The foundation of the twostage 

administrative procedure is a presumption overwhelmingly in favour of disclosure: “*i+n 

principle, all documents of the institutions should be accessible to the public.” 17 With respect to 

documents containing environmental information, as here, the Aarhus Regulation essentially 

“guarantee*s+ the right of public access to environmental information received or produced by 

Community institutions or bodies and held by them.” 18 

8 Public Access Regulation, Article 7(1). 



11 See Public Access Regulation, Recitals 1-4. 

12 Aarhus Regulation, Recital 5 and Article 1. 

13 Aarhus Regulation, Recital 5 and Article 9. 

14 Public Access Regulation, Article 1(a)-(c). 




18 Aarhus Regulation, Article 1(1)(a). 

4 | P a g e


The first stage of the two-stage administrative procedure begins when an application is “made in 

any written form.” 19 In the instance of a request for access to Commission documents, as here, the 

application may be submitted to any relevant department, Directorate-General, or the Secretary- 

General. 20 The Public Access Regulation requires that the application “be handled promptly.” 21 

Within 15 working days, the institution is required to “either grant access to the document 

requested... or, in a written reply, state the reasons for the total or partial refusal and inform the 

applicant of his or her right to make a confirmatory application.” 22 The reasons stated in the written 

reply will serve as the basis for an applicant later seeking reconsideration. 23 Only in “exceptional 

cases” may the time-limit be extended. 24 A refusal to disclose a document or a failure to reply within 

the prescribed time-limit entitles the applicant to submit a confirmatory application. 25 This is in 

recognition of the time-sensitive nature of most document requests, especially those containing 

environmental information, and represents the balance struck by the Community legislature 

between administrative review and timely disclosure. 

The second stage of the two-stage administrative procedure begins when a confirmatory application 

is submitted requesting reconsideration of the refusal. In the instance of a request for access to 

Commission documents, as here, the confirmatory application should be submitted to the Secretary- 

General. 26 The Public Access Regulation requires that the confirmatory application “be handled 

promptly.” 27 Within 15 working days, the institution is required to “either grant access to the 

document requested... or, in a written reply, state the reasons for the total or partial refusal.” 28 Only 

in “exceptional cases” may the time-limit be extended. 29 A refusal to disclose or failure to reply 

entitles the applicant to certain remedies, “namely instituting court proceedings against the 

institution and/or making a complaint to the Ombudsman.” 30 The institution must state its reasons 

for refusal “in a written reply” during the two-stage administrative procedure, not after. 

In Internationaler Hilfsfonds v. Commission, the European Court of Justice described the aims of the 

two-stage administrative process: 

With regard to Regulation No 1049/2001, it should be pointed out that Articles 7 

and 8 of that regulation, by providing for a two-stage procedure, aim to achieve, 

first, the swift and straightforward processing of applications for access to 

documents of the institutions concerned and, second, as a priority, a friendly 

settlement of disputes which may arise. For cases in which such a dispute cannot be 

resolved by the parties, the abovementioned Article 8(1) provides two remedies, 

namely the institution of court proceedings or the lodging of a complaint with teh 

Ombusdman. 

That procedure, in so far as it provides for the making of the confirmatory 

application enables in particular the institution concerned to re-examine its position 

before taking a definitive refusal decision which could be the subject of an action 


20 Commission Decision 2001/937/EC, Annex, Article 2. 


22 Public Access Regulation, Article 7(1)(emphasis added). 

23 See, e.g., Public Access Regulation, Articles 7(2) and 8(1). 


25 Public Access Regulation, Article 7(2) and (4). 

26 Commission Decision 2001/937/EC, Annex, Article 4. 


28 Public Access Regulation, Article 8(1)(emphasis added). 



5 | P a g e


before the courts of the Union. Such a procedure makes it possible to process initial 

applications more promptly and, consequently, more often than not to meet the 

applicant’s expectations, while also enabling the institution to adopt a detailed 

position before definitively refusing access to the documents sought by the 

applicant, in particular where the applicant reiterates the request for disclosure of 

those documents notwithstanding a reasoned refusal by that institution. 

The Public Access Regulation and jurisprudence identify several important public policies promoted 

through the two-stage administrative process, two of which are particularly relevant here. 

First, it provides the basis upon which applicants may determine whether to challenge the refusal, 

providing them the ability to weigh the presumption in favour of disclosure against the reasons for 

refusal to determine whether their rights have been violated or a claim to exception is vitiated by 

error. In Kingdom of Sweden v. Commission of the European Communities, the Court found that the 

“information will allow the person who has asked for the document to understand the origin and 

grounds of the refusal of his request and the competent court to exercise, if need be, its power of 

review.” 31 In WWF European Policy Programme v. Council of the European Union, the Court found 

that this obligation to state the reasons for denial is “to provide the person concerned with 

sufficient information to make it possible to determine whether the decision is well founded or 

whether it is vitiated by an error which may permit its validity to be contested.” 32 It also makes 

common sense. Stating the reasons in written form provides the applicant with the ability to secure 

counsel to review the legal merits of the refusal and, if necessary, initiate court proceedings or make 

a complaint to the Ombudsman, as circumstances may require. 

Second, it creates an administrative record upon which judicial review is based. In WWF European 

Policy Programme v. Council of the European Union, the Court stated that “settled case-law provides 

that the purpose of the obligation on the institution to state the reasons for its decision to refuse 

access to a document is... to enable the Community judicature to review the lawfulness of the 

decision.” 33 In order to ensure the most efficient use of limited judicial resources, the institution 

must raise the reasons for withholding the document in written form during the course of the twostage 

administrative procedure, or otherwise waive its ability to raise them later. Reasons offered 

orally or after the two-stage administrative procedure are not within the scope of judicial review. 

For these reasons, the institution must give detailed reasons for the refusal. In Kingdom of Sweden 

v. Commission of the European Communities and Others, the Court found that “as is apparent in 

particular from Articles 7 and 8 of the regulation, the institution is itself obliged to give reasons for a 

decision to refuse a request for access to a document.” 34 In Kingdom of Sweden and Maurizio Turco 

v. Council of the European Union, the Court found that “it is incumbent on the institution concerned 

to give a detailed statement of reasons for such a refusal.” 35 It goes without saying that failing to 

respond in writing—much less to give detailed reasons—during the first stage of the administrative 

hinders the development of the issues for public and judicial review since it precludes applicants, as 

here, from knowing the basis for any denial and responding accordingly. 

31 Case C-64/05 P, Kingdom of Sweden v. Commission of the European Communities (2007), paragraph 89. 

32 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case T- 

187/03, Scippacercola v Commission (2005), paragraph 66. 

33 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36. 

34 Case C-64/05 P, Kingdom of Sweden v. Commission of the European Communities and Others (2007), paragraph 89. 

35 Joined cases C-39/05 P and C-52/05 P, Kingdom of Sweden and Maurizio Turco v. Council of the European Union (2008), 

paragraph 69. 

6 | P a g e


The institution must also carry out a concrete, individual assessment of the content of the 

documents referred to in the request. 36 Courts have found that “where an institution receives a 

request for access under [the Public Access Regulation] it is required, in principle, to carry out a 

concrete, individual assessment of the content of the documents referred to in the request.” 37 This is 

made apparent in “that all exceptions mentioned in Article 4(1) to (3) are specified as being 

applicable to ‘a document.’” 38 On this point, in Verein für Konsumenteninformation v. Commission of 

the European Communities, the court rejected as insufficient an assessment of documents by 

reference to categories rather than on the basis of the actual information contained in those 

documents, “since the examination required of an institution must enable it to assess specifically 

whether an exception invoked actually applies to all the information contained in those 

documents.” 39 A concrete, individual assessment is also needed to ensure compliance with other 

provisions of the Public Access Regulation, including whether redaction is appropriate under Article 

4(6) and the period of time protection is justified under Article 4(7). 40 The purpose of this 

assessment must be forwarded to the applicant to serve as the basis for determining the 

applicability of the exception with respect to the document in question. 41 

In order to be withheld, a document falling under the purview of the Public Access Regulation must 

fall under one of the exceptions provided in the regulation. The Aarhus Regulation provides a special 

rule of interpretation when reviewing the grounds for refusal for environmental information. It 

states that “the grounds for refusal... should be interpreted in a restrictive way,” particularly when 

the “information requested relates to emissions in the environment,” such as GHGs: 

“The grounds for refusal as regards access to environmental information should 

be interpreted in a restrictive way, taking into account the public interest served 

by disclosure and whether the information requested relates to emissions in the 

environment.” 42 

The term “environmental information” is expansively defined, “encompass*ing+ information in any 

form on the state of the environment.” 43 It includes “reports on the implementation of 

environmental legislation,” “the state of the elements of the environment... and the interaction 

among these elements,” and “measures (including administrative measures)... and activities 

affecting or likely to affect [the environment] as well as measures or activities designed to protect 

those elements.” 44 

In addition to its restrictive interpretation of exceptions, the Public Access Regulation requires even 

wider access where, as here, the documents relate to the Commission’s delegated legislative 

36 Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraphs 69-74; 

see also Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. Council of 

the European Communities (1999), paragraph 67. 

37 Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraphs 69-74; 

see also Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. Council of 

the European Communities (1999), paragraph 67. 

38 See Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraph 70. 

39 Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraph 73, 

citing Case T-123/99, JT’s Corporation v. Commission of the European Communities (2000), paragraph 46. 

40 See Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraph 73; 

see also Public Access Regulation, Article 4(6), Article 4(7), and Article 11(1). 

41 See, e.g., Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), 

paragraphs 69-74; Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. 

Council of the European Communities (1999), paragraph 67. 

42 See Aarhus Regulation, Recital 15 and Article 6(1). 

43 Aarhus Regulation, Recital 8. 

44 Aarhus Regulation, Article 2(1)(d)(i), (iii), and (iv). 

7 | P a g e


capacity. Under Recital 6, the documents should be made accessible to the greatest possible extent 

in matters related to legislative activities: 

“Wider access should be granted to documents in cases where the institutions 

are acting in their legislative capacity, including under the delegated powers, 

while at the same time preserving the effectiveness of the institution’s 

decision-making process. Such documents should be made directly accessible 

to the greatest possible extent.” 45 

This wider access has been expansively interpreted. For example, in Kingdom of Sweden and 

Maurizio Turco v. Council of the European Union, the Court rejected the Council’s argument that 

disclosure of legal documents advising the Council on legislative matters would undermine the 

Council’s decision-making. 46 Citing Recital 6, the Court found that, on the contrary, openness 

contributed to “strengthening democracy by allowing citizens to scrutinize all the information which 

has formed the basis of a legislative act,” adding further that the “possibility for citizens to find out 

the considerations underpinning legislative action is a precondition for the effective exercise of their 

democratic rights.” 47 That the Court would not protect legal documents containing legal advice 

given to the Council—a category of documents that has traditionally enjoyed far more privilege 

under the law than inter-departmental communications or scientific and technical findings— 

underscores the particularly restrictive application of any exception when serving in a legislative 

capacity. 

The Public Access Regulation also contains an exception to certain exceptions. Assuming a 

document falls under the narrow category of documents within an exception, an “overriding public 

interest in disclosure” will nevertheless compel its release. 48 In short, the burden on an institution an 

exception with respect to environmental information is significant. 

The review of the basis to a claim to exception is limited to the written record generated during the 

course of the two-stage administrative procedure. Any reason for the total or partial refusal offered 

after the prescribed time-limit is not subject to judicial review. A document may only be withheld if 

proper grounds for refusal are established by the Commission in written form. An overriding public 

interest in disclosure is sufficient to defeat an otherwise valid claim to most exceptions. 49 

II. Failure to Provide Detailed Reasons for Requesting Extension 

The Commission failed to provide detailed and cognizable reasons for extending the statutory timelimit 

for responding. Article 7(3) sets out the process for requesting a time extension: 

In exceptional cases, for example in the event of an application relating to a very 

long document of to a very large number of documents, the time-limit provided for 

in paragraph 1 may be extended by 15 working days, provided that the applicant is 

notified in advance and that detailed reasons are given. 

Time extensions are only for exceptional cases. This is not one of them. Further, in recognition of the 

exceptional nature of a time extension and its impact on meaningful public participation, the 



paragraph 46. 


paragraph 46. 


49 Public Access Regulation, Article 4(2)-(3). 

8 | P a g e


Community legislature requires detailed reasons to justify the extension. In the context of denials, 

the courts have interpreted the detailed-reason requirement as necessary to allow applicants to 

determine whether the decision is vitiated by error. These arguments are equally applicable to time 

extensions too. In Kingdom of Sweden v. Commission of the European Communities and Others, the 

European Court of Justice found that “as is apparent in particular from Articles 7 and 8 of the 

regulation, the institution is itself obliged to give reasons for a decision to refuse a request for access 

to a document.” 50 As a result, the court found that “it is incumbent on the institution concerned to 

give a detailed statement of reasons for such a refusal.” 51 In WWF European Policy Programme v. 

Council of the European Union, the court found that this obligation to state the reasons for denial is 

“to provide the person concerned with sufficient information to make it possible to determine 

whether the decision is well founded or whether it is vitiated by an error which may permit its 

validity to be contested.” 52 

Despite the clear language of the Public Access Regulation, and the exceptional nature of a time 

extension, the Commission failed to provide detailed reasons for the extension, simply alleging a 

unquantified number of documents: 

Your application is currently being dealt with. However, in view of the number of 

documents applied for, we have to extend the prescribed period by another of 15 

working days before you receive a reply. We apologize for this delay. 

This request is inadequate on its face. Time extensions are intended to be rare. Notifying time 

extensions as a matter of course is an abusive and thwarts the principles of transparency, openness, 

and democracy that underlie the Public Access Regulation and Aarhus Regulation. It is this type of 

abuse that is the very reason that detailed reasons are required, and those reasons must be 

proportionate to justify the exceptional nature of the extension. 

Here, the Commission flaunts the limitations on time extensions in several ways. First, the 

Commission failed to confer. The Public Access Regulation provides an informal mechanism to 

resolve requests for a very large number of documents. Under Article 6(3) of the Public Access 

Regulation, “in the event of an application relating to... a very large number of documents, the 

institution concerned may confer with the applicant informally, with a view to finding a fair 

solution.” 53 Though the act of conferring is not a mandatory requirement, it is likely a condition 

precedent to taking the exceptional step of requesting a time extension. It is, in other words, an 

indication of good faith and duty incumbent on the Commission to avoid making exceptional time 

extensions. The Commission made no such effort to confer despite Applicants invitation to do so. 

Second, The Commission failed to provide detailed reasons for the time extension. The simple recital 

of “in view of the number of documents applied for” is inadequate to meet the detailed-reason 

requirement. It is dismissive of the “concept of openness” 54 “transparency of the decision-making 

process,” 55 and the “right of public access to documents.” 56 In the response given, the Commission 

neither divulges the number of documents covered in the request nor gives an indication of their 

length. The Commission also fails to describe any reasons for the delay or the circumstances 

surrounding the need for more time, leaving Applicants to divine them. A knock-on impact of the 

50 Case C-64/05 P, Kingdom of Sweden v. Commission of the European Communities and Others (2007). 


52 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case T- 

187/03, Scippacercola v Commission (2005), paragraph 66. 





9 | P a g e


delay is that the Commission precludes, in effect, the democratic right to participate early in the 

environmental decision-making process on biofuels – a right protected under Article 9 of the Aarhus 

Regulation. 57 

In order to avoid any adverse impacts from a time extension, Applicants requested that all 

documents be produced on a rolling basis: 

Please produce the documents on a rolling basis. At no point should the 

Commission’s search for—or deliberations concerning—certain documents delay 

the production of others that the Commission has already retrieved and elected to 

produce. 

The request that documents be disclosed on a rolling basis was made in response to the 

Commission’s history of abusing time extensions without basis or justification – akin to a delay tactic 

that thwarts public participation. To date, the Commission has not produced a single document. 

Third, the Commission has exhibited a pattern and practice of maladministration on documents 

revealing the science underlying its biofuel policies. Similar obstructionism and delay have already 

resulted in one lawsuit for similar documents. 58 In that lawsuit, as here, the Commission failed to 

provide detailed reasons and justifications for its actions. The result is that the public—and, at times, 

the courts when compelled to review—must expend scarce time and resources to respond to the 

Commission’s nonfeasance in the absence of a fully developed record upon which to base a decision. 

Such antics violate the text and the spirit of the law. 

III. There is an Overriding Public Interest in Disclosure of the Request Documents 

There is an overriding public interest in disclosure of the requested documents. As a general matter, 

the public interest in reducing greenhouse-gas emissions to curb climate change is irrefutable. In 

fact, at the time of submission of this confirmatory application, the leaders of the world are urgently 

negotiating a new climate treaty. The public has every right to be fully informed and involved to 

ensure that EU climate policies, such as promoting biofuels, do not overstate greenhouse-gasemission 

reductions or, as many suspect, actually increase greenhouse-gas emissions. 59 The public 

also has an interest in ensuring that the biofuel targets do not result in the destruction of forests and 

loss of biodiversity. 60 The increase in biofuel consumption without adequate accounting for indirect 

land-use change, however, is expected to do just that. 61 Both these interests—the change in the 

Earth’s climate and the conservation of biological diversity—are recognised as “common concern*s+ 

for humankind” in treaties signed and ratified by the European Union. 62 

More specifically, however, provisions in RED and FQD set targets that artificially increase the 

demand for biofuels. That is their purpose under the assumption that it will reduce GHG emissions. 

But several scientific studies published in reputable periodicals have concluded that biofuels may 

actually increase greenhouse-gas emissions, especially when taking into consideration the impacts of 

57 Aarhus Regulation, Article 9. 

58 ClientEarth e.a. v. Commission, Case No t_120/10 (filed on 8 March 2010). 

59 European Commission, Terms of Reference, Administrative Arrangement between JRC and DG ENV on Indirect Land Use 

Change Emissions from Biofuels, Annex No. 1 to the Offer No. H01-IES/MAM/D(08)(3349)29546, p. 1. 

60 RED, Recital 69; FQD, Recital 11. 

61 See, e.g., RED, Recital 85. 

62 United Nations Framework Convention on Climate Change (UNFCCC), Recital 1; Convention on Biological Diversity (CBD), 

Recital 3. 

10 | P a g e


indirect land-use change. 63 If the Commission is serious about promoting only those biofuels that 

reduce GHG emissions, the public has a clear interest that overrides any other in knowing which 

those biofuels are and what are the risks associated with the policy as a whole. The gravity of the 

climate crisis puts a lot of effort on all actors in the society to reduce GHG emissions. It is urgent that 

the accounted GHG reductions—those claimed on paper—correspond to GHG reductions in reality, 

which may not be the case for biofuels. 

The requested study and correspondence also address biofuel usage scenarios that are likely to exist 

within the EU, completing the GHG picture for biofuels. From the beginning, the Commission 

envisioned biofuels being the “primary” beneficiary of the 10% target in transport. 64 Early in the 

legislative process, the Commission minces no words when discussing the objectives of its proposal: 

[I]t is proposed that each Member State shall achieve at least a 10% share of 

renewable energy (primarily biofuels) in the transport sector by 2020. This is done 

for the following reasons: (1) the transport sector is the sector presenting the most 

rapid increase in greenhouse gas emissions of all sectors of the economy; (2) 

biofuels tackle the oil dependence of the transport sector, which is one of the most 

serious problems of insecurity in energy supply that the EU faces; (3) biofuels are 

currently more expensive to produce than other forms of renewable energy, which 

might mean that they would hardly be developed without a specific requirement. 65 

The Commission launched four studies to examine ILUC issues, including the study by the 

International Food Policy Research Institute (IFPRI). 66 The IFPRI study uses a global computable 

general equilibrium (CGE) model to estimate the impact of EU biofuels policies. 67 As with any model, 

the inputs and assumptions influence the conclusions. The IFPRI study suffers from at least two 

limitations that have the practical effect of reducing ILUC impacts. First, the IFPRI study assumes that 

biofuel consumption comprises only a 5.6% share of the mix of biofuels and fossil fuels despite the 

10% target. 68 Furthermore, the sensitivity analysis shows that the higher the percentage of biofuel 

usage is, the greater the adverse impacts we can expect. The authors confirm this, noting that 

“*s+imulations for EU biofuels consumption above 5.6% of road transport fuels show that ILUC 

emissions can rapidly increase and erode the environmental sustainability of biofuels.” 69 The 

requested documents contain simulations and conclusions under a 7% scenario, allowing the public 

to assess more accurately the true impacts of the 10% target. Second, the study assumes a 45%/55% 

ratio between biodiesel and bioethanol, which makes the biofuel policy look better in terms of GHG 

emissions. 70 The authors promote investigating the assumption behind the ratio, noting that it 

“strongly influence the results.” 71 It is important to understand how this ratio was determine and 

what impact it has on results,. Despite these curious figures, the IFPRI study concludes that there “is 

indeed indirect land use change associated with the EU biofuels mandate.” 72 The IFPRI study further 

finds that “*i+t is clear... that increased demand for biofuels will have impact on the demand for land 

63 See, e.g., Searchinger and Fargionne, Science Magazine (2008); The Gallagher Review for the UK Government (2008); The 

German Study by WBGU (2008); UNEP Sensitivity Analysis of GHG balances of Biofuels (2009). 

64 COD/2008/0016. 

65 COD/2008/0016. 

66 International Food Policy Research Institute, Global Trade and Environmental Impact Study of the EU Biofuels Mandate, 

Final Draft Report (March 2010) *hereinafter “IFPRI Study”+. 

67 IFPRI Study, p. 9. 






11 | P a g e


and will result in potentially significant land use changes.” 73 In this context, the IFPRI study comes 

out with feedstock-specific figures on the marginal ILUC impacts. 

The figures are startling. The IFPRI study computed the marginal effect for each feedstock in 2020: 

EU 5.6% Biofuel Mandate 

(existing trade barriers, no peatland effects) 

Marginal ILUC emissions from biofuels in 2020 

Marginal ILUC Emissions in gCO 2 /MJ Per Year 

(emissions annualised over 20 years) 

Sugar beet ethanol 16.07 

Sugar cane ethanol 17.78 

Wheat ethanol 37.26 

Maize ethanol 54.11 

Rapeseed biodiesel 53.01 

Soybean biodiesel 74.51 

Sunflower biodiesel 59.87 

Palm oil biodiesel 46.40 

In order to count toward the 10% target, a biofuel must meet the 35% GHG threshold, which means 

that it must emit less than 54,47 gCO 2eq /MJ on the life cycle basis. The table above demonstrates 

that several biofuels are already close to or have exceeded those emissions on indirect land-use 

change alone – without incorporating any other sources of GHG emissions. 74 In fact, to put this into 

perspective, several biofuels that the European Union might rely on to meet its targets are shown to 

be worse than fossil fuels themselves under a 5.6% scenario, much less the 10% target for 

renewables in transport. 75 A 5.6% scenario seems to be a very unrealistic assumption considering 

that the Commission originally attested and continues to attest that the 10% target is to be primarily 

met with biofuels. 

These impacts begin to outline that the policy promoting biofuels might contradict other EU policies, 

such as the ones that aim at reducing CO2 emissions, halting deforestation and biodiversity loss. It 

also reveals the winners and losers in the biofuel industry in terms of GHG emissions, information 

that should serve to direct the public and private investments of climate-conscious investors. The 

public has an overriding public interest in ensuring that biofuel science is not manipulated to fit 

biofuel policy but that the biofuel policy is based on the best available science. The Public Access 

Regulation affirms this public interest, stating in Article 12(3) that, “*w+here possible, other 

documents, notably documents relating to the development of policy or strategy, should be made 

directly available.” 76 

CONCLUSION 

The Commission is a public institution funded with public moneys making environmental public 

policy. The public has the right to participate. With this confirmatory application for reconsideration, 

Applicants respectfully request that the Secretary-General grant access to the requested documents 

and information therein, providing access in accordance with Article 10. 


74 FQD, Annex IV(C)(19); see also RED, Annex V(C)(19). 

75 FQD, Annex IV(C)(19); see also RED, Annex V(C)(19). 


12 | P a g e


Sincerely, 

Tim Grabiel 


Staff Lawyer 



Policy Officer 

Ariel Brunner 

Birdlife International – European Division 

Head of European Union Policy 

Pieter Depous 


Senior Policy Officer 

For further information please contact: 

Tim Grabiel 


m +33 (0)6 32 76 77 04 t +32 (0)2 893 0846 

tgrabiel@clientearth.org m +32 (0)488 574 418 

nusa.urbancic@transportenvironment.org 

www.clientearth.org 

www.transportenviornment.org 

Avenue de Tervuren 36 

Rue d’Edimbourg, 3rd Floor, Mundo-B 

Brussels 1040 Brussels 1050 

*** *** *** 

ClientEarth is a non-profit organization dedicated to safeguarding the planet—its flora, fauna, ecosystems, 

people—for the benefit current and future generations. Founded in 2007, ClientEarth attorneys are ushering a 

new era of environmental protection in Europe and beyond, pioneering innovative ways to protect the 

environment through the power of law. With offices in London and in Brussels, ClientEarth provides legal and 

technical capacity to the environmental movement. Activities focus on transformational changes to the 

European legal and legislative landscape, including opening up the European courts to citizen suits, advocating 

for effective environmental legislation with binding and enforceable provisions, bringing transparency to 

European decision-making; and empowering non-governmental organizations. 

Transport & Environment is an independent not-for-profit organisation whose mission is to promote transport 

policy that is based on science and the principles of sustainable development to both minimise the use of 

energy and land—and associated impacts on the environment and health—while maximising safety and 

guaranteeing sufficient access for all. The Brussels-based team focuses on the areas where European Union 

policy has the potential to achieve the greatest environmental benefits, including technical standards for 

vehicle fuel efficiency and pollutant emissions, environmental regulation of international transport, such as 

aviation and shipping, European rules on infrastructure pricing, and environmental regulation of energy used 

in transport. Established in 1990, Transport & Environment represents over 50 organisations across Europe, 

mostly environmental groups and sustainable transport campaigners. 

BirdLife International is a global Partnership of conservation organizations that strives to conserve birds, their 

habitats and global biodiversity, working with people towards sustainability in the use of natural resources. 

13 | P a g e


BirdLife partners operate in me than 100 countries and territories worldwide. BirdLife International is 

represented in 43 countries in Europe and is active in all EU Member States. 

The European Environmental Bureau is a federation of more than 140 environmental citizens’ organisations 

based in all EU Member States and most Accession Countries, as well as in a few neighbouring countries. These 

organisations range from local and national, to European and international. The aim of the European 

Environmental Bureau is to protect and improve the environment of Europe and to enable the citizens of 

Europe to play their part in achieving that goal. 

14 | P a g e

ANNEX A.6


Ref. Ares(2010)378723 - 29/06/2010 

EUROPEAN COMMISSION 

SECRETARIAT-GENERAL 

Direction E 

SG-E-3 

Transparency, Relations with Stakeholders and External Organisations 

Brussels, 29/06/2010 

SG.E3/HP/psi - sg.e.3(2010)420798 

MrTimGrabiel 

By email only. tgrabiel(ã),clientearth.org 

Subject: 

Your confirmatory application under Regulation (EC) 1049/2001 for 

access to documents - GESTDEM No 2010/1595 


I refer to your e-mail dated 8 June 2010, by which, pursuant to Regulation 1049/2001 

regarding public access to European Parliament, Council and Commission documents 1 , 

you lodge a confirmatory application concerning access to documents registered under the 

above-mentioned case number. 

Your application is currently being handled. However, due to the complexity of the issue 

and the need to consult all the involved internal services, we have not yet been able to 

carry out a proper analysis of the requested documents in order to take a final decision. 

Consequently, we will not be able to reply to your confirmatory request within the 

prescribed time limit which expires today. Therefore, we have to extend this period by 

another 15 working days in accordance with Article 8(2) of Regulation 1049/2001. The 

new deadline expires on 20/07/2010. I apologise for any inconvenience this delay may 

cause. 


Marc MAES 

Deputy Head of unit 

1 

OJL145, 31.05.2001, p.43. 

Commission européenne, B-1049 Bruxelles / Europese Commissie, B-1049 Brussel - Belgium. Telephone: (32-2) 299 1111. 

littp://ec.europa.eu/dgs/secretariat_geiieral 

E-mail: sg-acc-docfSec.euiOpa.eu

ANNEX A.7




Ref. Ares(2010)440238 - 19/07/2010 

Direction E 

SG-E-3 


Brussels, 

SG.E3/HP/rc - 

Mr Tim Grabiel 

By email only: tgrabiel(a),clientearth.org 

Subject: 




I refer to your e-mail dated 8 June 2010, by which, pursuant to Regulation 1049/2001 

regarding public access to European Parliament, Council and Commission documents 1 , 

you lodge a confirmatory application concerning access to documents registered under the 

above-mentioned case number. 

I also refer to our letter of 29 June 2010 in which the time limit for handling your 

confirmatory request was extended by another fifteen working days. This time limit will 

expire on 20 July 2010. 

Unfortunately, we will not able to provide you with a final reply to your request within this 

extended time limit, since the required analysis of the documents and the consultation 

with the third party concerned, in accordance with Article 4(4) of Regulation 1049/2001, 

as well as the internal consultations take more time than usual. However, we aim to send 

you a reply within the shortest possible time limit. I regret this additional delay and 

apologise for any inconvenience this may cause. 


fi 

/ 

ALVAREZ CUARTERO 

Maria Isabel 

OJL145, 31.05.2001, p.43. 


http://ec.europa.eu/dgs/secretariat_general 

E-mail: SLi-acc-cloctajec.europa.eu

ANNEX A.8

24/02/2010 Transport & Environment Mail - acces… 


Nuša Urbančič 

access to documents request 

Nuša Urbančič 15 October 2009 15:22 

To: Hilkka.SUMMA@ec.europa.eu 

Dear Ms. Hikka SUMMA, 

With this email we would like to submit a request for access to 

documents under Regulation (EC) No 1049/2001. 

Specifically, we would like to request all the documentation 

(including terms of reference, proposals from researchers, 

correspondence from and to the Commission, minutes of working 

meetings, datafiles with inputs and results, draft, interim, and final 

reports etc.) related to the modelling of the impacts of indirect land 

use change caused by increased biofuels production performed for the 

Commission by the IPTS section of JRC, and by other consultants if 

applicable, as of 1 January 2009. 

Thank you in advance and best regards, 

Nusa Urbancic 

Policy officer for Low Carbon Fuels 

Transport & Environment (T&E) 

www.transportenvironment.org 

** NEW ADDRESS AND FIXED LINE FROM JUNE 10th ** 

3rd floor, Mundo-B 

rue d'Edimbourg, 26 

1050 Brussels 

t. +32 (0)2 893 0846 

m. +32 (0)488 574 418 

…google.com/…/transportenvironmen… 1/1

ANNEX A.9

24/02/2010 Transport & Environment Mail - FW: a… 



FW: access to documents request - No GESTDEM 

2009/4261 - URBANCIC 

AGRI-ACCESS-DOCUMENTS@ec.europa.eu 

To: nusa.urbancic@transportenvironment.org 

3 November 2009 

16:28 

Dear Ms. Urbancic, 

Thank you for your e-mail dated 15/10/2009 registered on 15/10/2009 , requesting access to documents 

under Regulation No 1049/2001 regarding public access to European Parliament, Council and Commission 

documents. 

Your application will be dealt with as quickly as possible. However, because of administrative reasons, we 

have to extend the prescribed period by another of 15 working days before you receive a reply. We apologize 

for this delay. 


Vassiliki Anagnostou 

DG Agriculture et Développement Rural K2 

Accés aux documents 

LOI 130 4/070 

B-1049 Bruxelles 

Tel. +32 2 29 59894 

---------------------------------------------------------------------------------------------------- 

With this email we would like to submit a request for access to 

documents under Regulation (EC) No 1049/2001. 

Specifically, we would like to request all the documentation 

(including terms of reference, proposals from researchers, 

correspondence from and to the Commission, minutes of working 

meetings, datafiles with inputs and results, draft, interim, and final 

reports etc.) related to the modelling of the impacts of indirect land 

use change caused by increased biofuels production performed for the 

Commission by the IPTS section of JRC, and by other consultants if 

applicable, as of 1 January 2009. 

Thank you in advance and best regards, 

Nusa Urbancic 




** NEW ADDRESS AND FIXED LINE FROM JUNE 10th ** 




24/02/2010 Transport & Environment Mail - FW: a… 

1050 Brussels 

t. +32 (0)2 893 0846 

m. +32 (0)488 574 418 



ANNEX A.10

24/02/2010 Transport & Environment Mail - Reply … 



Reply to your request GESTDEM N° 4261/2009 

AGRI-ACCESS-DOCUMENTS@ec.europa.eu 


Cc: Maciej.KUCZYNSKI@ec.europa.eu, Vassiliki.Anagnostou@ec.europa.eu 

27 November 2009 

15:22 

Dear Mrs. Urbancic, 

Pleas find attached the reply to your request to access to documents. 

Best regards. 

 

B. Bagueiro 

DG AGRI.K.2 

Accès aux documents 

Loi 130 4/70 

Tel: 32-2-299.27.36 

Beatriz.Bagueiro@ec.europa.eu 

20091127140615516.tif 

2422K 



























ANNEX A.11

24/02/2010 Transport & Environment Mail - confir… 



confirmatory application concering T&E's request for the 

access to documents request to DG AGRI 

Nuša Urbančič 17 December 2009 10:36 

To: sg-acc-doc@ec.europa.eu 

Cc: Tim Grabiel , Ariel Brunner , Pieter Depous 

 

Dear Sir or Madame, 

Please find in the attachment the Confirmatory Application for access 

to documents request that was originally sent by Transport & 

Environment’s to DG AGRI on 15 October 2009. We are sending the same 

application by ordinary mail. 

Best regards, 

Nusa Urbancic 






1050 Brussels 

t. +32 (0)2 893 0846 

m. +32 (0)488 574 418 

2009 12 17 T&E Confirmatory Application.pdf 

229K 



17 December 2009 

VIA ELECTRONIC MAIL AND POSTAL SERVICE 

Ms. Catherine Day 

The Secretary-General 



E-mail: sg-acc-doc@ec.europa.eu 

RE: 

Confirmatory Application for Reconsideration of the DG AGRI’s Denial of 

Transport & Environment’s Application Requesting Access to Documents 

Containing Environmental Information 

Dear Ms. Catherine Day, 

Transport & Environment submits this confirmatory application for reconsideration of the denial of 

15 October 2009 application requesting access to environmental documents, and is joined by 

ClientEarth, European Environmental Bureau, and Birdlife International (collectively “Applicants”). 1 

The original application requested documents related to modelling of impacts from indirect land-use 

change as a result of European Union biofuel targets. Those documents provide information 

necessary for meaningful public participation that, if not released, will effectively foreclose the 

public’s ability to engage in the decision-making process on this important issue. 

On 27 November 2009, the Directorate-General for Agriculture and Rural Development (“DG AGRI”) 

denied our request. The proffered basis for the denial relied on a narrow and restrictive exception in 

Regulation (EC) No 1049/2001 regarding public access to European Parliament, Council, and 

Commission documents. Through this confirmatory application for reconsideration, Applicants 

respectfully request that the Secretary-General reconsider the denial and grant access to the 

requested documents. 

FACTUAL BACKGROUND 

In April 2009, the European Parliament and Council approved Directive 2009/28/EC on the 

promotion of the use of energy from renewable sources and amending and subsequently repealing 

Directives 2001/77/EC and 2003/30/EC (hereinafter “RED” for Renewable Energy Directive), which is 

designed to promote wind power, solar energy, hydropower, and energy from biomass. 2 RED 

requires Member States to source 20% of their energy needs from renewables by 2020. It also 

outlines a 10% target for renewables in transportation, which is expected to be met through the 

increased use of biofuels. On the same day, the European Parliament and Council approved Directive 

2009/30/EC amending Directive 98/70/EC as regards the specification of petrol, diesel and gas-oil 

and introducing a mechanism to monitor and reduce greenhouse gas emissions (hereinafter “FQD” 

for Fuel Quality Directive), which includes sustainability criteria and targets a 6% reduction in 

1 ClientEarth, European Environmental Bureau, and Birdlife International were not parties to the original application but join in 

Transport & Environment’s confirmatory application as a result of a shared interest in gaining access to the requested documents. 

In the event the Commission requires ClientEarth, European Environmental Bureau, or Birdlife International to submit an 

additional application under Article 6 requesting identical documents—an application that has already been substantively 

denied—we respectfully request that this be considered an Article 6 application for those organisations only. In either instance, 

for purposes of Transport & Environment, this is a confirmatory application under Article 8 that requests reconsideration of DG 

AGRI’s denial of its 15 October 2009 application in the form and manner outlined therein. 

2 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 (hereinafter “RED” for 

Renewable Energy Directive). 

1 | P a g e


lifecycle greenhouse-gas emissions from fuels consumed in the EU by 2020. 3 During the legislative 

process, many recognised that these policies could be less-effective than envisioned and, at times, 

counter-productive, adversely impacting both forests and climate, especially from the potential 

impacts arising from indirect land-use change. As a result, though RED and FQD moved forward, 

both included an identical provision requiring the Commission to report by 31 December 2010 on 

the impacts of indirect land-use change and, if appropriate, make proposals to incorporate those 

greenhouse-gas emissions into the regulatory framework of the directives: 




shall, if appropriate, be accompanied, by a proposal, based on the best available 



Directive [...]. 

Such a proposal shall include the necessary safeguards to provide certainty for 

investment undertaken before that methodology is applied. With respect to 

installations that produced biofuels before the end of 2013, the application of the 

measures referred to in the first subparagraph shall not, until 31 December 2017, 

lead to biofuels produced by those installations being deemed to have failed to 

comply with the sustainability requirements of this Directive if they would otherwise 

have done so, provided that those biofuels achieve a greenhouse gas emission 

saving of at least 45 %. This shall apply to the capacities of the installations of 

biofuels at the end of 2012. 

The European Parliament and the Council shall endeavour to decide, by 31 

December 2012, on any such proposals submitted by the Commission. 4 

It is envisioned that this will take the form of amendments to the directives themselves or a separate 

legislative proposal. At the present, the Commission is drafting this report and considering the form 

of accompanying proposal. But a significant amount of information and communications have 

already been produced that must be made available to the public. 

On 15 October 2009, Transport & Environment submitted a request to DG AGRI for access to 

documents under Regulation (EC) No 1049/2001 regarding public access to European Parliament, 

Council, and Commission documents. The request detailed several documents for disclosure: 

Specifically, we would like to request all documentation (including terms of 

reference, proposals from researchers, correspondence from and to the 

Commission, minutes of working meetings, datafiles with inputs and results, draft, 

interim, and final reports etc.) related to the modelling of the impacts of indirect 

land-use change caused by increased biofuels production performed for the 

Commission by the IPTS Section of JRC, and by other consultants if applicable, as of 

1 January 2009. 

3 Directive 2009/30/EC of the European Parliament and of the Council of 23 April 2009 amending Directive 98/70/EC as regards 

the specification of petrol, diesel and gas-oil and introducing a mechanism to monitor and reduce greenhouse gas emissions and 

amending Council Directive 1999/32/EC as regards the specification of fuel used by inland waterway vessels and repealing 

Directive 93/12/EEC (hereinafter “FQD” for Fuel Quality Directive). 


2 | P a g e


This request was initially met with silence. Then, on 3 November 2009, DG AGRI responded, granting 

itself an additional 15 working days to comply. It cited “administrative reasons” for the delay. 

On 27 November 2009, several days after expiration of its own deadline extension, DG AGRI 

responded with an effective denial of the request. Although technically a partial denial, DG AGRI 

substantially denied the application by withholding all consequential information, releasing instead 

just four “terms of reference” for studies on indirect land-use change. For all other categories of 

documents—proposals from researchers, correspondence to and from the Commission, minutes of 

working meetings, data-files with inputs and results, draft reports, interim reports, and final 

reports—DG AGRI denied the request outright, arguing that those documents are covered by an 

exception: 

I regret to inform you that the study requested, the correspondence concerning the 

work process and the minutes of the working meetings are covered by one of the 

exceptions provided for by the policy relating to access to documents and that they 

cannot be made available to you. The exception which applies to the documents 

you requested is the one mentioned in Article 4(3) of Regulation (EC) No 1049/2001. 

According to it, the institutions could refuse access to a document: 

drawn up by an institution for internal use or received by an 

institution, which relates to a matter where the decision has not 

been taken by the institution, if disclosure of the document would 

seriously undermine the institution’s decision-making process, 

unless there is an overriding public interest in disclosure. 

containing opinions for internal use as part of deliberations and 

preliminary consultations within the institution concerned even 

after the decision has been taken if disclosure of the document 

would seriously undermine the institution’s decision-making 

process, unless there is an overriding public interest in disclosure. 

The studies are not yet finalized and it is our opinion that the public disclosure of all 

of the information that you requested (namely, proposals from researchers, 

correspondence, minutes of working meetings, data files and draft interim reports) 

on such a complex and sensitive issue currently under analysis and validation by the 

Commission is not appropriate. 

However, on the question of the modelling of the potential impacts of indirect land 

use change caused by increased biofuel production, the Commission will make 

studies public once the work has been completed. 5 

The practical effect of this denial is to preclude access to any substantive documentation that would 

allow the public to meaningfully engage in the environmental decision-making process. 

With this confirmatory application for reconsideration, Transport & Environment, joined by 

ClientEarth, European Environmental Bureau, and Birdlife International, now request that the 

Secretary-General reverse this improper denial and grant access to all documents requested by 

Applicants. 

5 European Commission, Directorate-General for Agriculture and Rural Development, Communication Denial Request for Access to 

Documents (27 November 2009), p.2. 

3 | P a g e


VIOLATIONS OF REGULATIONS PROVIDING PUBLIC ACCESS TO ENVIRONMENTAL INFORMATION 

In denying the request, DG AGRI violates two bedrock regulations providing access to environmental 

information. The first is Regulation (EC) No 1049/2001 of the European Parliament and of the 

Council of 30 May 2001 regarding public access to European Parliament, Council and Commission 

documents (hereinafter “Public Access Regulation”), which establishes the right of public access to 

environmental documents. It ushered in a new era of accessibility and legitimacy to Community 

institutions, 6 codifying the principles of openness, transparency and democracy to promote 

legitimacy, accountability, and effectiveness in Community decision-making. It also reaffirmed the 

“right” of public access to documents. 7 

The second is Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 

September 2006 on the application of the provisions of the Aarhus Convention on Access to 

Information, Public Participation in Decision-making and Access to Justice in Environmental Matters 

to Community institutions and bodies (hereinafter “Aarhus Regulation”), which gives the public’s 

right to environmental information fuller effect when relating to environmental information in the 

possession of Community institutions. 8 The Aarhus Regulation was adopted five years after the 

Public Access Regulation, reaffirming and strengthening these principles under its first pillar, “access 

to environmental information.” 9 Together, these regulations grant to Applicants the right to the 

documents and environmental information sought. 

DG AGRI claimed the Article 4(3) exception in the Public Access Regulation in refusing the request. 

That exception allows for denial when disclosure would seriously undermine the institution’s 

decision-making process: 

Access to a document, drawn up by an institution for internal use or received by an 

institution, which relates to a matter where the decision has not been taken by the 

institution, shall be refused if disclosure of the document would seriously 

undermine the institution’s decision-making process, unless there is an overriding 

public interest in disclosure. 

Access to a document containing opinions for internal use as part of deliberations 

and preliminary consultations within the institution concerned shall be refused even 

after the decision has been taken if disclosure of the document would seriously 

undermine the institution’s decision-making process, unless there is an overriding 

public interest in disclosure. 

The courts have interpreted this to establish an obligation to disclose the basis for the denial. 10 In 

other words, Article 4(3) contains a “seriously-undermine standard” that must be met. If a 

document either “relates to a matter where a decision has not been taken” or contains “opinions for 

internal use as part of deliberations and preliminary consultations,” the institution must disclose the 

document unless it would seriously undermine its decision-making process. The courts have found 


European Parliament, Council and Commission documents (hereinafter “Public Access Regulation”), Recital 3. 


8 Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 September 2006 on the application of the 

provisions of the Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in 

Environmental Matters to Community institutions and bodies (hereinafter “Aarhus Regulation”). 

9 Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in Environmental 

Matters to Community institutions and bodies (hereinafter “Aarhus Convention”), Article 1. 

10 Case T-264/04 WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case T-187/03 

Scippacercola v Commission (2005), paragraph 66; See also Case C-64/05 P, Kingdom of Sweden v. Commission of the European 

Communities and Others (2007). 

4 | P a g e


that the exception “must be interpreted and applied strictly.” 11 The word “seriously” connotes a 

strong presumption toward disclosure, in line with the Public Access Regulation’s stated purpose in 

the preamble to “give the fullest possible effect to the right of public access to documents,” 12 taken 

to mean that “all documents of the institutions should be accessible to the public.” 13 The recitals 

further clarify that the exception only entitles institutions “to protect their internal consultations 

and deliberations where necessary to safeguard their ability to carry out their tasks.” 14 As a result, 

under the Public Access Regulation, only in very rare instances is denial warranted under this 

exception – where it would seriously undermine the decision-making and withholding information is 

necessary to prevent that. 

In addition, EU law provides a special rule of statutory interpretation applicable to requests for 

environmental information. The Aarhus Regulation states that, when claiming the Article 4(3) 

exception for environmental information, “the grounds for refusal... should be interpreted in a 

restrictive way,” particularly when the “information requested relates to emissions in the 

environment,” such as greenhouse gases: 

The grounds for refusal as regards access to environmental information should be 

interpreted in a restrictive way, taking into account the public interest served by 

disclosure and whether the information requested relates to emissions in the 

environment. 15 

The term “environmental information” is expansively defined to include “reports on the 

implementation of environmental legislation,” 16 “the state of the elements of the environment... and 

the interaction among these elements,” 17 and “measures (including administrative measures)... and 

activities affecting or likely to affect [the environment] as well as measures or activities designed to 

protect those elements.” 18 As a result, not only is Article 4(3) narrow on its face, but subsequent 

legislation underscores that it is to be interpreted in a restrictive way when involving information on 

the environment. The presumption is, undeniably, strongly in favour of disclosure. 

The Public Access Regulation also contains an exception to the Article 4(3) exception, further 

underscoring its narrow and restrictive application. Assuming a document falls under the narrow 

category of documents whose disclosure would seriously undermine the decision-making process 

and when withholding it is necessary to carry out the Commission’s tasks—two conditions that are 

narrow to begin with and then restrictively interpreted thereafter—an “overriding public interest in 

disclosure” will nevertheless compel its release. In short, the burden on an institution claiming the 

Article 4(3) exception with respect to environmental information is significant. 

In its performance and response to the request for access, DG AGRI committed several violations 

that compel reconsideration and, in the final analysis, disclosure. These violations are addressed in 

turn. 

11 Case C-64/05 P Kingdom of Sweden v Commission of the European Communities and Others (2007), paragraph 66; See Joined 

cases C-39/05 P and C-52/05 P, Kingdom of Sweden and Maurizio Turco v Council of the European Union (2008) 



14 Public Access Regulation, Recital 11 (emphasis added). 

15 Aarhus Regulation, Recital 15 and Article 6(1). 

16 Aarhus Regulation, Article 2(1)(d)(iv). 

17 Aarhus Regulation, Article 2(1)(d)(i) 

18 Aarhus Regulation, Article 2(1)(d)(iii). 

5 | P a g e


I. Failure to Provide Detailed Reasons for Withholding the Requested Documents 

DG AGRI failed to provide detailed and cognizable reasons for denying the application. The reasons 

justifying a claim to an exception under Article 4(3) must be explicitly stated. Article 7 sets out the 

process and requirements to deny a request: 

1. An application for access to a document shall be handled promptly. An 

acknowledgement of receipt shall be sent to the applicant. Within 15 working 

days from registration of the application, the institution shall either grant 

access to the document requested and provide access in accordance with 

Article 10 within that period or, in a written reply, state the reasons for the 

total or partial refusal and inform the applicant of his or her right to make a 

confirmatory application in accordance with paragraph 2 of this Article. 

2. In the event of a total or partial refusal, the applicant may, within 15 working 

days of receiving the institution’s reply, make a confirmatory application asking 

the institution to reconsider its position. 

3. In exceptional case, for example in the event of an application relating to a very 

long document of to a very large number of documents, the time-limit 

provided for in paragraph 1 may be extended by 15 working days, provided 

that the applicant is notified in advance and that detailed reasons are given. 

4. Failure by the institution to reply within the prescribed time-limit shall entitle 

the applicant to make a confirmatory application. 19 

The courts have interpreted this to require detailed reasons for the denial. In Kingdom of Sweden v. 

Commission of the European Communities and Others, the European Court of Justice found that “as 

is apparent in particular from Articles 7 and 8 of the regulation, the institution is itself obliged to give 

reasons for a decision to refuse a request for access to a document.” 20 As a result, the court found 

that “it is incumbent on the institution concerned to give a detailed statement of reasons for such a 

refusal.” 21 In WWF European Policy Programme v. Council of the European Union, the court found 

that this obligation to state the reasons for denial is “to provide the person concerned with sufficient 

information to make it possible to determine whether the decision is well founded or whether it is 

vitiated by an error which may permit its validity to be contested.” 22 

Despite availing itself of additional time to respond, DG AGRI offers only a perfunctory and 

categorical rebuff. In fact, DG AGRI expresses the reasons for withholding the documents under the 

Article 4(3) exception in a mere 41 words: 


of the information that you request... on such a complex and sensitive issue 

currently under analysis and validation by the Commission is not appropriate. 

This statement of reasons is inadequate on its face. Article 7(1) places on DG AGRI the obligation to 

“state the reasons for a total or partial refusal.” 23 Moreover it represents an outright dismissal of the 

“concept of openness” 24 “transparency of the decision-making process,” 25 and the “right of public 

19 Public Access Regulation, Article 7 (emphasis added). 

20 Case C-64/05 P, Kingdom of Sweden v. Commission of the European Communities and Others (2007). 


22 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case T-187/03, 

Scippacercola v Commission (2005), paragraph 66. 




6 | P a g e


access to documents.” 26 There is no detailed statement with reasons for withholding any specific 

requested document included in that response. On the contrary, it merely opines generally that it 

would be “inappropriate” to release documents based on the nature of the issue (complex and 

sensitive) and its timing (currently under analysis and validation). But there is no dispute that 

biofuels are complex and sensitive issues – most environmental issues are. Nor is it disputed that the 

issues may be under analysis and validation. In fact, exercising the democratic right to participate 

early in the environmental decision-making process is one of the purposes for the request, and 

protected under Article 9 of the Aarhus Regulation titled “Public Participation Concerning Plans and 

Programmes Relating to the Environment.” 27 Therefore, reciting the nature and timing and the issue 

is not sufficient to deprive the public its right to access documents. The underlying reasons must be 

divulged, in detail, with respect to each document withheld. If the EC legislature wanted a 

categorical exception for documents related to issues of that share this nature and timing, it would 

have done so through an explicit exception. As it stands now, however, DG AGRI’s attempt to claim 

an Article 4(3) exception without providing even the most cursory elaboration for the 

inappropriateness of disclosure should be rejected outright. 

II. Disclosure Does Not Seriously Undermine Decision-making 

Even assuming that inappropriateness is a valid reason, it violates Article 4(3) by failing to meet the 

seriously-undermine standard in the exception. In other words, an Article 4(3) exception may only be 

claimed in instance in which it would seriously undermine the institution’s decision-making. The 

burden to meet the standard rests with the institution. The courts have found that the exception 

“must be interpreted and applied strictly.” 28 As noted above, the word “seriously” connotes a strong 

presumption toward disclosure, in line with the Public Access Regulation’s stated purpose in the 

preamble to “give the fullest possible effect to the right of public access to documents,” 29 taken to 

mean that “all documents of the institutions should be accessible to the public.” 30 This connotes a 

grave or acute impact from the release of the document that would effectively preclude the 

institution’s ability—here, the Commission—to make decisions. The recitals further clarify that the 

exception only entitles institutions “to protect their internal consultations and deliberations where 

necessary to safeguard their ability to carry out their tasks.” 31 The result is that only in very rare 

instances would disclosure of a document seriously undermine the decision-making and its 

withholding be necessary to safeguard their ability to carry out their tasks. Although there may be 

instances in which the Article 4(3) exception applies to environmental information, this request is 

not one of them. Moreover, the reasons offered by DG AGRI are inadequate as a matter of law. 

DG AGRI violated its obligation to provide cognizable reasons that allow Applicants to understand 

the origin and grounds for how decision-making would be seriously undermined. For the seriouslyundermine 

standard to have any force and effect, denials must be supported with reasons that 

would allow the aggrieved party to understand the basis for how decision-making would be 

undermined. This is what the interplay between Article 4(3) and Article 7(1) requires. Here, however, 

Applicants are left to divine or intuit how such disclosure would seriously undermine the institution’s 

decision-making process. As the Court stated in Kingdom of Sweden v. Commission of the European 

Communities and Others, it “is apparent in particular from Articles 7 and 8 of the regulation, the 


27 Aarhus Regulation, Article 9. 

28 Case C-64/05 P, Kingdom of Sweden v. Commission of the European Communities and Others (2007), paragraph 66; Joined cases 

C-39/05 P and C-52/05 P, Kingdom of Sweden and Maurizio Turco v. Council of the European Union (2008) 



31 Public Access Regulation, Recital 11 (emphasis added). 

7 | P a g e


institution is itself obliged to give reasons for a decision to refuse a request for access to a 

document.” 32 

In addition to narrowness of the Article 4(3) exception and the restrictive interpretation thereof, the 

Public Access Regulation requires even wider access where, as here, the documents related to the 

Commission’s delegated legislative capacity. Under Recital 6, the documents should be made 

accessible to the greatest possible extent in matters related to legislative activities: 

Wider access should be granted to documents in cases where the institutions are 

acting in their legislative capacity, including under the delegated powers, while at 

the same time preserving the effectiveness of the institution’s decision-making 

process. Such documents should be made directly accessible to the greatest 

possible extent. 33 

This wider access has been expansively interpreted. For example, in Kingdom of Sweden and 

Maurizio Turco v. Council of the European Union, the European Court of Justice rejected the 

Council’s argument that disclosure of legal documents advising the Council on legislative matters 

would undermine the Council’s decision-making. 34 Citing to Recital 6, the Court found that, on the 

contrary, openness contributed to “strengthening democracy by allowing citizens to scrutinize all 

the information which has formed the basis of a legislative act,” adding further that the “possibility 

for citizens to find out the considerations underpinning legislative action is a precondition for the 

effective exercise of their democratic rights.” 35 That the Court would not protect legal documents 

containing legal advice given to the Council—a category of documents that has traditionally enjoyed 

far more privilege under the law than inter-departmental communications or scientific and technical 

findings—underscores the particularly restrictive application of this exception when serving in a 

legislative capacity. 

Here, the documents relate to the Commission’s delegated legislative capacity. In Article 19(6) of the 

RED, the Commission is charged with submitting a report and, if appropriate, a legislative proposal: 




shall, if appropriate, be accompanied, by a proposal, based on the best available 



Directive, in particular Article 17(2). 

In the terms of reference for the “Administrative Arrangement between JRC and DG ENV on Indirect 

Land Use Change Emissions from Biofuels,” one of the few documents released, DG AGRI 

acknowledges the relationship to its delegated legislative capacity: 

The Commission has recognised that careful consideration of this problem is 

needed, further to its proposal for a Directive on the promotion of Renewable 

Energy, which includes a greenhouse gas calculation methodology as part of the 

sustainability criteria. These criteria are currently under discussion in Council and 



34 Case Nos C-39/05 P and C-52/05 P, Kingdom of Sweden and Maurizio Turco v. Council of the European Union (2008), paragraph 

46. 


46. 

8 | P a g e


the European Parliament for elaboration of a common text to be included in both 

the Renewable and Fuel Quality Directives. 36 

Because the report will form the scientific and technical basis for the any accompanying legislative 

proposal, wider access must be afforded. In Kingdom of Sweden and Maurizio Turco v. Council of the 

European Union, in debating the interest in protecting the independence of the Council’s legal 

service, the Court found that an overriding public interest is constituted by the fact that disclosure 

of documents “on legal questions arising when legislative initiatives are being debated increases the 

transparency and openness of the legislative process and strengthens the democratic right of 

European citizens to scrutinize the information which has formed the basis of a legislative act.” 37 An 

identical argument exists here. 

III. Failure to Provide a Concrete, Individual Assessment for Each Document 

It is well-settled that an institution must carry out a concrete, individual assessment of the content 

of the documents referred to in the request. 38 Courts have found that “where an institution receives 

a request for access under [the Public Access Regulation] it is required, in principle, to carry out a 

concrete, individual assessment of the content of the documents referred to in the request.” 39 This is 

made apparent in “that all exceptions mentioned in Article 4(1) to (3) are specified as being 

applicable to ‘a document.’” 40 On this point, the Court has rejected as insufficient an assessment of 

documents by reference to categories rather than on the basis of the actual information contained 

in those documents, “since the examination required of an institution must enable it to assess 

specifically whether an exception invoked actually applies to all the information contained in those 

documents.” 41 

A concrete, individual assessment is also needed to ensure compliance with other provisions of the 

Public Access Regulation, including whether redaction is appropriate under Article 4(6), the period of 

time protection is justified under Article 4(7), and compliance with obligation to provide “public 

access to a register” under Article 11(1) with an itemised list of the documents. 42 The purpose of this 

assessment must be forwarded to the applicant to serve as the basis for determining the 

applicability of the exception with respect to the document in question. 43 In response to Applicants’ 

request for access, however, these concrete, individual assessments were neither made nor made 

available. 

The request to access documents targeted all documentation on biofuels modelling in addition to 

related information: 

36 European Commission, Terms of Reference, Administrative Arrangement between JRC and DG ENV on Indirect Land Use Change 

Emissions from Biofuels, Annex No. 1 to the Offer No. H01-IES/MAM/D(08)(3349)29546, p. 1. 


67. 

38 Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraphs 69-74; see also 

Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. Council of the European 

Communities (1999), paragraph 67. 

39 Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraphs 69-74; see also 

Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. Council of the European 


40 See Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraph 70. 

41 Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraph 73, citing Case 

T-123/99, JT’s Corporation v. Commission of the European Communities (2000), paragraph 46. 

42 See Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraph 73; see also 

Public Access Regulation, Article 4(6), Article 4(7), and Article 11(1). 

43 See, e.g., Case T-2/03, Verein für Konsumenteninformation v Commission of the European Communities (2005), paragraphs 69- 

74; Case T-188/98 Kuijer v. Council of the European Union (2000), paragraph 38; Case T-14/98, Hautala v. Council of the European 


9 | P a g e


Specifically, we would like to request all the documentation (including terms of 

reference, proposals from researchers, correspondence from and to the 

Commission, minutes of working meetings, datafiles with inputs and results, draft, 

interim, and final reports etc.) related to the modelling of the impacts of indirect 

land-use change cause by increased biofuels production performed for the 

Commission by the IPTS section of JRC, and by other consultants if applicable, as of 

1 January 2009. 

The word “including” in the parenthetical is illustrative in nature, not exhaustive, of the various 

categories of documentation sought. It was intended to provide guidance on the broad range of 

documents that Applicants considered “related to the modelling.” 

But DG AGRI reveals that it assessed the documents in the aggregate, rather than individually, if at 

all, since it simply repeats the categories of documents included in the descriptive parenthetical in 

the original request: 


of the information that you requested (namely, proposals from researchers, 

correspondence, minutes of working meetings, data files and draft interim reports) 

on such a complex and sensitive issue currently under analysis and validation by the 

Commission is not appropriate. 

This response evinces a failure to assess in concrete and individual manner each document 

requested, as required. 44 Nor did DG AGRI provide an itemised list of the documents on which the 

assessments were purportedly performed and the reasons for the claim to exception for each 

document in question. In addition, no public access to a register of documents under Article 11(1) 

was likewise provided. As a result, Applicants are precluded from being able to determine for each 

document whether the decision is well founded or whether it is vitiated by an error which may 

permit its validity to fall within Article 4(3) to be contested. 45 

IV. Failure to Consider Redaction of Documents to Allow Disclosure or Determine the 

Period of Application of the Exception 

Only in the rarest of occasions are documents with environmental information relating to emissions 

in the environment to be withheld from public access. This request is not one of them. 

Nevertheless, to the extent that any document was properly withheld, DG AGRI failed to consider 

redaction of the documents or determine the period of application of the exception. 

Upon a finding that documents with environmental information may be withheld, two additional 

determinations must be made. First, under Article 4(6), “if only parts of the requested document are 

covered by any of the exceptions, the remaining parts of the document shall be released.” 46 In WWF 

European Policy Programme v. Council of the European Union, the Court interpreted this language to 

mean that documents must be redacted, if possible, to allow their disclosure: 

It is clear from the wording itself of Article 4(6) of [the Public Access Regulation] that 

an institution is required to consider whether it is appropriate to grant partial access 

to documents requested and to confine any refusal to information covered by the 

44 Public Access Regulation, Article 4(3); Case T-2/03, Verein für Konsumenteninformation v Commission of the European 


45 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case T-187/03, 

Scippacercola v. Commission (2005), paragraph 66. 


10 | P a g e


relevant exceptions. The institution must grant partial access if the aim pursued by 

that institution in refusing access to a document may be achieved where all that is 

required of the institution is to blank out the passages which might harm the public 

interest to be protected. 47 

This requirement is further supported by the restrictive interpretation afforded to documents 

containing environmental information, especially information related to emissions in the 

environment. 48 By its own admission, DG AGRI, in denying the request, stated that “the subject 

matter of your request relates to emissions into the environment.” 49 Second, under Article 4(7), the 

exceptions “shall only apply for the period during which protection is justified on the basis of the 

content of the document.” 50 This determination is, likewise, document and content specific. DG AGRI 

simply failed to perform either analysis and make it available to Applicants in its response. 

V. Any Claim to Exception Is Defeated by an Overriding Public Interest in Disclosure 

The Public Access Regulation also contains an exception to the Article 4(3) exception. In other words 

if a document falls under the narrow category of documents whose disclosure would seriously 

undermine the decision-making process and whose withholding is necessary to carry out the 

Commission’s tasks, an “overriding public interest in disclosure” will nevertheless compel its release. 

The burden on an institution claiming the Article 4(3) exception with respect to environmental 

information is heightened. Here, the disclosure of the requested documents falls within the 

exception to the Article 4(3) exception since there is an overriding public interest in disclosure. 

As an initial matter—and it bears repeating—to even consider the exception to the Article 4(3) 

exception, it must already be found that DG AGRI’s 41-word response: (i) provides a detailed 

statement of reasons with sufficient information to determine whether the decision is well-founded 

or vitiated by an error; (ii) sets out reasons that meet the seriously-undermine standard; (iii) 

establishes that denial was necessary for the Commission to perform its tasks; (iv) demonstrates that 

the documents cannot be redacted to nevertheless allow disclosure; and (v) overcomes the 

restrictive interpretation of the Article 4(3) exception for requests for environmental information. 

Moreover, this assessment must have been made for each document individually on each point. 

These requirements have not been met. 

Nevertheless, assuming arguendo a valid claim to an Article 4(3) exception here, the public interest 

overrides it. As a general matter, the public interest in reducing greenhouse-gas emissions to curb 

climate change is irrefutable. In fact, at the time of submission of this confirmatory application, the 

leaders of the world are urgently negotiating a new climate treaty in Copenhagen. The public has 

every right to be fully informed and involved to ensure that EU climate policies, such as setting 

biofuel targets, do not overstate greenhouse-gas-emission reductions or, as many suspect, actually 

increase greenhouse-gas emissions. 51 The public also has an irrefutable interest in ensuring that 

biofuel targets do not result in the destruction of forests and loss of biodiversity. 52 The increase in 

biofuel consumption without adequate safeguards, however, is expected to result in the conversion 

of natural and biodiverse lands into cropland as a result of biofuel production – or indirect land-use 

47 Case T-264/04, WWF European Policy Programme v. Council of the European Union (2007), paragraph 36, citing Case C-353/99 

P, Council v. Hautala (2001), paragraph 29. 

48 Aarhus Regulation, Recital 15 and Article 6(1). 

49 European Commission, Directorate-General for Agriculture and Rural Development, Communication Denial Request for Access 

to Documents (27 November 2009), p.1. 


51 European Commission, Terms of Reference, Administrative Arrangement between JRC and DG ENV on Indirect Land Use Change 

Emissions from Biofuels, Annex No. 1 to the Offer No. H01-IES/MAM/D(08)(3349)29546, p. 1. 

52 RED, Recital 69; FQD, Recital 11. 

11 | P a g e


change, as it were. 53 Both these interests—the change in the Earth’s climate and the conservation of 

biological diversity—are recognised as “common concern[s] for humankind” in treaties signed and 

ratified by the European Union. 54 

More specifically, however, provisions in RED and FQD set targets that artificially increase the 

demand for biofuels. That is their purpose under the assumption that it will reduce greenhouse-gas 

emissions. But several scientific studies published in reputable periodicals have concluded that 

biofuels may actually increase greenhouse-gas emissions, especially when taking into consideration 

the impacts of indirect land-use change. 55 Many legislators felt that the existing sustainability criteria 

in the RED and FQD were inadequate to safeguard against this, prompting amendments to the 

original Commission proposal: 

In the methodology the Commission has left out potential GHG emissions 

from indirect land use change, like when the production of bio-fuels leads to 

displacement of previously grown food crops on to other lands. Such effects are by 

many experts considered to be significant. However, no global model exists to 

provide us with precise estimates of the scale of such effects. 56 

*** *** *** 

Indirect land use changes mean that when land is used for biofuel production 

instead of food production somewhere else in the world food production increases 

at the expense of tropical forests or other non-agricultural land to substitute the 

lost food production. This has a negative impact on the GHG-performance of 

biofuels, the Commission does not take this into account. Therefore it is necessary 

to introduce a factor that deducts the impact of indirect land use change from the 

default or actual GHG-values of biofuels. The factor will provide an incentive to use 

more efficient crops and to more efficient land use. 57 

With these concerns in mind, the European Parliament and Council included identical provisions in 

RED and FQD charging the Commission with producing a report “reviewing the impact of indirect 

land-use change on greenhouse gas emissions and addressing ways to minimise that impact.” 58 If 

appropriate, the report shall include an accompanying proposal that contains “a concrete 

methodology for emissions from carbon stock changes caused by indirect land-use changes,” 59 which 

may include an indirect land-use change factor. The Commission is also mandated with basing both 

the report and the proposal on the “the best available scientific evidence.” 60 The documents 

requested relate to the environmental information related to the modelling underpinning the report 

and accompanying proposal. The notion that the issues are too complex and sensitive for disclosure 

is misplaced. On the contrary, it is an argument for—not against—disclosure because it would bring 

greater transparency and peer review to the underlying technical, scientific and political 

considerations upon which this critical policy decision is based. 

53 See, e.g., RED, Recital 85. 

54 United Nations Framework Convention on Climate Change (UNFCCC), Recital 1; Convention on Biological Diversity (CBD), Recital 

3. 

55 See, e.g., Searchinger and Fargionne, Science Magazine (2008); The Gallagher Review for the UK Government (2008); The 

German Study by WBGU (2008); UNEP Sensitivity Analysis of GHG balances of Biofuels (2009). 

56 Opinion of the Committee on the Environment, Public Health and Food Safety for the Committee on Industry, Research and 

Energy on the proposal for a directive of the European Parliament and of the Council on the promotion of the use of energy from 

renewable sources (COM(2008)0019 – C6-0046/2008 – 2008/0016(COD)). 

57 European Parliament, Report of 26 September 2008 on the proposal for a directive of the European Parliament and of the 

Council on the promotion of the use of energy from renewable sources, (COM(2008)0019 – C6-0046/2008 – 2008/0016(COD)), 

Opinion of the Committee on the Environment, Public Health and food safety, Amendment 86 (emphasis added). 




12 | P a g e


Moreover, the requested documents are related to the development of indirect land-use change 

policy and strategy. Article 12(3) states that, “[w]here possible, other documents, notably 

documents relating to the development of policy or strategy, should be made directly available.” 61 In 

fact, the emphasis on the issues being “under analysis” is misguided. In the instance of policy and 

strategy development, as here, it is precisely the analysis that should be made available. Moreover, 

DG AGRI failed to state any reason why the public interest does not override its refusal to disclose 

the documents, as required in Article 7(1) and case-law. 62 Applicants are therefore prevented from 

determining whether the exception actually applies to all the information contained in the 

documents. 

CONCLUSION 

With this confirmatory application for reconsideration, Applicants respectfully request that the 

Secretary-General grant access to the requested documents and information therein, providing 

access in accordance with Article 10. 

In the event that that this confirmatory application is rejected, however, Applicants request that the 

following actions be taken: 

• Preserve all documents, as defined in Article 3(a) of the Public Access Regulation, and 

environmental information, as defined in Article 2(1)(d) of the Aarhus Regulation, that were 

requested in the 15 October 2009 application for access to documents; and 

• Provide a concrete, individual assessment of compliance with the requirements in both the 

Public Access Regulation and Aarhus Regulation for each requested document, including, but 

not limited to, the reasons for denial of access, compliance with the relevant requirements, 

and time duration of the refusal. 

In addition, in the event of any denial, Applicants hereby request under Article 6 of the Public Access 

Regulation access to all documents relevant to the processing of the initial application made on 15 

October 2009 and this confirmatory application for reconsideration, including, but not limited to, 

communications from other Directorate-Generals regarding the request. 

Sincerely, 

Jos Dings 


Director 

Tim Grabiel 


Climate & Forests Lawyer 

Konstantin Kreiser 

Birdlife International – European Division 

EU Policy Manager 

__________________________ 

John Hontelez 


Director 


62 Case T-2/03, Verein für Konsumenteninformation v. Commission of the European Communities (2005), paragraph 73, citing Case 

T-123/99, JT’s Corporation v. Commission of the European Communities (2000), paragraph 46. 

13 | P a g e


For further information please contact: 


Tim Grabiel 

t +32 (0)2 893 0846 m +33 (0)6 32 76 77 04 

m +32 (0)488 574 418 

tgrabiel@clientearth.org 

nusa.urbancic@transportenvironment.org 

www.transportenviornment.org 

www.clientearth.org 

Rue d’Edimbourg, 3rd Floor, Mundo-B Avenue de Tervuren 36 

Brussels 1050 Brussels 1040 

*** *** *** 

Transport & Environment is an independent not-for-profit organisation whose mission is to promote transport 

policy that is based on science and the principles of sustainable development to both minimise the use of 

energy and land—and associated impacts on the environment and health—while maximising safety and 

guaranteeing sufficient access for all. The Brussels-based team focuses on the areas where European Union 

policy has the potential to achieve the greatest environmental benefits, including technical standards for 

vehicle fuel efficiency and pollutant emissions, environmental regulation of international transport, such as 

aviation and shipping, European rules on infrastructure pricing, and environmental regulation of energy used 

in transport. Established in 1990, Transport & Environment represents over 50 organisations across Europe, 

mostly environmental groups and sustainable transport campaigners. 

ClientEarth is a non-profit organization dedicated to safeguarding the planet—its flora, fauna, ecosystems, 

people—for the benefit current and future generations. Founded in 2007, ClientEarth attorneys are ushering a 

new era of environmental protection in Europe and beyond, pioneering innovative ways to protect the 

environment through the power of law. With offices in London and in Brussels, ClientEarth provides legal and 

technical capacity to the environmental movement. Activities focus on transformational changes to the 

European legal and legislative landscape, including opening up the European courts to citizen suits, advocating 

for effective environmental legislation with binding and enforceable provisions, bringing transparency to 

European decision-making; and empowering non-governmental organizations. 

BirdLife International is a global Partnership of conservation organizations that strives to conserve birds, their 

habitats and global biodiversity, working with people towards sustainability in the use of natural resources. 

BirdLife partners operate in me than 100 countries and territories worldwide. BirdLife International is 

represented in 43 countries in Europe and is active in all EU Member States. 

The EEB is a federation of more than 140 environmental citizens’ organisations based in all EU Member States 

and most Accession Countries, as well as in a few neighbouring countries. These organisations range from local 

and national, to European and international. The aim of the EEB is to protect and improve the environment of 

Europe and to enable the citizens of Europe to play their part in achieving that goal. 

14 | P a g e

ANNEX A.12


ANNEX A.13

24/02/2010 Transport & Environment Mail - Confir… 



Confirmatory application for access to documents 

pursuant to Regulation 1049/2001 --- GESTDEM 

2009/4261 

Sg-Acc-Doc@ec.europa.eu 9 February 2010 18:29 


Dear Ms Urbancic, 

Kindly find herewith a letter concerning your confirmatory application for access to documents (gestdem 

2009/4261). 

 


Paul SIMON 

Unit SG.E3, Transparency, Relations with Stakeholders and External Organisations 


Urbancic (4261) 2nd letter- EN.pdf 

218K 





Ref. Ares(2010)70321 - 09/02/2010 

Direction E 

SG-E-3 


Brussels, 09.02.2010 

SG.E3/MM/psi - Ares (2010)70321 

Ms Nuša Urbančič 


Rue d'Edimbourg, 26 

1050 Brussels 

By email only: 

nusa.urbancic(a)transDortenvironment.org 

Subject: Confirmatory application registered under GESTDEM No 2009/4261 

- proposal for a fair solution 

Dear Ms Urbančič, 

Thank you for your letter of 8 February 2010, by which you react to our proposal for a 

fair solution sent earlier on the same day. 

We are aware that the extended time limit for handling your application expires today. 

However, as indicated in our previous letter, we have not completed the analysis of the 

requested documents and are, therefore, not in a position to take a final decision on your 

application for access. Formally, you are entitled to bring proceedings to the General 

Court of the EU or to lodge a complaint to the European Ombudsman. I would like to 

stress, however, that we are handling your application and the fact that we are not in a 

position to take a final decision should not be interpreted as an implicit rejection of your 

application. 

Regulation 1049/2001 lays down time limits for the handling of applications for access to 

documents with a view to ensure that applicants may seek redress when an institution 

fails to respond to a request for access after the deadline has expired. However, the 

Regulation also provides in Article 6(3) for a possibility to confer with applicants in 

order to find a fair solution when an application concerns a very large number of 

documents. In your letter you mention that the Commission did not use this possibility 

when handling your application at the initial stage and you suggest that this provision 

cannot apply at the confirmatory stage. There is nothing in the wording of the Regulation 

that supports this interpretation of Article 6(3). As we explained in our previous letter, 

the Secretariat-General conducts an independent review of the application. This means, 

first, a proper identification of all documents covered by the application and, second, a 

concrete analysis of all documents with a view to determine whether or not full or partial 

access can be granted. 

Commission européenne, B-1049 Bruxelles / Europese Commissie, B-1049 Brussel - Belgium. Telephone: (32-2) 299 lill. 

hltp://cc.europn.eu/dgs/secretariat_general 

E-mail: sii-acĽ-cloĽflJtiľĽ.curona.Ľu

We have by now identified around 200 documents which fall within the scope of your 

request and we have made progress with the concrete analysis of a number of them. 

Therefore, we are not in a situation where access has implicitly been refused. On the 

contrary, we are carrying out a concrete analysis of all documents with a view to grant 

you the widest possible access. 

You indicate in your letter that you wish the following documents to be disclosed by 

close of business today: 

• the external study carried out by IFPRI; 

• the study using the AGRI/OECD partial equilibrium AgLink model; 

• the assessment of other existing modelling exercises carried out by the Joint Research 

Centre; 

• the literature review. 


While it is clear that these documents cannot be disclosed today, we will take this order 

of priority into account when carrying out our analysis. 

Once again, I would like to stress that we are committed to granting you the widest 

possible access to the documents which you seek to obtain, based on a proper 

examination of each and every document. Given the great number of documents, the 

complexity of the subject matter and the need to consult third parties on some documents, 

it cannot be reasonably expected that this work be completed within the normal time 

limits et out in Regulation 1049/2001. We therefore maintain our proposal for a fair 

solution on the meaning of Article 6(3) of the Regulation. 

Finally, you request a meeting with the relevant Commission staff. It is not clear what the 

purpose of such a meeting would be. A discussion on the substance would be premature 

as the Commission has not completed the studies on the modelling of impacts from 

indirect land-use change as a result of European Union biofuel targets. The Commission 

intends to publish the studies upon completion. This will be the right opportunity to have a 

discussion on the substance. For the moment, we are only looking at the possibility to 

disclose documents at this early stage of the decision-making. 

I hope that we can find a solution which reconciles your right of access with the 

administrative burden entailed by your application. 


Marc MAES 

Deputy Head of unit

ANNEX A.14

24/02/2010 Transport & Environment Mail - Your … 



Your confirmatory application for access to documents 

under Regulation (EC) N° 1049/2001 - GESTDEM 

2009/4261 

Sg-Acc-Doc@ec.europa.eu 22 February 2010 11:03 


Dear Mrs. Urbancic, 

Kindly find a first response to your confirmatory application concerning your request for access to 

documents pursuant to Regulation (EC) N° 1049/2001 regarding public access to European Parliament, 

Council and Commission documents (GESTDEM 2009/4261). 

Please, note that the following sentence contains an error, in the third paragraph on the second page, 

before the signature of Ms Day: "In addition, I am pleased to inform you that full access can also be 

granted to 60 other documents, all listed in the attached annex". Please, read "59 other documents" 

instead of "60 other documents". We apologise for this mistake. 


Paul SIMON 

Unit SG.E3, Transparency, Relations with Stakeholders and External Organisations 


2 attachments 

Urbancic (4261) - réponse partielle - EN.pdf 

219K 

Urbancic (4261) - réponse partielle - List of the fully disclosed documents.xls 

31K 




Ref. Ares(2010)93239 - 22/02/2010 


The Secretary General 

Brussels, 22/02/2010 

SG.E3/MIB/HP/rc/psi - Ares (2010)93239 




1050 Brussels 


nusa.urbancic(a>transDortenvironmenLors 

Subject: 




I refer to your letter dated 17 December 2009, registered on 18 December 2009 by which, 

pursuant to Regulation 1049/2001 regarding public access to European Parliament, 

Council and Commission documents 1 , you lodge a confirmatory application concerning 

the decision of Directorate-General for Agriculture and Rural Development of 27 

November 2009 to refuse access to "all the documents related to modelling of impacts of 

indirect land-use change caused by increased biofuels production performed for the 

Commission by the IPTS Section of the Joint Research Centre and by other consultants, if 

applicable, as ofl January 2009". 

I also refer to the letters of 19 January and of 8 and 9 February 2010 from my services by 

which, respectively, the time-limit for replying to your request was extended and a 

proposal for a fair solution in accordance with Article 6(3) of the Regulation was 

submitted to you. This proposal aimed at reconciling your interest in receiving a swift 

reply to your confirmatory application with the need for the Commission to carry out a 

concrete and individual analysis of the numerous documents concerned. It consisted of 

proposing that the Commission gives a first partial reply as soon as it has completed its 

examination of a first set of the documents concerned. One or more complementary 

replies would follow as the analysis of the remaining documents leads to a decision on 

disclosure. In this respect, you were invited, if you so wished, to indicate an order of 

priority for the handling of the request, indicating which documents you wished to 

receive first. 

' 

OJL145, 31.05.2001, p.43. 


http://ec.europa.eu./das/secretariat general 

E-mail: sa-acc-doc®ec.europa.eu

In your reply of 8 February 2010, you proposed the following list of documents - and any 

drafts - for expedited review: 

• the external study carried out by IFPRI based on a general equilibrium modelling 

approach and using an extended version of the GTAP database and the Mirage 

model; [item 1] 

• the study using the AGRl/OECD partial equilibrium AgLink model (performed for the 

Commission by the IPTS section ofJRC); [item 2] 

• the assessment of other existing modelling exercises carried out by the Commission's 

JR C; [item 3] 

• the literature review, [item 4] 

We have subsequently taken into account this order of priority when carrying out our 

analysis. 

As a result, having carefully re-examined your request and the undisclosed documents 

concerned on the basis of Regulation 1049/2001, I am pleased to inform you that full 

access can be granted to the documents falling under the second and fourth items of your 

priority list above. 

Regarding the second item. 12 documents were identified. These documents - listed in 

the attached annex - are successive draft versions of the study, which is not yet finalised. 

As for the fourth item, 2 documents were identified, which are two draft versions of the 

literature review, also not yet finalised. 

In addition, I am pleased to inform you that full access can also be granted to 60 other 

documents, all listed in the attached annex. 

Given the volume, we will send you a CD-Rom containing the concerned documents by 

normal mail or, if you prefer, you can receive it by hand at our office in the Berlaymont 

building. Please confirm your preference by e-mail to sg-acc-doc(a),ec.euiOpa.eu by 

Tuesday next. If no reply is received at that date, we will send you the CD-Rom by 

normal mail. 

With regard to the remaining documents falling under the scope of your request - 

including those falling under items 1 and 3 of your priority list above -1 have to inform 

you that we are still in the process of analysing them. We aim to send you one or more 

complementary replies regarding disclosure of these remaining documents separately 

within the shortest possible delay. 



Enclosures 

Catherine Day


LIST OF THE FULLY DISCLOSED DOCUMENTS 

Documents Other documents 

disclosed on full 

access 

falling under the 8/2/2010 request Doc Nr SUBJECT DATE FROM TO 

AGRI 1 Deliverable 2c IMAP2009 Biofuel report content v1 7/6/2009 JRC-IPTS X 

AGRI 2 Biofuel assumptions 6/22/2009 AGRI X 

AGRI 3 Powerpoint presentation Brussels 23/7/09 ESIM results 7/23/2009 JRC-IPTS X 

AGRI 4 Powerpoint presentation Brussels 23/7/09 CAPRI results 7/23/2009 JRC-IPTS X 

AGRI 5 Powerpoint presentation Brussels 23/7/09 AGLINK-COSIMO results 7/23/2009 JRC-IPTS X 

AGRI 6 IPTS biofuel report v1 draft 16July09.doc 7/16/2009 JRC-IPTS X draft v1 

AGRI 7 IPTS biofuel report v2 AGLINKpart-draft 7/20/2009 JRC-IPTS X draft v2a 

AGRI 8 IPTS biofuel report v2 CAPRIpart-draft 7/20/2009 JRC-IPTS X draft v2b 

AGRI 9 IPTS biofuel report v3 draft 31July09.doc 7/31/2009 JRC-IPTS X draft v3 

AGRI 10 IPTS biofuel report v5 draft 9/17/2009 JRC-IPTS X draft v5 




AGRI 14 new yield scenario on v 8 10/23/2009 JRC-IPTS X 


AGRI 16 IPTS biofuel report v10 final draft 10/30/2009 JRC-IPTS X draft v10 

AGRI 17 finaldraft v11 biofuel report 12/23/2009 JRC-IPTS X draft v11 

AGRI 18 additional comments on version 4 - email 9/9/2009 AGRI JRC-IPTS X 

AGRI 19 additional comments on version 4 - email 8/31/2009 AGRI JRC-IPTS X 

AGRI 20 questions from TRADE on presentations - email 7/23/2009 TRADE JRC-IPTS X 

AGRI 21 comments on structure of report - email 7/13/2009 AGRI JRC-IPTS X 

AGRI 22 comments on workplan - email 3/30/2009 AGRI JRC-IPTS X 

AGRI 23 email exchange on AGLINK model with OECD 6/4/2009 AGRI OECD X 

AGRI 24 email exchange on II generation in AGLINK 6/17/2009 AGRI JRC-IPTS X 

AGRI 25 note on biofuel modelling accompanying v11 12/23/2009 JRC-IPTS AGRI X 

AGRI 26 data transmission to JRC ISPRA 11/24/2009 JRC-IPTS JRC-ISPRA X 

AGRI 27 echange with TREN on fuel consumption 6/11/2009 TREN AGRI X 

AGRI 28 echange on DDGS 9/25/2009 AGRI JRC-IPTS X 

AGRI 29 central scenario results 7/17/2009 JRC-IPTS AGRI X 

AGRI 30 proposed structure 7/6/2009 JRC-IPTS AGRI X 

AGRI 31 model descriptions 5/11/2009 JRC-IPTS AGRI X 

AGRI 32 model limitations and description of workplan 4/17/2009 JRC-IPTS AGRI X 

AGRI 33 IPTS biofuel report v4 draft 8/19/2009 JRC-IPTS AGRI X draft v4 

TREN 1 Market Analysis Oils and Fats 12/17/2009 Frank Bergmans DEURWAARDER for Fuels Ewout (TREN) X 

TREN 2 Article from WARWICK LYWOOD*, JOHN PINKNEY* AND S AM COCKERILL : "Impact of protein concentrate coproducts on net land requirement for European biofuel production" 12/10/2009 Warwick Lywood European Commission services X 

TREN 3 Support material used during the 2nk policy workshop on "A land-use modelling framework for the European Commission" (Nov 2009) 11/25/2009 Patricia Benito ( European Commission services X 

TREN 4 European Forest Institute, Joensuu, Finland Paper on aforestation in Europe 11/20/2009 Warwick Lywood TREN D1 X 

TREN 5 Link to the webpage of the conference "A stakeholoder debate on the science and policy impacts of sustainable biofuel production" Lausanne, Switzerland, 9-10 November 2009 11/19/2009 Rob Cox (IPIECA) X 

TREN 6 Answer to a question from P. Hodson regarding Best use of land + sending of a representation from R. Matthews (forest research in UK) 11/18/2009 J. Woods (Impe P. Hodson X 

TREN 7 Two reports part of the RFA annual report on ILUC: 11/16/2009 Aaron Berry (Re TREN X 

- ECOFYS report: "Mitigatingindirect effects of biofuel production Case studies and Methodology" 

- Econometrica report : "Methodology and Evidence Base on the Indirect Greenhouse Gas Effects of Using Wastes, Residues, and By-products for Biofuels and Bioenergy" 

TREN 8 Peer Review of the REA biofuel scenario modeling by the Imperial College + slides presentation 11/16/2009 David Knibbs (V P. Hodson X 

TREN 9 Article in Sciencexpress: "Inidrect Emissions from Biofuels: How Important?" 11/11/2009 Rob Vierhout (ebP. Hodson X 

TREN 10 Report from ifeu - Institut für Energieund Umweltforschung Heidelberg GmbH: "Synopsis of Current Models and Methods Applicable to Indirect Land-Use Change (ILUC)" 11/9/2009 Dietrich Klein (B P. Hodson X 

TREN 11 Article REVIEW OF THE USEPA'S EISA NPRM, by Pelvin 10/22/2009 Michael O'Hare P. Hodson X 

TREN 12 Working "Killing Two Birds with One Stone: US and EU Biofuel Programmes" by David Tréguer and Jean Marc Bourgeon 10/22/2009 Sylvie La Mantia (INRA) X 

TREN 13 Lind to a UNEP report on biofuels 10/16/2009 Simon Worthing TREN D1 X 

TREN 14 Ecofys report on indirect impacts of biofuels 10/11/2009 Stijn Cornelissen (Ecofys) X 

TREN 15 Comments on Ozdemir paper on DDGS utilisation 10/9/2009 Warwick Lywood P. Hodson X 

TREN 16 Note from Warwick Lywood on Issues of concern with models for calculating GHG emissions from indirect land use change (version 1) 10/8/2009 Warwick Lywood (ensus group) X 

TREN 17 Paper from Jesper Hedal Klovierpris - Kenneth Baltzer - Per H. Nielsen on LCI modelling of Land Use 9/22/2009 Jesper Hedal Kl TREN D1 X 

TREN 18 Ensus paper: comparison between EPA/MODIS and FAO data of forest and crop area changes 9/18/2009 Warwick Lywood P. Hodson X 

TREN 19 spreadsheet used by Inconebrasil to calculateCO2 emissions using GTAP results 9/16/2009 Andre Nassar (icEuropean Commission services X 

TREN 20 Slides on Economic modelling of the expansion of Biofuels production 9/16/2009 Andre Nassar (icP. Hodson X 

TREN 21 Palm oil study from WWF 9/15/2009 WWF X 

TREN 22 Three articles on carbon footprints of biofuels 9/14/2009 Eric Johnson (A P. Hodson X 

TREN 23 Inconebrazil paper analyzing the CHG emissions associated to land use change due to sugarcane expansion in Brazil 9/14/2009 Andre Nassar (iconebrasil) X 

TREN 24 Slides on "How to link Blum and Aglink Models regarding sugar cane expansion" 9/13/2009 Andre Nassar (icP. Hodson X 

TREN 25 Comments from Steven Wonink (ECOFYS) to P. Hodson 9/8/2009 Steven Wonink P. Hodson X 

TREN 26 Answer to technical question from P. Hodson on reasons for soya growth 8/7/2009 Warwick Lywood (ensus group) X 

TREN 27 final versions of the Ensus papers: "The relative contributions of changes in yield and land area to increasing crop output" and "Impact of protein concentrate co-products on net land requirement for European biofuel production" 8/7/2009 Warwick Lywood P. Hodson X 

TREN 28 question relating to the conservatism of default values 7/17/2009 P. Hodson Aaron Berry (RFA) X 

TREN 29 Terms of reference of a study launched by RFA Pvovision of consultancy services to develop a methodology and evidence base on practical solutions foravoiding indirect land use change in the promotion fo bioenergy 7/16/2009 Aaron Berry (Re P. Hodson X 

TREN 30 Comments on the Commission pre-consultation on "Indirect land use change - Possible elements of a policy approach - preparatory draft for stakeholder/expert comments" 7/10/2009 Michael O'Hare P. Hodson X 

- Article from MO'HARE R J Plevin, J I Martin, A D Jones, A Kendall and E Hopson: "Proper accounting for time increases crop-based biofuels' greenhouse gas deficit versus petroleum" 

TREN 31 Link to a scientific article: "A common sense theory of indirect land use change" (biofueldigests) 7/10/2009 James Primrose P. Hodson X 

TREN 32 Ensus approach for ILUC modelling 6/29/2009 Warwick Lywood P. Hodson and E. Deurwaarder X 

TREN 33 Slides from J. Woods and Andre Faaij:" Accounting for LUC / ILUC for the biofuel sector" (roundtable discussion of 28 May on the issue of indirect land-use change (ILUC) associated with biofuels 6/11/2009 Nina Sjurseike (The Center) X 

TREN 34 Econometrica paper on Consequential and attributional approached to LCA 6/5/2009 Richard Tipper ( P. Hodson X 

TREN 35 Unica contribution sent to the Californian Air Resources Board chairman regarding the proposed low carbon fuel standard 4/26/2009 Arnaldo Walter P. Hodson X 

TREN 36 Note from M. Ruete to M. Demarty and M. O'Sullivan on Commission modelling exercises on land use change and biofuels 12/3/2009 X 

TREN 37 Reply from M. Demarty to M. RUETE's note on Commission modelling exercises on land use change and biofuels 12/11/2009 X 

TREN 38 Provisional outline of the part describing the modelling results 7/29/2009 X 

TREN 39 Literature review-DG TREN internal work-Last updated version 10/31/2009 X 

TREN 40 Literature review-DG TREN internal work-Version 21/10/2009 10/21/2009 X 

14 59

ANNEX A.15



Ref. Ares(2010)389128 - 02/07/2010 


The Secretary General 

Brussels, 02/07/2010 

SG.E3/MIB/HP/rc/psi - sg.e.3(2010)421121 




1050 Brussels 


nusa.urbancic(ū)transvortenvironment.ors 

Subject: 




I refer to your letter dated 17 December 2009, registered on 18 December 2009 by which, 

pursuant to Regulation 1049/2001 regarding public access to European Parliament, 

Council and Commission documents 1 , you lodge a confirmatory application concerning 

the decision of Directorate-General for Agriculture and Rural Development of 27 

November 2009 to refuse access to "all the documents related to modelling of impacts of 

indirect land-use change caused by increased biofuels production performed for the 

Commission by the IPTS Section of the Joint Research Centre and by other consultants, if 

applicable, as ofl January 2009". 

I also refer to the letters of 19 January and of 8 and 9 February 2010 from my services by 

which, respectively, the time-limit for replying to your request was extended and a 

proposal for a fair solution in accordance with Article 6(3) of the Regulation was 

submitted to you. This proposal aimed at reconciling your interest in receiving a swift 

reply to your confirmatory application with the need for the Commission to carry out a 

concrete and individual analysis of the numerous documents concerned. It consisted of 

proposing that the Commission gives a first partial reply as soon as it has completed its 

examination of a first set of the documents concerned. One or more complementary 

replies would follow as the analysis of the remaining documents leads to a decision on 

disclosure. 

On the basis of your application I have identified 140 documents as falling within the 

scope of your request. For ease of reference, you will find attached a list of all the 

concerned documents, along with the corresponding decision on disclosure, i.e. full 

access (FA), partial access (PA) or no access (NA). 

1 

OJL145, 31.05.2001, p.43. 


http://ec,europa.eu./das/secretariat general 

E-mail: sa-acc-doc@ec,europa.eu


With my decision of 22 February 2010 2 , Ml access was granted to 74 documents, namely 

documents AGRI 1 - 33 as well as TREN 1 - 40 in the annexed list 3 . 

As to documents AGRI 34, ENV 1 and TRADE 8, these documents have already been 

disclosed in the initial stage. 

The present decision deals with the remaining 63 documents falling within the scope of 

your above referred application, namely docs TREN 41 and 42, ENV 2 - 35, JRC 1 - 12, 

TRADE 1-7 and 9-16. 

ASSESSMENT UNDER THE REGULATION 

The Commission has made a careM assessment of the remaining documents and I 

am pleased to inform you that Ml access is granted to 37 documents, namely 

documents TREN 42, ENV 2, ENV 4-5, ENV 7-21, ENV 30 - 35, JRC 1 - 7, 

JRC 9-11, and TRADE 6 -7. Copies of these documents are attached to this letter. 

Please note that some parts of documents ENV 12 to ENV 21 (notes of meetings) 

concern issues not related to your request for access. These parts have been 

suppressed since they are out of the scope of your request. 

Moreover, I am pleased to inform you that, pursuant to Article 4(6) of the 

Regulation, partial access is granted to documents ENV 3 and ENV 6. You will find 

expunged versions of these documents attached to this letter. 

The non-disclosed parts of ENV 3 and ENV 6, as well as the remaining 24 

documents (TREN 41, ENV 22 - 29, JRC 8, JRC 12, TRADE 1- 5, TRADE 9-16), 

all identified in the attached list, have to be refused since they are covered by one or 

more of the exceptions provided for in Regulation 1049/2001. These are the 

protection of the public interest as regards international relations (Article 4(1) (a), 

3 rd indent) and the protection of the commercial interests of a natural or legal person, 

including intellectual property (Article 4(2), 1 st indent). 

THE PROTECTION OF INTERNATIONAL RELATIONS 

Document TREN 41 is an e-mail from the Mission of Canada to the European 

Union, with an annexed table regarding biodiesel imports from, among others 

Canada, into the EU in years 2008 and 2009. The e-mail discusses an issue related to 

Canada's biofuel exports and its market share in biofuels in a third country. 

In accordance with Article 4(4) of Regulation 1049/2001, the Commission has 

consulted Canada with a view to assessing whether an exception is applicable. 

I have careMly analysed the e-mail and its attachment in question which relate to 

aspects of international trade policy conducted by Canada. Disclosure would reveal 

a position expressed in these documents which would affect current and future trade 

relations of Canada with the EU and with other third counties, which in turn would 

undermine the protection of the public interest as regards the EU's international 

relations with Canada. One factor in that assessment is the fact that the Mission of 

Canada has opposed disclosure of this document. 

2 

3 

Ares (2010)93239 

TREN 7 includes two documents


Document JRC 8 contains the preliminary results of marginal calculations with AG- 

LINK model produced by the Organisation for Economic Co-operation and 

Development (OECD). The document concerned relates to research in progress 

which has not been cleared with the management of the OECD Secretariat or with 

Member Countries of the OECD. Public release of research results of such 

preliminary nature would be harmful to the image of the OECD. Taking into account 

that this international organisation has opposed disclosure of the document 

concerned, I conclude that public disclosure of this document would clearly damage 

the EU's relationship with the OECD. 

Moreover, since the information contained in this document is reflected in chapter 

4.3 (pages 25 to 32) of documents ENV 3 and ENV 6, this chapter can also not be 

released. As a consequence, partial access is granted to both these documents. 

The release of the concerned documents or parts of documents would seriously 

affect current and future relations with both Canada and the OECD, thereby 

undermining the protection of the public interest as regards the EU's international 

relations with both these actors. 

Indeed, on the one hand, it would reduce Canada's and the OECD's willingness to 

cooperate with the EU and, on the other hand, it would be prejudicial to the EU's 

strategy in its relations with both this third country and this international body. This 

would not only affect current agreements but might also interfere with future 

negotiations on similar issues with the concerned country and international 

organisation. 

Consequently, these documents or parts of documents are covered by the exception 

to the right of access expressly laid-down in Article 4(1) (a), 3 rd indent, of the 

above-mentioned Regulation according to which "the institutions shall refuse access 

to a document where disclosure would undermine the protection of the public 

interest as regards international relations", and cannot be disclosed. 

THE PROTECTION OF COMMERCIAL INTERESTS 

Documents ENV 24, 26, 29 and TRADE 1, 3, 10, 11 and 12 to 16 are draft reports, 

tables, notes, model documentation and preliminary results drawn up by the 

International Food Policy Research Institute (IFPRI), an external consultant engaged 

by the Commission's Directorate-General for Trade to carry out a study on biofiiels. 

Documents ENV 22, 23, 25, 27, 28 and TRADE 2, 4, 5 and 9 contain the 

Commission services' internal comments on those IFPRI intermediate papers. 

These provisional documents contain an incomplete and unfinished analysis from 

IFPRI and were not designed for public dissemination. They were meant to be 

internal papers submitted by IFPRI to the client (the Commission) for verification 

and discussion only. Their public release would seriously damage IFPRI's 

commercial and academic interests since they have not yet reached the same level of 

quality in content and format as that of a final product. 

Moreover, disclosure of these documents would affect IFPRI's intellectual property 

rights at an academic level. Indeed, these documents include numerous innovations 

which have a commercial value in terms of fund raising with donors. Their public 

disclosure would allow third parties to use IFPRI's research against them in future 

call for tenders. 

3

The same applies to document JRC 12, which is a draft report drawn up by an 

external consultant. 

This view was also confirmed by both external consultants. 


Therefore, these documents are covered by the exception laid down in Article 4 (2), 

first indent of Regulation 1049/2001, which stipulates that "The institutions shall 

refuse access to a document where disclosure would undermine the protection of 

commercial interests of a natural or legal person, including intellectual property". 

Consequently, they cannot be disclosed. 

The invoked exception protects commercial secrets including the intellectual 

property of a natural or legal person, as well as its commercial interests in a wider 

sense, including aspects relating to commercial reputation. 

As you know, the final version of the study carried out by IFPRI has been 

published 4 . 

OVERRIDING PUBLIC INTEREST 

Please note that the exceptions under Article 4(1) of the Regulation must not be 

balanced against an overriding public interest. 

The exception laid down in Article 4(2), first indent of the Regulation must be 

waived if there is an overriding public interest in disclosure. In order for such an 

interest to exist, it has to be public and it has to outweigh the interest to be protected 

under Article 4 of the Regulation. 

You claim that "the public interest in reducing greenhouse-gas emissions to curb 

climate change is irrefutable. (...). The public has every right to be fully informed 

and involved to ensure that EU climate policies, such as setting biofuel targets, do 

not overstate greenhouse-gas-emission reductions or, as many suspect, actually 

increase greenhouse-gas emissions. The public also has an irrefutable interest in 

ensuring that biofuel targets do not result in the destruction of forests and loss of 

biodiversity. (...). Both these interests - the change in the Earth's climate and the 

conservation of biological diversity — are recognised as "common concerns for 

humankind" in treaties signed and ratified by the European Union". 

It is obvious and certainly not contested that both the invoked interests are of a 

public nature. 

However, as above indicated, the documents concerned reflect the ĽFPRI's 

provisional analysis reached at a certain stage of their research. These provisional 

conclusions were subsequently altered in the light of their further analysis. 

Therefore, contrary to what you seem to assume, these documents cannot have the 

evidentiary value you claim. Consequently, their disclosure could not serve the 

argued public interests of "reducing greenhouse-gas emissions to curb climate 

change" or "ensuring that biofuel targets do not result in the destruction of forests 

and loss of biodiversity". The arguments you have put forward are, therefore, not 

linked to the content of the requested documents. 

Cfr. http://ec.europa.eu/trade/analvsis/chief-economist/


As you know, the final report submitted by IFPRI has been published so that the 

public can take cognisance of IFPRI's finalconclusions. 

5. MEANS OF REDRESS 

I draw your attention to the means of redress available against this decision. You 

may either bring proceedings before the General Court or file a complaint with the 

European Ombudsman under the conditions specified, respectively, in Articles 263 

and 228 TFEU. 


Enclosures 

Catherine Day

Ref. Ares(2010)389128 - 02/07/2010 

DATE 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

27/11/2009 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 

22/02/2010 


LIST OF ALL CONCERNED DOCUMENTS 

1 Deliverable 2c IMAP2009 Biofuel report content_v 

3 Powerpoint presentatioi 

4 Powerpoint presentatioi 

5 Powerpoint presentatio Brussels 23Я/09 AGLINK-COSIMO r« 

> ÌPTS_biofuei_repoit_v lrïO'sJü'iyoa'dö 

' IPTS_biofuel_report_v2_AGLINKpart-draf 

ÍÍPTSlb]o^"eiIreportlv2lČAPŘipart-draf 

i"iPTSlb]o^"eiIreportlv"Cdraft3ÏJÜÏyÖ9d"o 

) ÍPTS_biofuei_report_v5_draf 

i PT S_bii o f u e i_re ρ o ιϊ_ν 6_ d ra f 

l IPTS_biofuel_report_v7_draf 

> ÍPTS_b"io"fuei_repo"rt_v"8_draf 

Ί IPTS_biofuel_report_v9_draf 

im TRADE on presenl 

n structure of report - 

: on II generatio AGLINIipanymg 

vi 

modelling s 

6 data transmission to JRC ISPR/ 

7 échange with TREN on fuel c 

0 proposed structure 

model clescnptior 

mitations and descnptior 

> IPTS_biofuei_repo"rt_v4_dra"ft 

4 Administrative arrangement between AGRI and JRC-IPTS οι -nodellmg platform (IMAP200 

1 Market Analysis Oils and Fats for Fuels 

2 Artide from WARWICK LYWOOD", JOHN PINKNEY* AND S AM COCKERILL : "Impact of protein concentrate coproducts on net land requirement for European biofuel production" 

3 Support material used during the 2nk policy workshop on "A land-use modelling framework for the European Commission" (Nov 2009) 

4 European Forest Institute, Joensuu, Finland Paper on afore stati o n in Europe 

5 Link to the webpage of the conference 'A stakeholoder debate on the science and policy impacts of sustainable biofuel production'Lausanne, Switzerland, 9-10 November 200Ē 

6 Answer to a question from P. Hudson regarding Best use of land + sending of a representation from R. Matthews (forest research in UK) 

7 Two reports part of the RFA annual report on ¡LUC: 

ECOFYS report: "Mtigatingindirect effects of biofuel production Case studies and Methodology' 

report : "/Methodology and Evidence Base on the Indirect Greenhouse Gas Effects of Using Wastes, Residues, and By-products for Blofuels and Bioenergy 

8 Peer Review of the REA biofuel scenario modeling by the Imperial College Ides presentation 

9 Article in Sciencexpress:"ín/ďrecí Emissions from Biofuels: How important?' 

Report from ifeu -Institut für Energieund Umweltforschung Heidelberg GmbH."Synopsis of Current Models and Methods Applicable to Indirect Land-Use Change (ILUC)' 

Article RËVÏËWÖF THE ÜSEPÄ'S EISA ÑPRM, by Pelvin 

2 Working "Killing Two Birds with One Stone: US and EU Biofuel Programmes' by David Tréguer and Jean Marc Bourgeon 

3 Lind to a ¡JNËP report on biofuejs 

6 Note from Vteirwick Lywood o 

Dissions from Indirect land use 

7 Paper from Jesper Hedal K 

ing of Land Use 

8 Ensus paper: comparison between ЕРА/MODIS and FAO data of forest and crop an changes 

9 spreadsheet used by Inconebrasil to calculateC02 emissions using GTAP results 

0 Slides on Economic modelling of the expansion of Blofuels production 

il study from WWF 

2 Three articles on carbon footprints of biofuels 

I Inconebrazil paper analyzing the CHG e 

e change due to sugarcane expansion In Brazil 

4 Slides on "Howίο link Blum andAglink Models regarding sugara 

5 Comments from Steven Wbnink (ECÖFYSj to P. Hudson 

6 Answer to technical question from P. Hudson on reasons for soya growth 

7 final versions of the Ensus papers:'The reíaíiVe contributions of changes ir ι yield and land area to increasing crop output and "Impact of protein concentrate co-products on net land requirement for European biofuel product 

o the :ervatism of default valu« 

study launched by RFA Pvovis n of consultancy si s to develop a methodology and evidence base on practical solutions foravolding Indirect land use change in the promotion fo bioenergy 

0 Comments on the Commission pre-consultation on" indirect ¡and use change -Possible elements of a policy approach - preparatory draft for stakeholderfexp ert comments 

-Article from MO'HARE R J Plevin, J I Martin, A D Jones, A Kendall and E Hopson:"Properaccounf/ng for time increases crop-based biofuels' greenhouse gas deficit versus petroleum 

1 Link to a scientific artici 

e theory of indirect land u échange' (blofueldlgestsļ 

2 Ensus approach for ILUC modi Ijng 

3 Slides from J. Woods and Andre Faaij" Accounting for LUC ƒ ILUC for the biofuel sector" (roundtable discu: nof28Mayonthe e oìindirect land-use change (ILUC) associated with biofueli 

io metrica paper on Consequential and attributional approached to LCA 

a contribution sent to the Californian Air Resources Board chairman regarding the proposed low carbon fuel standard 

6 Note from M. Ruete to M. D e m arty and M. O'Sulllvan on Commission modelling exercises on land use change and biofuels 

7 Reply from M. Demarty to M. RUETE's note on Commission modelling exi ι land use change and blofuels 

8 Provisional outline of the part describing the modelling results 

■DG TREN internal work-Last updated vi 

-DG TRENïntemaï work-Version 21/10/2009 

1 Available figures on blodi' rts into the EU: e-mail from the rr n of Canada to th EU 

2 Answer to a technical questi 

f reference of administrative arrangement n" 070307/2008/517067/C3 betweer 

2!ЕхраГ ■m of administrative arrangement n" Ö703Ö7/2008/517067/C 

π Report for Administrative Agreement n" 070307/2008/517067/C 

4;Comments by DO TREN on interim report for Admimstrive Agreement n" 070307/2008/517067/( 

5ÏJRc7epiyio"DG7REN ;^c"omrnents"fo"r Administrative Agreement n 0 070307/2008/517067/C 

ëlRevised version of JRC report foiiowmg DG TREÑ's comment 

DATE 

FROM 

6/07/2009 JRC-IPTS 

22/06/2009 AGRI 

23/07/2009 JRC-IPTS 

23/07/2009 JRC-IPTS 

23/07/2009 JRC-IPTS 

16/07/2009 JRC-IPTS 

20/07/2009 JRC-IPTS 

20/07/2009 JRC-IPTS 

31/07/2009 JRC-IPTS 

17/09/2009 JRC-IPTS 

9/10/2009 JRC-IPTS 

20/10/2009 JRC-IPTS 

22/10/2009 JRC-IPTS 

23/10/2009 JRC-IPTS 

27/10/2009 JRC-IPTS 

30/10/2009 JRC-IPTS 

23/12/2009 JRC-IPTS 

9/09/2009 AGRI JRC-IPTS 


23/07/2009 TRADE JRC-IPTS 



4/06/2009 AGRI OECD 


23/12/2009 JRC-IPTS AGRI 

24/11/2009 JRC-IPTS JRC-ISPRA 

11/06/2009 TREN AGRI 







8/06/2009 

17/12/2009 Frank Bergmar DEURWAARDER Ewo ut (TREN) 

10/12/2009 Warwick Lywoi European Commission ser 

25/11/2009 Patricia Benito European Commission ser 

20/11/2009 Warwick LywoiTREN DI 

19/11/2009 Rob Cox (IPIECA) 

18/11/2009 J. Woods (Impi P. Hudson 

16/11/2009 Aaron Berry (RTREN 

16/11/2009 David Knibbs ( P. Hudson 

11/11/2009 Rob Viertioutļi P. Hudson 

9/11/2009 Dietrich Klein (P. Hudson 

22/10/2009 Michael O'Hare P. Hudson 

22/10/2009 Sylvie La Mantia (INRA) 

16/10/2009 Worthin TREN DI 

11/10/2009 Stijn Cornelissen (Ecofys) 

9/10/2009 Warwick LywoiP. Hodson 

8/10J Warwick Lywood (ensus group) 

22/09/2009 JesperHedalKTRENDI 

18/09/2009 Warwick LywoiP. Hodson 

16/09/2009 Andre Nassar ( European Commissi 

16/09/2009 Andre Nassar ( P. Hodson 

15/09/2009 WWF 

14/09/2009 Johnson (, P. Hodson 

14/09/2009 Andre Nassar (¡conebrasil) 

13/09/2009 Andre Nassar ( P. Hodson 

8Л9/2009 Steven Wbnink P. Hodson 

7ЮЗГ2003 Warwick Lywood (ensus group) 

7ЮЗГ2003 Warwick LywoiP. Hodson 

17/07/2009 P. Hodson Aaron Berry (RFA) 

16/07/2009 Aaron Berry (R P. Hodson 

10/07/2009 Michael O'Hare P. Hodson 

10/07/2009 James Primros P. Hodson 

29/06/2009 Warwick Lywoi P. Hodson and E. Deurwaarder 

11/06/2009 NinaSjurseike (The Center) 

5Ä)6/2009 Richard Tipper P. Hodson 

26/04/2009 Arnaldo Walter P. Hodson 

3/12/2009 

11/12/2009 

29/07/2009 

31/10/2009 

21/10/2009 

10/12/2009 Jeannette Pate P. Hodson 

16/07/2009 Thomas Weber (Bundesministerium für Umwel 

17/02/2009 

29/07/2009 Philip Owen (EN H OssennnkandF Raes (JRC) 

7/10/2009 R Edwards, L Marelh, A Leip, R Koeble.M Bran 

26/10/2009 DG TREN DG ENV 

e WG 

14/01/2010 JRC 

e WG

DECISION I 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

NA 

NA 

NA 

NA 

NA 

NA 

NA 

NA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

FA 

NA 

FA 

FA 

FA 

DATE 



DocNr 

ENV7 

L.Ç....baÇSr?.UM.5.9.Çy.[G.Sn 

7 ci list of parti ci pant E 

IIIŅVJ Report^fo71he"parallellessi"ons"oi"jRŒEA/OECD Expert Consultatio 

8a sessioni 

ENV9 

ENV ï Ö 

SUBJECT 

Background document, agenda, list of participants to JRC/EE/VOECD Expert Consultation "Review and inter-companson of modelling land use change effects of bioene 

7 a final draft agendi 

7 lii list of parameters/questions to be answere 

^bisessionī;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; 

8 c draft report from group ; 

^а^Ш^аА^ШШШ^^^Ш^Е^Е'!^^^^5^0^ on "Marginal Yields and land Allocation in ILUC emissions estimat 

Slides from presentations of the JRC workshop on "Marginal Yields and land Allocation in ILUC emissions estimati 

ÏÖ a Meihodoiogy and.Expenences^.aiiheÖËC 

ÏÖ lii Land Cise and Land Äiiocation in Brazilian Agncultur 

Iancl 

]^^1Е^Ш^Е^ЗША^§Ш}°^}Е^}ЕАЛ? 

allocation models and CGE modell 

ÏÖ ci Spatial Estimations of Land Carbon Stock Changes and N2 0 Emissior 

110 Θ Spatial Variation and Productivity Gap 

hO f Modeling land use change in a CGE mod« 

hO g Mapping Global Agriculture New Datasets on Agricultural Land Cover and Land U 

hO h Spatial allocation of extra areas resulting from ILUC modelir 

ENV 11 lEmails relating to the administrative agreemei 

hi a e-mail 07/10/200ί 

IH b e-mail 04/06/2009 (19 04 

111 c e-mail 04/06/2009(17 50 

111 d e-mail 04/10/200ί 

111 e e-mail 10/06/200ί 

111 f e-mail 21/04/200Ē 

111 g e-mail 17/09/200ί 

111 h e-mail 15/05/200ί 

111 ι e-mail 19/10/200Ē 

ENV 12;Notesofthe meeting of 18/02/2009 


ENV 14;Notes of the meeting of 13/03/2009 








ENV 22;Comments DG AGRI on draft report 

ENV 23;Comments DG JRC on draft report 

ENV 24;Draft final report 

ENV 25;Correspondence on scenanos forATLASS consortium report 

ENV 26;Additional tables 

ENV 27;Comments DG TREN 

ENV 28;Correspondence on draft final report among Commission seivices 

ENV 29;Answerfrom consultants to questions from Commission sen/ices 

ENV 30;Slides presented by JRC IPTS 

ENV 31 iCorrespondence between sen/ices on comments on draft final report 

ENV 32;Correspondence between sen/ices on comments on draft final report 

ENV 33;Correspondence between sen/ices on comments on draft final report 

ENV 34;Email on current status of model companson initiatives 

ENV 35;Note on "Spatial allocation of agricultural land demand" 

JRC 1 ;E-mail on "Model Comparison for ILUC: EU Commission timing" 

h a : tranmission e-mail 

h b : annex 

JRC 2;Misslon report on IEA Bloenergy task 38 Workshop on "Land Use Changes due to Bioenergy: Quantifying and Managing Climate Change and Other Environmental Impacts" 

;2 a : transmission e-mail 

;2 b : annex 

JRC 3;Mission report on Biofuel Modelling Workshop at the EEA - Copenhagen 

:3 a : transmission e-mail 

13 b : annex 

JRC 4;JRC comments on Greenergy/Econometrica paper on ILUC 


;4 b : annex 

JRC 5;JRC note on "model comparison for ILUC: overview of model comparison initiatives" 

:5 a : transmission e-mail 

15 b : annex 

JRC 6;Technical Annex for ¡LUC (marginal calculations) modeling worfc with DART model 


16 b : annex 1 

16 c : annex 2 

JRC 7;Technical Annex for Hue (marginal calculations) modeling wori( with G-TAP based model 




JRC 8 Preliminary results of marginal calculations with AG-LINK model (OECD 

JRC 9 

JRC 10 

JRC 11 

Results of Marginal calculations with DART model 

9 a : transmission e-mail 

9 b : annex 1 (Appendix DART Scenario A) 

9 c : annex 2 (Appendix DART Scenario BD) 

9 d : annex 3 (Appendix DART Scenario C) 

9 e : annex 4 (DART Calculation of marginal biofuel scenarios) 

Results of Marginal calculations with IMPACT model 


10 b : annex 

TREN's note on "Next steps for the Commission modelling of land use change and biofuels" 


DATE FROM TO 

Transmission e-mail JRC F08 I Hodgson (ENV C3), A c eDomir 

Transmission e-mail JRC F08 I Hodgson (ENV C3), A c eDomir 

Transmission e-mail L Marelli (JRC - Interseivice WG 

Transmission e-mail L Marelli (JRC - F08) 

Email 07/01/2009 

Email 09/01/2009 

11/03/2009 ATLASS consortium 

Vanous emails between 08/04/2009 and June 2009 

Email from 01/10/20ATLASS consortium 

Email Oct 2009 

Vanous emails as of November and December 2009 

Email 09/12/2009 

13/11/2009 JRC IPTS 

Email from November 2009 



10/09/2009 Jan-Erik Peters« Wide list of recipients including Co 

14/12/2009 F Ramos et al ( Interseivice WG 

24/03/2009 JRC-IE F08 Various EU-US experts 

21/04/2009 JRC internal mail (IES) 

23/04/2009 JRC internal mail (IPTS) 

29/04/2009 JRC-F08 (R. Ed Interservice WG 

10/06/2009 JRC (IE F08) Various EU-US experts ■» Interse 

12/06/2009 JRC (IE F08) KIEL Institute of the World Econ 

ЗЮ712009 JRC (IE F08) Life Cycle Associates 

24/07/2009 OECD JRC-F08 

29/07/2009 KIEL Institute c JRC-F08 

8/12/2009 IFPRI JRC-F08 

14/12/2009 TREN Interservice WG 

I

— I DECISION | DATE 



2 Results and draft report of marginal calculations with G-TAP model 

1 Biofuels Global Trade and Environmental Impact Study Final Report - ATLASS Consorţii Specific Contract №312 499 331 implementing Framework Contract N o TRADE/07/A2 

2 Biofuels - Global Trade and Env 

3 Final report - Atlass consortium 

lentmg Framework Contract № TRADE/07/A2 Biofuels Global Trade and Environmental Impact Study 

"Biofuels global trade and environmental impact study" - draft final report 

5 Report on the impact of Biofuels ËÜ Policy- draft March 2009 

/v study under the framework contract TRADE07/A2 Biofuels II 

7 The indirect land use change impact of biofuels (final) 

8 Terms of Reference for the study "Biofuels Global trade environ dm enta I impact study" (contract N 

9 CËPÜ/FPRi biofueis study Phase li note o largmal ILUCci 

0 Biofuels II Global Trade and Env ental Impact Study Preli nmary Draft Report 

1 Biofuels study II the report S tables 

TRADE 12 Questions and Answers Biofuels Study IFPRI 

TRADE 13 Sensitivity analysis strategy Directions to take for the Biofuels 

TRADE 14;Biofuel Impact II Extended model 

TRADE 15;Biofuel Impact II Preliminary results 

TRADE leiBaselmes, Scenanos and Sensitivity Analysis 

Alia Golub JRC-F08 

Antoine Bouet 

DG ENV 

Antoine Bouet 

DG TREN 

mars-09 DG AGRI 

24/06/2009 Bertin Martens Lionel Fontagné 

29/06/2009 

29/06/2009 

30/09/2009 

9 Bertin Martens Inters 

9 

David Laborde 

David Laborde 

David Laborde

Schedule of Annexes and Appendix - ClientEarth

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?