1

Supplementary Information for Market-mediated responses confound policies to limit deforestation from oil palm expansion in Malaysia and Indonesia Farzad Taheripour, Thomas Hertel, and Navin Ramankutty Corresponding Author: Farzad Taheripour Email: tfarzad@purdue.edu 1. Introduction

This document presents the following materials:

- Literature of environmental impacts of producing oil crops in tropical areas,

- Sustainability certificates,

- Links to the contemporary debate about imposing a ban on palm oil produced in M&I

and its economic and environmental impacts,

- GTAP-BIO model and its background,

www.pnas.org/cgi/doi/10.1073/pnas.1903476116

2

- Description of the examined experiments,

- Instructions for replicating results,

- Systematic Sensitivity Analysis,

- Supporting figures and tables.

2. Literature of environmental impacts of producing oil crops in tropical areas

Previous studies (1-16) have examined the environmental consequences of agricultural

activities in tropical areas in considerable detail. A group of these studies highlighted the

land use changes, releases of terrestrial carbon, and losses in biodiversity due to oil palm

plantations in Southeast Asia, basically in Malaysia and Indonesia (1-8). Others examined

these consequences for soybean and beef production and in South America, basically in

Argentina and Brazil (8-16). Byerlee et al. (17) have reviewed many of these studies and

highlighted their main findings and conclusions.

3. Sustainability Certificates

The certified palm oil produced under these schemes should meet certain sustainability

standards including, but not limited to, limiting land use emissions and preserving

biodiversity. The first and most important scheme is the Roundtable on Sustainable Palm

Oil (RSPO) which was initiated in 2004 by the World Wildlife Fund (WWF), the

Malaysian Palm Oil Association, the largest trader of palm oil (Unilever), and several other

non-governmental stakeholders. It comprises a market share of 21% (7). The International

Sustainability and Carbon Certification (ISCC) is a newer scheme which has been

developed in 2010 in response to the EU Renewable Energy Directive and its requirement

and has a market share of 6 – 7% (7). The Indonesian Sustainable Palm Oil (ISPO) and

3

Malaysian Sustainable Palm Oil (MSPO) were also established in 2011 and 213,

respectively. For a complete lists of these types of schemes and their configurations see (7,

17)

4. Debates on imposing a ban on palm oil produced in M&I and its economic and

environmental impacts

The idea of imposing a ban on palm oil was first initiated in the European Union (EU).

Several countries including Denmark, France, UK, Germany, Netherlands, and Norway

signed an agreement in 2015 (Amsterdam Declaration signed on December 2015) to

support consumption of 100% sustainable palm in the EU region and end illegal logging

and deforestation by 2020 (18). Then, in 2016, the ENVI Directorate of the EU

Commission prepared a report on palm oil and deforestation of rainforest and argued that

deforestation causes climate change and that generates social and economic problems (19).

The ENVI report called for a halt to deforestation in rainforests. In 2017, the members of

EU Parliament voted in favor of this report and passed a resolution (EU Resolution of

2017) to support it. Following this resolution several proposals offered to impose a ban on

imports of non-sustainable palm oil into the EU and to stop using food-based vegetable

oils for biofuel production. In January 2018, the EU Parliament approved amendments for

the EU Resendable Energy Directorate II and prohibited use of palm oil in biofuel

production by 2021. In February of 2019, France pledged to halt the importation of

deforestation-related commodities by 2030. More generally, public media, environmental

groups, palm oil producers -- both opponents and supporters of palm oil have expressed

their opinions and arguments against and or in favor of imposing a ban on palm oil. Below

we present several links that represent some of these debates:

4

- France pledges to stop ‘deforestation imports’ by 2030: https://news.mongabay.com/2019/02/france-pledges-to-stop-deforestation-imports-by-2030/

- How palm oil ban has made the EU a dirty world in Malaysia: https://www.euandgvc.nl/documents/publications/2015/december/7/declarations-palm-oil,

- Coalition protests EU’s planned ban of palm oil: https://punchng.com/coalition-protests-eus-planned-ban-of-palm-oil/

- Palm oil: economic and environmental impacts: https://epthinktank.eu/2018/02/19/palm-oil-economic-and-environmental-impacts/

- The EU’s war on palm oil: https://www.aseantoday.com/2018/03/the-eus-war-on-palm-oil/

- The EU Says A New Law Will Expel Palm Oil From Europe. The Evidence Suggests Otherwise: https://www.forbes.com/sites/davekeating/2018/06/15/the-eu-says-a-new-law-will-expel-palm-oil-from-europe-the-evidence-suggests-otherwise/#2d2389a056c1

- Palm Oil Production Poses Problems for the Climate: https://www.climatecentral.org/news/palm-oil-production-climate-18338

- Europe’s palm oil ban has no basis: https://www.eco-business.com/news/europes-palm-oil-ban-has-no-basis-says-sinar-mas-boss-franky-widjaja/

5. GTAP-BIO model and its background

The original GTAP model is documented in Hertel (20) with detailed discussion on the

theory and derivation of the behavioral equations involved in the model. GTAP is a multi-

commodity, multi-regional, Computable General Equilibrium (CGE) model which models

economic activities (including crops, livestock sectors, food, and feed) at a global scale.

Fig. S1 provides the main components of this model. As represented in this figure, a

regional household (e.g., the United States) collects all the income in its region and spends

it over three expenditure types – private household (consumer), government, and savings.

In this model in each sector a representative firm uses primary production factors (labor,

5

land, capital, and resources) and intermediate inputs to produce a final good/service. Firms

pay wages/rental rates to the regional household in return for the employment of land,

labor, capital, and natural resources. Firms sell their output to other firms (intermediate

inputs), private households, government, and investment.

Since this is a global model, firms also export the tradable commodities and import the

intermediate inputs from other regions. These goods are assumed to be differentiated by

region, following the Armington assumption, and so the model can track bilateral trade

flows. Taxes (and subsidies) go as net tax revenues (subsidy expenditures) to the regional

household from private household, government, and the firms. The rest of the world gets

revenues by exporting to private households, firms and government. These revenues are

spent on export taxes and import tariffs, which eventually go to the regional household.

This rest of world composite is actually made up of many other regions – with the same

utility and production functions as for the regional household at the top of Fig. S1.

In this paper we used an advanced version of this model, dubbed GTAP-BIO. This model

has been developed, modified, and used frequently to examine the economic and

environmental impacts of economic-environmental-energy-trade policies (21-31). Fig. S2

shows the main elements of this model. Taheripour, Cui, and Tyner (32) and Taheripour,

Zhao, and Tyner (33) explain the latest version of this model and its background.

Similar to the standard GTAP model, GTAP-BIO traces production, consumption, and

trade of all goods and services (aggregated into various categories) at the global scale.

However, unlike the standard model, GTAP-BIO disaggregates oil crops, vegetable oils,

and meals into several categories including: soybeans, rapeseed, palm oil fruit, other oil

seeds, soy oil, rapeseed oil, palm oil, other oils and fats, soy meal, rapeseed meal, palm

6

kernel meal, and other meals. In addition to the standard commodities and services, this

model handles production and consumption of biofuels (e.g. corn ethanol, sugarcane

ethanol, and biodiesel) and their by-products (DDGS and meals). Therefore, unlike the

standard GTAP model, it takes into account the use of commodity feedstock for food and

fuel and the competition or trade-offs between those and market uses. Hence, this model is

capable of tracing the markets for oil crops, other crops, vegetable oils, and meals produced

in conjecture with vegetable oils. In addition, it traces land use (and changes in land prices)

across the world at the level of Agro-Ecological Zone (AEZ). The latest version of this

model handles intensification in crop production due to technological progress, multi-

cropping, and conversion of unused cropland to crop production as well. Finally, the

parameters of this model were tuned to recent observations. This model traces the links

between crop, livestock, feed, and food sectors and links them with biofuels sectors while

it takes into account forward and backward linkages between these sectors and other

economic activities. This model also takes into account resource constraints and

technological progress. Hence, it provide a comprehensive framework to assess the impacts

of a restriction on palm oil.

The left panel of Fig. S2 shows the main inputs of this model including: 1) Benchmark data

consists of reginal Input-Output tables (known as Social Accounting Matrices (SAMs),

economic parameters, and tax/subsidy rates; 2) International trade data; 3) Land cover,

harvested area, crop production, at Agro-Ecological Zone level; 4) Emission data by

sources and region; 5) Changes in government policy (e.g. changes in taxes, tariffs,

subsidies or changes in biofuel policy); 6) Changes in technology (e.g. changes in total

factor productivity and advances in crop production).

7

The middle panel of Fig. S2 shows that the GTAP-BIO model takes into account the links

between: 1) Forestry, livestock, and crop industries as the main land using sectors; 2) Food,

feed, and biofuels as the user of livestock and crop products; 3) Energy sectors as the main

sources for energy products that interact with other activities in particular with crop and

livestock products; and 4) all other industries and services which interact among

themselves and also with crop, livestock, food, feed and biofuel sectors. As depicted in this

panel, production activities generate intermediate demands for goods and services and

demands for primary inputs including labor, land, capital and resources (see the bottom of

middle panel of Fig. S2). The producers of each region use these inputs and provide goods

and services to be consumed domestically or traded with other regions. As shown in the

right panel of Fig. S2, the GTAP-BIO model provides a wide range of simulation results

including: 1) Changes in production, consumption, and trade of all goods and services and

their prices/costs; 2) Induced Land Use (ILUC) changes; 3) Changes in emissions

associated with production and consumption of good and services by source; and 4)

Changes in economic well-being (welfare).

The GTAP-BIO model traces economic and biophysical variables by region to better

represent real world. For instance, the separation between Brazil and the rest of South

America can better reflect their competition in crop markets. Also the land use emissions

factors of Brazil and the rest of South America are very different. Hence, it is important to

keep these regions separate. However, one can define its own designed geographical

aggregation.

The original GTAP-BIO model was using a two-nest Constant Elasticity of Transformation

(CET) land supply system to govern land allocation among forest, pasture, and forest and

8

also allocate cropland among alternative crops. The land conversion elasticities used in the

original model were determined using a calibration processes developed by Ahmed, Hertel,

and Lubowski (34). These authors determined a global land transformation elasticity to

govern movement of land among forest, pasture, and cropland for a short to medium time

horizon. They also determined a global land transformation elasticity to govern allocation

of cropland across alternative crops. Taheripour and Tyner (26) then extended the land

supply system of the model to a three-nest format. A nest that allocates land among forest

and agricultural land (pasture and cropland), a nest that governs land allocation among

pasture and crop land, and finally a nest that distributes cropland among crops. These

authors divided the whole world into 19 regions and determined a geographical distribution

for the land transformation elasticity applied to each nest of the land supply tree. Due to

the lack of available data for time series estimation of region-specific land transformation

elasticities, these authors used a tuning process based on the recent observations on land

use across the world provided by the Food and Agriculture Organization of the United

Nations to calibrate these elasticities to recent observations.

In general, the GTAP-BIO model uses demand and supply functions and market clearing

conditions to endogenously determine production and consumption of goods/services

(including biofuels) and their prices. In this model demands and supplies are functions of

relative prices and exogenous variables (e.g. consumption taxes or production subsidies).

However, it is possible to determine production and or consumption of goods and services

exogenously as well in the case of binding biofuel mandates. In this case the user can swap

endogenous variables (e.g. consumption of ethanol) with the appropriate exogenous

variables (e.g. a tax credit to blend more ethanol in gasoline) to determine biofuel use

9

exogenously. Following this approach, a module is defined to target

production/consumption of biofuels exogenously when needed (e.g. in the presence of

biofuel mandates). We used this module in the baseline experiment to replicate observed

global biofuel production and use.

6. Description of the examined scenarios

Four simulations were developed for this paper:

- A historical baseline: This case simulates changes in the global economy from 2011

to 2016. This simulation takes into account changes in population, gross national

product, capital formation, labor force, managed land, and biofuel policy by region

according to real observations. The required data for this simulation were obtained from

databases of the World Bank and the Food and Agricultural Organization (FAO) of the

United Nations. We exogenously shocked the model for these variables. In return the

model determined changes in the production, consumption, and trade for all goods and

services and also estimate the rate of technological progress by region. The model

provides changes in harvested areas and land cover items by AEZ and country at the

global scale. For more details on a historical simulation using a static CGE model see

Yao, Hertel, and Taheripour (30). These authors develop a GTAP-BIO baseline for the

time period of 2004 to 2011.

- Experiment I: Baseline is combined with a regulation policy which freezes production

of oil palm in M&I at its 2011 level via a domestic production tax (TAX).

- Experiment II: Baseline plus TAX supported by an economic incentive (subsidy) to

freeze forest area in M&I at its 2011 level (TAXAREA).

10

- Experiment III: Baseline, plus a uniform international tariff imposed on the world

imports of palm oil from M&I which freezes production of palm oil in this region at

its 2011 level, along with the freeze on forest area in M&I (TARIFFAREA).

The required production tax rate, the conservation forest subsidy rate, and the tariff rate for

the introduced policies in this box are all determined endogenously by the model to achieve

the stated policy goal. The results of each experiment were compared with the simulation

results obtained from the baseline simulation to determine the impacts of each policy.

All of these simulations were developed under the RunGTAP program. This is a free and

user-friendly program available to the public. A simplified version of the examined

experiments, their closures, shocks, and results were archived in a zip file which can be

directly loaded into RunGTAP to permit readers to replicate our results, provided they have

a license for the underlying data base, the continuing development of which is supported

by these contributions.

7. Replicating the results using the RunGTAP software interface.

A simple version of this model which closely replicate the main finding of this paper using

a graphical user interface (RunGTAP) will be available for download. The following link

provides information about an online and in-person GTAP short courses:

https://www.gtap.agecon.purdue.edu/events/GTAP101/2019-1/default.asp

The registration costs for the short course are:

$1,400 Developed Country Professional Rate $1,100 Developing Country Professional Rate $750 Developed Country Student Rate $500 Developing Country Student Rate

11

The introductory course -- GTAP 101 -- is fully online and is offered in the spring and fall

semesters.) The first thing the user must do is download the zip archive for this application.

(The link to this archive will be posted online when the paper is published.) With the free

RunGTAP software installed on the computer, users can point to the zip archive, load it

and run the applications through this interface. Key decisions, such as parameter values

and the size of the exogenous shocks are automatically loaded, but can be altered by the

user. Results may be readily viewed from the results page. Fig. S8 offers a screenshot of

the RunGTAP software main page.

One can find the examined experiments under the “solve” button of the main page and then

select each option and run them.

8. Systematic Sensitivity Analysis (SSA)

The SSA program of our model runs a large numbers of simulations to examine the impacts

of systematic changes in the model parameters on the simulation results. Fig. S9 represents

the main components of a typical systematic sensitivity test on GTAP model parameters,

The GTAP-BIO model uses a large number of parameters. However, many of them may

have insignificant impacts on the variables of focus in this study. To concentrate on the

markets and variables that are important for this study we run the SSA test for the following

parameters:

- Trade elasticities (ESUBD and ESUBM) for oil crops and vegetable oils only,

- Substitution elasticities among primary inputs at in value added nest (ESUBVA)

for crops

- Substitution elasticities among vegetable oils in consumption, for household

(VOLH) and firms (VOLF) using vegetable oils,

12

- Land transformation elasticities among land cover items (forest, pasture and

cropland) and between alternative crops,

- Income and demand elasticities (INCPAR and SUPAR) for oilseeds and vegetable

oils.

The SSA varies each parameter over its baseline value by +/- 30%, except for the INCPAR

and SUBPAR that determine demand elasticities. The SSA varies these two parameters by

+/- 25% to maintain constancy in demand elasticities. The sensitivity analysis assumes a

triangular distribution for each parameter and no correlation among the examined

parameters. Using the results, a 95% confidence interval is defined for each variable.

13

Fig. S1. An overview of GTAP Model

14

Fig. S2. Structure of GTAP-BIO model

15

Fig. S3. Production of oil crops (left panels) and vegetable oils and fats (right panels). Panels plot observed outcomes in 2011 and 2016, and the model projections for 2016 obtained from the baseline simulation starting in 2011. Projection errors (% difference between model projected and observed values for 2016) are reported.

16

Fig. S4. Global production of oil crops: Actual observations in 2016, and the model projections for 2016 obtained from the baseline simulation.

17

Fig. S5. Production of oil crops by region: Actual observations in 2016 and the model projections for 2016 obtained from the baseline simulation. RSA represents Rest of South America.

18

Fig. S6. Global harvested area of oil crops: Actual observations in 2016, and the model projections for 2016 obtained from the baseline simulation.

19

Fig. S7. Harvested area of oil crops by region: Actual observations in 2016 and the model projections for 2016 obtained from the baseline simulation. RSA represents Rest of South America.

20

Fig. S8. The RunGTAP Software interface for replicating results

21

Fig. S9. Components of a typical sensitivity test on GTAP model parameters

Researcher task list before conducting a sensitivity test

Select important parameters from the pool of model parameters

Specify the distributions for the selected parameters

Specify type $ size of variations for the selected parameters

Select the Gaussian Quadrature approach

Select resources, commodities, industries, & regions

RunGTAP Sensitivity Test Module

Generate discrete distributions for

selected parameters

Random selections for the generated

distributions

Solve the model for each section

Determine means and standard deviations for endogenous

variables

Extract the means and standard deviations for endogenous variables

Construct confidence intervals and determine error bars for selected

endogenous variables using the extracted means and

standard deviations

Report the results

Researcher task list after getting the solution results

22

Fig. S10. Changes in consumption and production of vegetable oils due to restriction on palm oil expansion under alternative policies. Error bars represent 95% confidence intervals.

23

Table S1. Land use impacts of TAX, TAXAREA, and TARIFFAREA (1000 hectares)

Description TAX TAXAREA TARIFFAREA

M&I Rest Total M&I Rest Total M&I Rest Total

Harvested area

Soybeans 83 471 555 20 511 532 32 731 762 Palm -2,965 -48 -3,013 -3,061 -49 -3,109 -3,029 -206 -3,236 Rapeseed 0 509 509 0 533 533 0 710 710 Other oilseeds 839 1,414 2,253 412 1,477 1,889 34 1,755 1,790 Other crops 1,687 -2,181 -494 132 -2,092 -1,960 373 -2,547 -2,173 Harvested area -356 165 -191 -2,496 381 -2,115 -2,590 443 -2,147

Multiple cropping or unused cropland 0 -94 -94 0 -216 -216 0 -253 -253

Land cover

Cropland -356 70 -286 -2,496 165 -2,331 -2,590 190 -2,400 Forest 363 -56 307 3,096 -618 2,478 3,188 -645 2,543 Pasture -8 -14 -22 -600 452 -147 -598 455 -143

24

Table S2. Decomposition of welfare impacts of TAX, TAXAREA, and TARIFFAREA (million USD)

Description TAX TAXAREA TARIFFAREA

Terms of trade Other Total Terms

of trade Other Total Terms of trade Other Total

USA 483 -68 415 768 -83 685 889 255 1,145 EU27 -640 -258 -898 -665 -343 -1,009 318 -339 -22 Brazil 315 39 354 548 59 607 721 46 768 China -1,036 -2,147 -3,184 -1,248 -2,377 -3,625 116 -3,138 -3,022 R. S. America 382 36 418 552 50 602 502 24 526 M&I 1,547 320 1,867 1,120 -68 1,053 -4,524 -169 -4,693 Other -1,005 -2,268 -3,273 -1,042 -2,803 -3,845 2,004 -4,104 -2,100 World 46 -4,346 -4,300 33 -5,565 -5,532 26 -7,424 -7,398

25

References 1. Agus F, et al. (2013) Historical CO2 emissions from land use and land-use change from

the oil palm industry in Indonesia, Malaysia, and Papua New Guinea, in Reports from the Technical Panels of the 2nd Greenhouse Gas Working Group of the Roundtable on Sustainable Palm Oil, Killeen TJ, Goon J, (eds.) Roundtable on Sustainable Palm oil: pp 65-88.

2. Byerlee D (2014) The fall and rise again of plantations in tropical Asia: History

repeated? Land 3 (3): 547–597.

3. Margono BA, Potapov PV, Turubanova S, Stolle F, Hansen MC (2014) Primary forest cover loss in Indonesia over 2000–2012. Nature Climate Change 4: 730–735.

4. Obidzinski K, Andriani R, Komarudin H, Andrianto A (2012) Environmental and

social impacts of oil palm plantations and their implications for biofuel production in Indonesia. Ecology and Society 17: 481–499.

5. Austin KG, et al. (2017) Shifting patterns of oil palm driven deforestation in Indonesia

and implications for zero-deforestation commitments. Land Use Policy 69: 41-48.

6. Alisjahbana AS, Busch JM. (2017) Forestry, Forest Fires, and Climate Change in Indonesia. Bulletin of Indonesian Economic Studies 53: 111-136.

7. Barthel M, et al. (2018) Study on the environmental impact of palm oil consumption

and on existing sustainability standards. DG Environment, European Commission, European Union.

8. Henders S, Persson UM, Kastner T (2015) Trading forests: Land-use change and carbon emissions embodied in production and exports of forest-risk commodities. Environmental Research Letters 10 (12): 1–13.

9. Brown JC, Koeppe M, Coles B, Price KP (2005) Soybean production and conversion

of tropical forest in the Brazilian Amazon: the case of Vilhena, Rondonia. Ambio J. Hum. Environ. 34: 462–469.

10. Morton DC, et al. (2006) Cropland expansion changes deforestation dynamics in the

southern Brazilian Amazon. Proc. Natl. Acad. Sci. 103: 14637-14641.

11. Richards PD, Walker RT, Arima EY (2014) Spatially complex land change: the indirect effect of Brazil's agricultural sector on land use in Amazonia. Glob. Environ. Change 29: 1–9.

12. Walker R, DeFries R, Vera-Diaz C, Shimabukuro Y, Venturieri A (2009) The expansion of intensive agriculture and ranching in Brazilian Amazonia, in Amazonia and Global Change, Keller M, Gash JCH, Dias PS, Eds. American Geology Union, pp. 61-81.

26

13. Arima EY, Richards P, Walker R, Caldas MM (2011) Statistical confirmation of

indirect land-use change in the Brazilian Amazon. Environmental Research Letters 6: 1–7.

14. Barretto AG, Berndes G, Sparovek G, Wirsenius S (2013) Agricultural intensification in Brazil and its effects on land-use patterns: An analysis of the 1975–2006 period. Global Change Biology 19 (6): 1804–1815.

15. Soares-Filho B, et al. (2014) Cracking Brazil’s forest code. Science 344 (6182): 363–364.

16. Fehlenberg V, et al. (2017) The role of soybean production as an underlying driver of deforestation in the South American Chaco. Glob. Environ. Change 45: 24–34.

17. Byerlee D, Falcon WP, Naylor RL (2017) The tropical oil crop revolution. Oxford

Univ. Press, New York, NY.

18. Amsterdam Declaration (2015) Towards Eliminating Deforestation from Agricultural Commodity Chains with European Countries. Amsterdam, Netherlands.

19. Konecna K (2017) On palm oil and deforestation of rainforests. Committee on the

Environment, Public Health and Food Safety, European Parliament.

20. Hertel T (1997). Global Trade Analysis: Modeling and Applications. Cambridge university press.

21. Hertel T, et al. (2010) Effects of U.S. maize ethanol on global land use and greenhouse

gas emissions: Estimating market-mediated responses. Bioscience, 60(3), 223–231.

22. Taheripour F, Hertel T, Tyner W, Beckman J, Birur D (2010) Biofuels and their by-products: global economic and environmental implications. Biomass Bioenerg. 34: 278–89.

23. Taheripour F, Hertel T, Tyner W (2011) Implications of biofuels mandates for the

global livestock industry: a computable general equilibrium analysis. Agric Econ. 42:325–42.

24. Beckman J., et al. (2012) Structural change in the biofuels era. European Review of

Agricultural Economics, 39(1): p. 137-156. 25. Taheripour F, Hertel T, Liu J (2013) Role of irrigation in determining the global land

use impacts of biofuels. Energy Sustain Soc. 3(4):1–18.

26. Taheripour F, Tyner W (2013) Biofuels and Land Use Change: Applying Recent Evidence to Model Estimates. Applied Sciences, 2013. 3: 14-38.

27

27. Liu J, Hertel T, Farzad T, Zhu T, Rigal C (2014) International trade buffers the impacts

of future irrigation shortfalls. Glob Environ Change, 29: 22–31. 28. Brookes G, Taheripour F, Tyner W (2017) The Contribution of Glyphosate to

Agriculture and Potential Impact of Restrictions on Use at the Global Level. GM Crops & Food 8: 216-228.

29. Taheripour F, Mahaffey H, Tyner W (2016) Evaluation of Economic, Land Use, and

Land Use Emission Impacts of Substituting Non-GMO Crops for GMO in the U.S., AgBioForum 19(2): 156-172.

30. Yao G, Hertel T, Taheripour F. (2018) Economic drivers of telecoupling and terrestrial

carbon fluxes in the global soybean complex. Global Environmental Change, 50; 190–200.

31. Taheripour F, Tyner W (2018) Impacts of Possible Chinese 25% Tariff on U.S.

Soybeans and Other Agricultural Commodities. Choices, 2nd Quarter 33 (2): 1-6

32. Taheripour F, Cui H, Tyner W (2017) An exploration of agricultural land use change at the intensive and extensive margins: Implications for biofuels induced land use change, in Z. Q in, U. Mishra & A. Hastings (eds.), Bioenergy and Land Use Change: American Geophysical Union (Wiley) pp 19-37.

33. Taheripour F, Zhao X, Tyner W (2017) The Impact of Considering Land Intensification

and Updated Data on Biofuels Land Use Change and Emissions Estimates, Biotechnology for Biofuels, 10 (191): 1-16.

34. Ahmed S. Hertel T, Lubowski R. (2008) Calibration of a Land Cover Supply Function

Using Transition Probabilities; GTAP Research Memorandum No 12; Center for Global Trade Analysis: West Lafayette, Indiana, USA, 2008.