The triennial conference of the International Association of Agricultural Economists (IAAE) provides a platform for the Global Futures and Strategic Foresight (GFSF) teams of the CGIAR centers to showcase their work. The first symposium organized by these teams was on ‘Bio-economic modeling to assess options for enhancing food security under climate change in the developing world’ and it took place during the 29th IAAE conference in Brazil in 2012. The teams came again together in 2015 to organize a second symposium on ‘Interpreting results from using bio-economic modeling for global and regional ex ante impact assessment’ at the 30th IAAE conference which took place in Milan on August 8-14, 2015.

1. Global Futures and Strategic Foresight
Analysis at CGIAR:
Application of the geo-spatial bio-economic
modeling framework to inform decision
making
S Nedumaran, G Sika, D Enahoro, Keith Wiebe and Cynthia Bantilan
CGIAR Research Program on Policies, Institutions and Markets
On behalf of the CG centers working on GFSF
Bernardo Creamer, Ulrich Kleinwechter, Guy Hareau, Daniel Mason-D'Croz,
Nelgen Signe, Roberto Telleria
ICAE Conference 9-14 August Milan, Italy

2. Outline
 Global Futures and Strategic Foresight (GFSF)
 Why foresight analysis?
 Framework developed in GFSF
 Case studies undertaken by CG centres
 ICRISAT – Evaluation of Groundnut Promising technologies
 CIP – Evaluation of Potato promising technologies
 CIMMYT - Impact of climate change on production and food security of maize systems in SSA
 ILRI - Quantification of global livestock futures
 Summary and way forward

3. CG centers partners in GFSF

4. Why Foresight Analysis?
Global food economy is in a state of FLUX
Growing population
Rising incomes
Changing diets
Restrictive trade policies
Climate change
Natural Resource degradation
Food crops used for bio-fuel
Higher and more volatile food
prices and increasing food and
nutritional insecurity
Drivers of Change

5. 0
500
1000
1500
2000
2500
3000
1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013
US$perMetricton
Groundnuts Oil Soybeans Maize Sorghum Rice, Thai 5%
Prices for Agricultural Commodities, 1971-2013
Stable and low
Source: World Bank (2014)
Note: Price are in real 2010 US$.

6. Foresight Analysis Reveals!
0
20000
40000
60000
80000
100000
120000
2010 2015 2020 2025 2030 2035 2040 2045 2050
'000metricton
Supply Demand
Demand and supply of grain legumes in Low Income Food Deficit Countries
(LIFDC)*
Source: IMPACT model projection
Note: Grain legumes - groundnuts, chickpea, pigeonpea and soybeans; *There are 62
countries classified under LIFDC by FAO
Projected population(Millions) under poverty in 2050

7. Goal of Global Futures and Strategic
Foresight
Increasing the yield by
developing crop
varieties with promising
traits
Increased production,
reduces the prices,
increase the
consumption
Reduce malnutrition
and Poverty
 To support increases in agricultural productivity
and environmental sustainability by evaluating
promising technologies, investments, and policy
reforms
 Evaluation of selected promising technologies
under development in CG centers
Source: Nelson et al., PNAS (2014)
Modeling climate impacts on agriculture:
Incorporating economic effects

8. Strategic Foresight@ICRISAT
• GFP activities integrated in CRP-PIM
• Multidisciplinary team created and
institutionalized (14 member team)
• Promising technologies were
identified and prioritized for
evaluation
• Collaboration with other CRPs and
Global Projects like AgMIP (data
sharing, model enhancement,
capacity building)
Multi-disciplinary team @ ICRISAT
•Cynthia Bantilan,
Nedumaran, Kai
Mausch, N Jupiter
Economists
•Ashok Kumar, SK
Gupta, CT Hash, PM
Gaur, P Janila, Ganga
Rao, Jana Kholova
Breeders and
Physiologists
•P Singh
•Dakshina Murthy
•Gumma Murali Krishna
Crop
Modelers/GIS/RS
ImpactAssessment

9. Integrated Model Framework
Hydrology
model-WSM
Climate
model
Source: Rosegrant et al. (2012)
Crop model

10. Evaluation of Promising Technologies: Virtual
Cultivars
 Target of the crop improvement scientists – develop promising
technologies with higher yield
PotentialYield(Kg/ha)
2014 2020
Incorporating the traits in elite cultivars
better root system
Extractmorewater

11. Crop Model Calibration and Development
of Virtual Promising Cultivars
 DSSAT Crop Model
 Baseline Cultivars selected - JL 24, M 335 and 55-437
 Location
 Anantapur and Junagadh sites in India
 Samanko (Mali) and Sadore (Niger) sites in West Africa
 Calibrate and validate baseline cultivars
 Manipulated the genetic co –efficient of baseline
cultivars and developed the virtual promising cultivars
for each locations
 Drought Tolerant
 Heat Tolerant
 Drought + Heat + yield Potential
 Estimate the yield change for each technology compare
to baseline cultivars in each location
 Data source: Breeders yield trial
data; NARS trial data
 National Bureau of Soil Survey and
Land Use Planning, Nagpur India;
WISE soil database
 India Meteorological Department
(IMD); NASA website
(http://power.larc.nasa.gov/)

12. 1171
1225
1270
1477
1100
1150
1200
1250
1300
1350
1400
1450
1500
JL 24 Drought
Tolerance
Heat Tolerance Drought + Heat
tolerance +
Yield potential
Kg/ha
1228
1271
1246
1451
1100
1150
1200
1250
1300
1350
1400
1450
1500
JL 24 Drought
Tolerance
Heat Tolerance Drought + Heat
tolerance + Yield
potential
Kg/ha
Step 1: Simulated Yield of Promising
Technologies of Groundnuts
Source: Singh, P., Nedumaran, S., Ntare, B.R., Boote, K.J., Singh, N.P., Srinivas, K., and Bantilan, M.C.S. 2013. Potential
benefits of drought and heat tolerance in groundnut for adaptation to climate change in India and West Africa. Mitigation and
Adaptation Strategies for Global Change.
18%
Yield under Current Climate
26%
Yield under Climate change 2050
(HADCM3, A1B scenario)
Location: Anantapur, India; Baseline Cultivars: JL 24

13. Step 2: Spatial Change in Groundnut Yield
Baseline cultivar (Current Vs Future Climate)
Promising Technology – Drought Tolerant

14. Step 3: Technology Development and Adoption
Pathway Framework
2012 2018 2035
India60%
Nigeria40%
20372020
Research lag Adoption lag
Promising Technologies of
Groundnut development
Technology development
Technology dissemination and
adoption
Outcomes and
Impacts
• Change in
Production
• Change in prices
• Change in
consumption
• Poverty level
Nedumaran et al. (2013)
Target countries
Target countries
Production
share (%)
Burkina Faso 1.2
Ghana 1.62
India 12.99
Malawi 0.7
Mali 1.01
Myanmar 3.84
Niger 0.51
Nigeria 10.72
Tanzania 1.73
Uganda 0.76
Vietnam 1.84
Total 36.92

15. Potential Welfare Benefits and IRR (M US$)
Technologies
Net Benefits
(M US$)
IRR (%)
Heat Tolerant 302.39 30
Drought Tolerant 784.08 38
Heat + Drought + Yield
Potential
1519.76 42
Nedumaran et al. (2013)
0
50
100
150
200
250
300
350
Malawi Tanzania Uganda Burkina
Faso
Ghana Mali Nigeria Niger India Myanmar Vietnam
ESA WCA SSEA
MUS$
Heat Tolerance Drought Tolerance Heat+Drought+yield potential

16. Evaluation of improved potato
varieties for SSA
• Key traits
• Higher yield potential
• Late-blight and virus resistance
• Heat tolerance
• Processing quality
• 30% higher yields
• Nine target countries
• Total investment: 9.8m US$ (4.29m NPV, 2000
constant prices)
• Project duration: 12 years
Source: Theisen and Thiele (2008).
EthiopiaUganda
Rwanda
Burundi
DR Congo
Kenya
Tanzania
Mozambique
Malawi

17. Welfare Benefits and IRR
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
Net welfare changes (M $)
Low adoption Medium adoption High adoption
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
IRR
Low adoption Medium adoption High adoption
 Positive production impacts in target countries
 Positive net welfare effects and high ROI in target countries
 Comparable with findings from previous impact evaluations of improved varieties
 Investment in improved potato varieties justified from economic point of view

18. Strategic Foresight @ CIMMYT:
Impact of Climate change on Maize
Production in SSA

19. GCM monthly
gridded data
Regional/global crop productivity
under various climate models and
technologies
Evaluated
DSSAT
model
DSSAT
Crop
Model
Site/farm level
simulation
Site
soil
Daily
site
climate
Crop
Crop
management
Model
calibration
Model
evaluation
27 FAO soil
groups
daily pixel
climate
Crop
management
Weather
generator
Crop per
MME
Evaluated DSSAT
model
Evaluated
IMPACT model
GIS
GIS
Projections on
population
and income
growth
Trade-offs
(elasticities)
on inputs,
production
and
consumption
patterns
Projections on
trade barriers
Projected world and
domestic prices
Projected
demand,
supply and net
trade
Nutrition
results
DSSAT Spatial DSSAT IMPACT
Bio-economic modelling framework

20. Changes in yield and area of maize under low N level in SSA by 2050 (a & b) and 2080 (c & d)
relative to the baseline (2000) using climate projection from CSIRO and MIROC global circulation
models under the A1B emission scenario
Impact of Alternate Climate Scenarios on Maize

21. SSA
Eastern SSA
Southern SSA
Central SSA
Western SSA
0
10
20
30
40
CSI-A1 vs. Base2050
MIR-A1 vs. Base2050
Changein#ofpeople
atriskofhunger(mil.)
Caloric intake in 2050 under MIR-A1 scenario
1500 2000 2500 3000 3500 4000
Changeinpeopleatriskofhunger(mil.)
-2
0
2
4
6
8
10
BUR
DJIERI
ETH
KEN
MAD
MLW
MOZ
RWA
SOM
TAN
UGA
ZAM
ZIM
ANG
CAM
CAR
CHA CON
DRC
EQG GAB
BEN
BUF
GAM
GHA
GUIGUBIVC
LIB
MAL
MAU
Niger
Nigeria
SEN
SLETOG
Effect of Climate Change on Food
Security in SSA

22. Summary
Maize production in SSA
 Reduction of up to 12% and 20% by 2050 and 2080,
respectively
 Sahel and southern Africa: reduction in maize yields due
to increasing temperatures and decreasing rainfall
 Highlands in eastern Africa: increase in maize production
due to small changes in rainfall and increasing
temperature
 Food security in SSA: hardest-hit is eastern Africa; DRC in central
Africa and Nigeria in western Africa
Tesfaye et al., 2015

23. Strategic Foresight @ ILRI
Global Livestock Futures

24. Global Livestock Futures
Objective: Improve representation of livestock sector in
IMPACT model to:
Better account for (agro-ecological and
management system) barriers to sector growth
Better assess potential for sector expansion
Improve capacity to simulate response, growth and
recovery to shocks - including climate change
Enhance model usefulness as policy assessment tool
for livestock sector development

25. Original Specification Suggested Updates
Supply response is relatively homogenous
within countries
Livestock supply disaggregated by system
types (intensive/extensive)
Livestock feed basket composed only of
internationally-traded feeds (mostly coarse
grains and meals)
Pasture grasses, crop residues and occasional
feeds added to livestock feeding possibilities
Yield is exogenously determined, and does
not respond to quantity or quality of fed
rations
Meat and milk yield response functions are
endogenous, responding to changes in feed
quantities and nutritive values
Total herd size includes milk-producing and
slaughtered meat animals only
Total herd count includes replacement and/or
follower herds in dairy and meat production
Animal productivity only indirectly affected
but not affected by feed availability through
price effects
Explicit feed-availability constraints imposed
on animal productivity
Source: Msangi et al., 2014
Suggested Enhancements to Livestock Sector
representation in IMPACT

26. Source: Msangi et al., 2014
Baseline Projections of Meat Production to
2030 for Key Countries
Baseline Projections of Milk Production to
2030 for Key Countries
Baseline Results

27. Summary
 More dynamic growth for meat and lamb production
in China, milk in India; Brazil meat production to
surpass US by 2030
 Supply-side response to growing demand for livestock
products is more constrained in the enhanced model
 Growth in feed demand and pressure on land
resources more apparent, with important implications
for the more extensive production systems

28. Way forward
 Evaluate the additional promising technologies (biotic stress tolerant and
management options) with current GF/PIM Strategic foresight tool
 Provide evidence to better targeting of technology and inform priority
setting for CG centres and CRPs
 Identify and collaborate with pest and diseases modelling team
 Consider linking results from global models with household data for ex ante
impact assessment at lower scales
 Gender lens in foresight analysis and technology evaluation
 Development of ‘stand alone’ module in IMPACT with enhanced
representation of livestock
 Test current and alternative (technology and policy) strategies for livestock
sector development under a range of plausible future scenarios - including
global climate change

29. Thank you!