Group presentationsApplications of the GTAP-AEZ model
23th Annual Short Course in Global Trade Analysis, July 11-17, 2015
GTAP-AEZ Model• Standard GTAP model (Hertel 1997) plus key
elements for the analysis of competition for land and related issues:– Heterogeneous land endowment
• Agro-Ecological Zones (AEZ)– Land supply to land using activities within AEZ– Changes in crop yields
• Intensive margin• Extensive margin
– Emissions from land use change
2
Heterogeneous Land: AEZs• 18 Agro-Ecological Zones
– 6 growing periods (6 categories x 60 day intervals)– 3 climatic zones (tropical, temperate and boreal)
• The competition for land within a given AEZ across uses is constrained to include activities that have been observed to take place in that AEZ
3
Global Distribution of AEZs
5
Land Supply in the GTAP-AEZ Model
Application 1
23th Annual Short Course in Global Trade Analysis, July 11-17, 2015
Experiment motivation
• Climate change is expected to decrease yields (on average) world wide– In the absence of tech developments, what does this
infer for land use change?• Does trade liberalization provide a
‘countervailing’ force – What does it mean for de/re-forestation?
7
How was this implemented?Two steps• Yield changes explored as:
– What if today’s crops grew in 2025’s climate?… infers a lower production per ha for a given set of inputs … shock to productivity terms aoall()
• Estimates were taken from IFPRI differentiated by developed and developing countries– Oilseed and wheat effects in developing less than developed,
opposite rice and maize (other grains)• What about trade?
– Tariffs on agro-food products set to zero worldwide(tms)– Note: tariff levels are high relative to climate shocks
8
Starting at the end
Land use change (000s ha)Emissions from land cover change (CO2 tons)
-2,000
-1,000
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
Forestry Cropland Grazing Total
Climate Trade
-10,000
-5,000
0
5,000
10,000
15,000
Forestry Cropland Grazing
Climate Trade
Closer look at Europe and the America’sClimate effects• Demand inelastic for staples
• Prices for good rise – across agrofood, but fall for services and mnfcs• … But change in quantity produced is generally smaller than the fall in
productivity• Result due to shock to worldwide productivity• Land shifts out of forestry to agriculture as a result
• With trade…• Reduction in tariffs sees America’s production increase, and a fall in Europe
(existing high tariffs)• Brazil has a large increase in grazing production driven by domestic
demand (from food sector which is exporter)• Agriculture sector expands increasing deforestation
Europe and the America’s: land and emissionsLand use change (000s): deforestation
Emissions from land cover change (CO2 tons): increase!
-12000
-10000
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
Forestry Cropland Grazing
Brazil EU USA Brazil EU USA
-2000
-1000
0
1000
2000
3000
4000
5000
Forestry Cropland Grazing
Brazil EU USA Brazil EU USA
The key drivers: Brazil re-enforcing influencesBrazil• Relative effects of climate effects on yield play to its
strengths• Oilseed and wheat sectors expand
• Driven by additional exports at the expense of other developed country producers who face larger yield losses
• Trade liberalization encourages trade growth for agro-food sector overall• Large increase in livestock sector, plus expansion of crops• Drawing land out of forestry into agriculture
The key drivers: USA & EU – a tale of competing effectsUSA• Climate induced effects
• Initial fall in crop production – relatively worse off due to productivity effects• Land shifts to forest sector
• Trade induced effects• Expansion in agriculture to increase world demand for US produce• Mainly wheat, maize, rice but also livestock sector• Land shifts out of forestry
EU-27• Climate effects see push towards crops and deforestation• Trade liberalization works the other way – output in forestry expands and ag production falls
due to removal of relatively large tariffs• … But a shift in location forestry area expands greater than output change (%)
Factor prices with climate effects: an illustration of scarcity [ps()]Land (selected zones) Other value added
0
50
100
150
200
250
300
3 AEZ3 8 AEZ8 12 AEZ12 17 AEZ17
1 BRA 4 EU27 7 USA
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
19 labor 20 capital 21 NatRes
1 BRA 4 EU27 7 USA
Land Use and Emissions Change in China
15
-5000.00
-4000.00
-3000.00
-2000.00
-1000.00
0.00
1000.00
2000.00
3000.00
4000.00
frs cropland grazeliv
Land Use Change in China (1000 ha)
CHN (climate) CHN (climate+tradelib) CHN (tradelib)
-1000.00
-800.00
-600.00
-400.00
-200.00
0.00
200.00
400.00
600.00
800.00
1000.00
frs cropland grazeliv Total
Emissions Change in China (tCO2)
CHN (climate) CHN (climate+tradelib) CHN (tradelib)
Land Use and Emissions Change in East Asia
16
-60.00
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
100.00
frs cropland grazeliv Total
Emissions Change in East Asia (tCO2)
xea (climate) xea (climate+tradelib) xea (tradelib)
-800.00
-600.00
-400.00
-200.00
0.00
200.00
400.00
600.00
frs cropland grazeliv
Land Use Change in East Asia (1000 ha)
xea (climate) xea (climate+tradelib) xea (tradelib)
China and East Asia• Climate Change
– Helps reforestration and increase the use of cropland however, causes to decrease the use of pasture
– Correspondingly, CO2 emissions decreases due to climate change
• Trade Liberalization on AGROFOOD – Dominates the productivity impacts by climate change– Leads to further reforesration both in China and East Asia– However, causes to increase in cropland in China yet decrease in East Asia.– On the other hand, helps expand pasture in East Asia and reduce pasture in
China
• Explanation on China’s Forestry Expansion – qo(forestry, China)= -0.8%– qfeland(forestry, China)= -0.8%– p_LANDCOVER= 2%
17
Changes in Land Use Patterns and CO2 Emissions in South Asia
18
-2500
-2000
-1500
-1000
-500
0
500
1000
1500
Forest Cropland Pasture
Chan
ges i
n La
nd C
over
100
0 ha
Changes in Land Use Pattern
Climate Change
Trade Liberalization
Climate change and Trade Liberalization
-400
-300
-200
-100
0
100
200
300
Forest Cropland Pasture
mill
CO2
Changes in CO2 Emissions in South Asia
Climate Change
Trade Liberalization
Climate change and Trade Liberalization
Changes in Land Use Patterns and CO2 Emissions in India
19
-2500
-2000
-1500
-1000
-500
0
500
1000
1500
Forest Cropland Pasture
Land
Cov
er C
hang
e (1
000
ha)
Changes in Land Use Pattern
Climate Change
Trade Liberalization
Climate change and Trade Liberalization
-300
-250
-200
-150
-100
-50
0
50
100
150
200
Forest Cropland Pasture
mill
t CO
2
Changes in CO2 Emissions
Climate Change
Trade Liberalization
Climate change and Trade Liberalization
South Asia – Land Use Pattern• Climate change caused
– Deforestation– Decrease in pasture lands– More land allocation for crop production
• Trade Liberalization in Agri-food industry caused– More land allocation for forestry and pasture and reduced land
allocation for crop production.• Net Effect
– Dominates by India– Increased land use for forestry and pasture– Reduced land use for cropland– Dominates by trade liberalization effect in the agri-food sector
20
Major driving factors of the Net Effect
• Impact on output (qo) of agri-food sector– Total = -4%– Vegetable oils and fats = -44.7– Oil seeds (shocked) = -17.5– Livestock production increases but very small (0.4)
• Supply price in agri-food sector (ps)– Total = -3%– Oil seeds (shocked) = -7.1– Price increased in paddy rice (2.2)
21
A note on forestry
• Output of forestry has increased (1.55)– Forestry area is expanding– Increase in output is driven by both increases in domestic and
export demands.
22
South Asia – Changes in Emissions
• Deforestation due to climate change has increased CO2 emissions.
• Net Effect is dominated by trade liberalization in Agri-Food sector.
23
Application 2
23th Annual Short Course in Global Trade Analysis, July 11-17, 2015
Effects on Land UseBased on improvements in technology in soybean production in Brazil
Experiment design Scenario 1: shock aoall("osd","BRA") = 6.38;
Technological improvement in soybean production
Yearly tech/yield improvement improvement of 0.66% (Ludena, 2006)
Reduced final shock value to 85% based on ratio of soybeans in the “osd” aggregation
Scenario 2: S1 + improved conversion ratio of forest/grazing land to cropland. 0.66 Universal 2004 ratio updated to more recent and more accurate ratios.
Provide comparison of results of changes in land use
Impact of TFP improvement in Brazil and World markets
Soy Beans (S1)
Vegetable Oils (S1)
Brazil Output 10.23 5.31
Brazil Supply Price
-5.36 -2.20
Brazil Exports 14.6 11.7
World Price -0.93 -0.40
World Output -0.081 0.053
*Brazil accounts for 24% of all world exports of Soybeans
0
5
10
15
Soy Beans Vegetable Oils
Brazil Output Destinations
Domestic Exports
Changes in land cover (1,000 ha)S1 Land cover type
Brazil China EU27 RoW Global
Forests -130.8(-0.08%)
-31.2(-0.02%)
52.6(0.03%)
71.8(0.004%)
-37.7(-0.002%)
Cropland 281.2(0.46%)
-13.0(-0.02%)
-75.8(-0.06%)
-488.8(-0.032%)
-296.3(-0.019%)
Pasture -150.4(-0.09%)
44.2(0.03%)
23.2(0.04%)
417.0(0.015%)
334.0(0.012%)
S2 Land cover type
Brazil China EU27 RoW Global
Forests -102.8(-0.07%)
-31.2(-0.02%)
52.6(0.03%)
71.8(0.004%)
-9.6(-0.001%)
Cropland 202.0(0.33%)
-13.0(-0.02%)
-75.8(-0.06%)
-488.8(-0.032%)
-375.5(-0.024%)
Pasture -99.2(-0.06%)
44.2(0.03%)
23.2(0.04%)
417.0(0.015%)
385.2(0.015%)
Impact of TFP improvement in Brazil on China
Soy Beans (S1)
Vegetable Oils (S1)
China Output -0.92 -0.07
China Supply Price
-0.36 -0.14
China Exports -3.36 -0.81-1
-0.8
-0.6
-0.4
-0.2
0
Soy Beans Vegetable Oils
China Output Destinations
Domestic Exports
Graphical comparison (1000 ha)-1
30.8
71.8
-37.
7
281.
2
-488
.8
-296
.3-15
0.4
417
334
BRAZIL ROW GLOBAL
LAND COVERAGE CHANGE (S1)
Forests Cropland Pasture
-102
.8
71.8
-9.6
202
-488
.8 -375
.5
-99.
2
417
385.
2
BRAZIL ROW GLOBAL
LAND COVERAGE CHANGE (S2)
Forests Cropland Pasture
Land rent change (%)S1 Item Brazil China EU27
Forestry 2.67 -0.36 -0.87
Oil seed prod. 7.97 -1.21 -3.14
Pasture 2.78 -0.24 -0.88
2.67
-0.36 -0.87
7.97
-1.21-3.14
2.78
-0.24 -0.88-5
0
5
10
Brazil China EU27
Land Rent Change
Forestry Oil seed prod. Pasture
Emissions change from land cover change (Million Co2 tons)
S1 Land cover type
Brazil China EU27 RoW Global
Forests 50.8 17.9 -18.9 -20.1 29.7Cropland -5.1 0.2 1.4 8.8 5.3Pasture 10.8 -8.7 -3.6 -42.1 -43.6Total 56.5 9.5 -21.2 -53.5 -8.6
S2 Land cover type
Brazil China EU27 RoW Global
Forests 39.9 17.9 -18.9 -20.1 18.8Cropland -3.6 0.2 1.4 8.8 6.8Pasture 7.1 -8.7 -3.6 -42.1 -47.3Total 43.4 9.5 -21.2 -53.5 -21.7
Graphical comparison (mill tonnes)50
.8
-20.
1
29.7
-5.1
8.8
5.310
.8
-42.
1
-43.
6
56.5
-53.
5
-8.6
BRAZIL ROW GLOBAL
CHANGE IN CO2 EMISSIONS (S1)
Forests Cropland Pasture Total
39.9
-20.
1
18.8
-3.6
8.8
6.8
7.1
-42.
1
-47.
3
43.4
-53.
5
-21.
7
BRAZIL ROW GLOBAL
CHANGE IN CO2 EMISSIONS (S2)
Forests Cropland Pasture Total
Final thoughts
23th Annual Short Course in Global Trade Analysis, July 11-17, 2015
Concluding comments
• Deforestation has a regional element• After liberalization, we see reforestation in East and South Asia, but further
deforestation in the Americas • Points to significant contribution of technical advancement
in reducing future deforestation• And adapting to climate change
• One market impacts are not sufficient to cope with carbon reduction…need global efforts
• Cost effective ways for land conversion would help reduce carbon emissions related to cropland expansion
Group presentationsGTAP-AEZ ModelHeterogeneous Land: AEZsGlobal Distribution of AEZsSlide Number 5Application 1Experiment motivationHow was this implemented?Starting at the endCloser look at Europe and the America’sEurope and the America’s: land and emissionsThe key drivers: Brazil re-enforcing influencesThe key drivers: USA & EU – a tale of competing effectsFactor prices with climate effects: an illustration of scarcity [ps()]Land Use and Emissions Change in ChinaLand Use and Emissions Change in East AsiaChina and East AsiaChanges in Land Use Patterns and CO2 Emissions in South AsiaChanges in Land Use Patterns and CO2 Emissions in IndiaSouth Asia – Land Use PatternMajor driving factors of the Net EffectA note on forestrySouth Asia – Changes in EmissionsApplication 2Effects on Land UseExperiment designImpact of TFP improvement in Brazil and World markets�Changes in land cover (1,000 ha)Impact of TFP improvement in Brazil on China�Graphical comparison (1000 ha)Land rent change (%)Emissions change from land cover change (Million Co2 tons)Graphical comparison (mill tonnes)Final thoughtsConcluding comments