Introducing Biodiversity issues in a Global Computable General

Introducing Biodiversity issues in a Global Computable General Equilibrium Model: the MIRAGE example DR. DAVID LABORDE D. LABORDE@CGIAR. ORG, IFPRI BASED ON JOINT WORKS WITH : Anouch Missirian, IFPRI-Columbia University for biodiversity focus Dr. Lauren Deason, IFPRI for nutrition focus Prof. Antoine Bouet, IFPRI-University of Bordeaux for Household modeling focus

Introducing biodiversity in a global CGE Multi dimensional issues (additivity, separability) Different scales Direct and indirect effects Conceptually: still some challenges ◦ ◦ ◦ ◦ Biodiversity as a Stock and factor of production Ecosystem services as a flow Role of time frame Role of scale: average vs concentration, average vs marginal Non linearity, irreversibility Moving across scale: No Bijection! Dynamic issues and discount factor! Quantification: huge challenges, especially when tackling uncertainties properly

Crop genetics, evolution, traits Crop varieties, diversification Biocontrol / Pollination Biodiversity in Food Loop effects Biodiversity Long term yield growth, adaptation capacity Higher resilience, stress response Reduced input uses, “higher” yield Better nutrition Ecosystem services beyond agriculture Total Factor Productivity (TFP) Lower yield/TFP volatility, joint production pattern, constraint e. g. on mechanization TFP, land productivity and input uses, constraint e. g. on human capital Nutrition metrics, Labor productivity Efficiency of the water system, Productivity of tourism, Health Land Use changes - Production Pattern Inpiut Uses - (flows, accumulation) Conceptual Framework: A simplified view on [Ag. ] biodiversity from a CGE modeler

MIRAGE-CGE Framework KEY ELEMENTS Global dynamic CGE, multi sectoral. Baseyear 2012 Main data sources: ◦ GTAP but a lot of modifications, fixes and updates ◦ Reconstruction of agricultural accounts using FAO, EUROSTAT, USDA, ◦ ◦ Bloomberg (consistency between physical and monetary flows) Trade policy instruments (specific sources) Foreign Direct Investments Household surveys …. Up to 87 sectors/products and 130 regions. However, most simulations focus on a 35 x 40 configuration. Different versions ◦ MIRAGE-CC: Long term, 2050, climate change focus ◦ MIRAGE-HH: Medium term, 2030, Focus on household heterogeneity. Full bottom up HH modelling. ◦ MIRAGE-Biof: Medium term, 2030, land use focus, different biomass-to-energy pathways, higher crop disagregation (e. g. 5 oilseed crops) FRAMEWORK Remark: CGE are bottom-up models. Should not confuse the level of aggregation and how entities interact in a model. Structural model, no reduced forms: ◦ Production function with a inputs and factors ◦ Utility function driven demand function: true welfare analysis Prices clear markets for each product, from each origin Product differentiation Factors of production (2 to 10 types of labor, capital, natural resources, land) Private and Public income (government finance) Recursive dynamic

Model focus: Land Markets – at the AEZ Level In average 200 -300 production unit by sector Infra-country modelling to capture land heterogeneity PAGE 6

Illustration MIRAGE-CC: Global Land Use Change by 2050: Cropland expansion. +2. 4 Mio Km 2 (+20. 3%). Pasture will be reduced by 2. 07 Mio Km 2 (still deforestation) Africa America Asia CIS EU 27 LAC Row Illustration. MIRAGE-CC by 2050. Central Scenario. Laborde 2014

Illustration MIRAGE-Biof: Differentiated Cropland increase of an increase of the EU biofuel demand, by region, by trade policy regime Laborde 2012

Illustration MIRAGE-Biof: More cropland or More Carbon (Ha by TJ and Tons CO 2 eq by Ha of cropland), by feedstock Note : The bars (left y-axis) show the amount of additional net cropland by TJ of biofuel produced for one feedstock. The line (right x-axis) shows the average tons of CO 2 equivalent by net Ha of cropland. Laborde 2012

Illustration MIRAGE-HH: Assessment of policy reforms or external shocks at the country and HH level. Illustration: global trade liberalization 4 One bubble = one household category Bubble size = Number of people in this household category 5 2 % Increase in Real Income 4 3 2 1 0 -1 0 2 4 6 -2 -4 -6 0 -2 2 4 6 Initial level of Income of the household Tanzania 8 10 -8 Initial level of Income of the household Brazil 8 10 12

MIRAGE Model outputs §Agricultural production §land use §input uses (chemicals, manure) §agricultural and non agricultural prices §Factor prices (wages, different skills, urban/rural, formal/informal, male/female) §Capital accumulation §income effects, including §Employment §Food consumption §At this stage, no feedback effect from biodiversit

Land Use Biodiversity. Top down approach 1. A generic approach to link land use and biodiversity measured as Species richness Different habitats (model outputs), at an AEZ level 2. A site specific approach: linking CGE output with a local land use model. Spatial econometric model estimated on detailed grid-information : ◦ for land use ◦ for bird population (STOC data, biodiversity metrics)

Focus: The role of a detailed bilateral trade database on nutrition contents. 800 products, 220 trading partners

Focus: Food Diversity and concentration Looking at imports for human consumption only. Average Number of Products # Products Herfindahl-Hirshman Index Product Space # Exporters Proteins Calories 1998 -2000 2011 -2013 Afghanistan 97 397 1. 5 3. 7 0. 731 0. 428 0. 516 0. 263 Argentina 514 429 5. 3 4. 8 0. 402 0. 064 0. 201 0. 067 Australia 548 546 10. 4 15. 7 0. 129 0. 041 0. 054 0. 021 Brazil 540 502 6. 4 7. 4 0. 338 0. 309 0. 341 0. 262 China 575 558 9. 4 14. 5 0. 371 0. 733 0. 132 0. 270 Ghana 310 491 3. 0 7. 1 0. 234 0. 088 0. 187 0. 093 Guatemala 491 495 3. 9 4. 7 0. 173 0. 139 0. 134 0. 132 Malawi 221 359 1. 7 2. 2 0. 207 0. 268 0. 160 0. 180 Mali 250 309 3. 0 3. 7 0. 137 0. 244 0. 146 0. 187 Paraguay 379 369 3. 0 3. 7 0. 209 0. 158 0. 135 0. 056 United States of America 601 585 20. 9 24. 9 0. 035 0. 030 0. 026 0. 024 Uzbekistan 230 299 2. 5 3. 2 0. 498 0. 364 0. 370 0. 263