Data Needs for Agricultural R&D Evaluation Liangzhi You Training Workshop on Impact Assessment of Agricultural Research in Tanzania
Parameters in R&D Evaluation Scenario Constants Market regions Base year Simulation period Real discount rate Market Information Initial price
Parameters in R&D Evaluation Market Information Supply Initial quantity Elasticity Exogenous growth rate Tax/subsidy Demand R&D Parameters
Market Region: Spatial Aggregation Nature of the evaluation Data availability Reduced data needs Easy to observe and interpret R&D effects Increased data needs Important markets and production zones Homogeneous Fewer regions Disaggregated More Regions
Many Disaggregation Opportunities Geopolitical regions Income groups Agro-ecological zones Adoption domains Production systems Processing/marketing chain
Use GIS Overlay to Define Spatial Analysis Unit + Slow Adoption Rapid Humid Zone Dry Zone Region B A SPECIFIC AGROECOLOGICAL ZONES MARKET REGIONS ADOPTION DOMAINS PRODUCTION AREAS 1 2 3 5 4 R GIS OVERLAY Homogeneous units with respect to the impact of new technology
Spatial and Socioeconomic Definition of regions
EXAMPLE: Egypt Rice
EXAMPLE: Uganda (Agroecological)
Discount Rate: Capital budgeting Scenario Constants Base Year: Benchmark Most recent 2~3 years (e.g. 1999-2000) Simulation Period Normally 10 ~ 30 years Discount Rate: Capital budgeting Nominal vs. real discount rate Real discount rate 3~5%
Initial Price Price Market level: farmgate, wholesale, border Border price CIF (cost, insurance and freight) FOB (Free-on-board) Exchange rate Data sources Market surveys FAOSTAT Different meaning price level
Supply and Demand Quantities Production Three-year average ROW production for multihorizontal market Consumption Market equilibrium over all regions Same form as production, price Data Sources National statistics FAO Production/consumption surveys Consumption = production + Import – Export – Stock change
Exogenous Growth Demand Growth w/o research Population growth Income growth Demand Growth=Pop. Growth + Income elasticity *Income Growth Production Growth w/o research Area expansion Yield growth not attributed to research Data sources UN Population Division World Bank: Word Development Indicators FAO National statistics
Elasticity Estimating elasticity needs time and resources Borrow elasticity from other situations but be careful Long-run vs. short-run elasticity Single commodity vs. aggregate of commodities Dynamic response Single commodity model
Elasticity (cont.) Supply Elasticity Demand Elasticity Big effect of the factors’ supply elasticities Agricultural product: 0.1~1.0 Demand Elasticity Very small demand elasticity for staples Higher elasticity for luxury foods Majority of food: 0 ~ 1.0
Income Elasticity Income Elasticity Reflect consumption pattern From household/consumption survey Income elasticity of multiple uses Si is the share of total quantity used for product i, and etaiI is income elasticity for product I. It means that if we have different uses of commodity j, the income elasticity is equal to the shared weighted sum of the income elasticities of demand for the final products. E.G.Maize: Feed vs Food.
Income Elasticity (cont.) Data Sources Isabelle Tsakok (1990), Agricultural Price Policy H. Askari and J. Cummings(1976), Agricultural Supply Response: A Survey of Econometric Evidence Estimation from household surveys
Income Elasticities in Egypt Commodities Urban Rural Egypt Fino bread 0.22 0.09 0.13 rice 0.27 0.24 Maize flour -- 0.03 0.02 Other cereal 0.31 0.38 0.37 Cooking oil 0.21 0.26 0.23 Sugar 0.52 0.61 0.56 Vegetables 0.59 0.68 0.63 Meat 0.81 0.73 Eggs& milk 0.66 0.62 Source:H. Bouis et al (1999), Patterns of food consumption and nutrition in Egypt, IFPRI
Demand Elasticities in Egypt Commodities Urban Rural Egypt Fino bread -0.85 -0.90 rice -0.86 -0.87 Maize flour -- -0.88 Other cereal -0.89 -0.91 Cooking oil Sugar -0.96 -0.98 Vegetables -0.97 Meat -1.00 -1.01 Eggs& milk -0.99 Source:H. Bouis et al (1999), Patterns of food consumption and nutrition in Egypt, IFPRI
Government Policies Many types tax/subsidy on production, inputs or trade Price fixing: target (minimum) price, ceiling (maximum) price Quantitative restrictions: Quotas Main findings (Alston and Martin, 1992 ) Patterns of research benefits changed relative to free trade The changes depend on specific interventions The gain or loss of total welfare of research is equal to the loss or gain of social cost of the market intervention Tax/subsidy can be treated as price distortion. They are included in DREAM.
Sensitivity Analysis for Egypt Rice Baseline Data Simulations period: 20 years Discount rate: 3% Supply Shift: 1% Region Supply Demand Price Supply Elasticity Demand Elasticity (1000 tons) US$/ton Egypt 3,277 2,520 240 1.0 0.5 ROW 50,908 51,665 210
Sensitivity Analysis for Market Parameters Elasticity (demand, supply or transmission) affect benefits distribution, not total benefits. Egypt is an exporter, so the impact of supply (quanity, growth) is felt on consumer benefits too. If an importer, demand will affect producer benefits.
Example Structure of simulations ASARECA example I cited before.
My Input Data Format
Market Data and Analysis Analysis is only as accurate as the underlying data Analyst’s own judgment and experience is important Address Sensitivity analysis Be careful of interpreting the results and any policy implications
Key Points for Market Data Deciding analysis regions is fundamental Consistency of commodity form for price, supply and demand Implication of different levels of prices at market chain Importance of elasticities Guides of Sensitivity analysis
Overview of Data Needs by Region Type