Ex Ante Evaluation of Investment in Research and Development for Major Food Commodities: Case Study of the Philippines Using the Welfare Impact Simulator for Evaluating Research (WISER) Model Roehlano M. Briones Research Fellow, Philippine Institute for Development Studies Presented at the 52 nd Meeting of the Philippine Economic Society, 14 November 2014, Hotel Intercontinental, Makati City
Background Research intensity ratio = 1.00 (public R&D spending on agriculture divided by agricultural GDP) has been cited as a benchmark for public expenditure on agricultural R&D For the Philippines, the 1997 AFMA elevates benchmark to a legal obligation: annual budgeted expenditure on agricultural R&D should reach 1% of the agricultural GVA two years earlier (starting 2001) However no annual budget since 2001 has reached this target. The current budget proposal for 2015 is nearest this benchmark; = 5.26 billion, vs AFMA benchmark of billion
Aims and scope This study: ex ante impact evaluation for meeting the AFMA benchmark (counter-factual scenario analysis) Applied to major food commodities starting 2013; these account fo 62% of agricultural GVA that year Method: economic surplus analysis, using Welfare Impact Simulator for Evaluating Research (WISER) Based on standard linear model and techniques in Alston, Norton, and Pardey (1995) Generates measures of project worth: NPV, BCR, IRR
Background: Shares in Agriculture GVA
Background: Output (million tons)
Background: Public outlays for agri R&D
Similar to allocation in Southeast Asia: Southeast Asian countries ranked among the last in a list of 60 countries: Myanmar (59) Vietnam (57) Indonesia (54) Laos (50). Even the Philippines, which ranks 36, has a research intensity ratio of 0.44, short of the median ratio of The highest ranking country in Southeast Asia is Malaysia at a ratio of 1.01 (ranked 16 th ).
Method Economic surplus analysis (linear model; implemented in GAMS) Thirty-year scenario starting % reduction in average cost during the first ten years after the research lag (i.e. years 5 – 15 from the baseline), then 20% reduction until the end of the scenario. Adoption follows a logistic process based on the following: adoption rate at baseline is 0.5%; in adoption in ten years is 60%; and ceiling adoption is 80%.
Method sensitivity analysis is conducted with regard to model parameters and exogenous variables. The variations are as follows: Parameters: 50% higher and 50% lower elasticities (respectively for demand and supply, in absolute value); Exogenous variables: Zero exogenous growth in demand and in supply; Research impact on processing and marketing efficiency.
Baseline data for simulatiosn
Counter-factual R&D expenditure per year
Results: Prices and quantities by 2043 Note: All prices in real terms.
Results: Net Present Values
Results: Benefit-Cost Ratios
Results: Internal Rates of Return
Conclusion Numerous qualifications: projections with and without the posited R&D investment technical assumptions related to elasticities of demand and supply, functional forms of supply, demand, and the adoption process; and single market competitive equilibrium under a closed economy. every effort has been taken in this instance to avoid arbitrariness Bottom line: Except for hogs and chicken under conditions of implausibly low supply response and intrinsic market growth, the worthiness of R&D investment at the 1% benchmark is robust.