Econometric modelling for agricultural policy analysis and forecasting: between theory and reality

• Main Author: Shamsudin, Mad Nasir
• Format: Inaugural Lecture
• Language: English; English
• Published: Universiti Putra Malaysia Press 2007
• Online Access: http://psasir.upm.edu.my/id/eprint/18196/1/Cover%20Econemetric.pdf; http://psasir.upm.edu.my/id/eprint/18196/7/MALAYSIAECONOMETRIC%20MODELLING.pdf; http://psasir.upm.edu.my/id/eprint/18196/
• Institution: Universiti Putra Malaysia

The inherent characteristics of the agricultural sector, like price instability, time lag between decisions and the resultant outputs, and risk and certainty, necessitate the development of econometric models for policy analysis and forecasting. The need for accurate forecasts and an empirical tool for policy evaluation are reinforced as the sector undergoes structural changes and becomes complex due to trade liberalization, globalisation, technological advances, and consumers’ changing tastes and preferences. Econometric models have long been viewed as a valuable aid to deal with this apparent complexity. The contribution of econometric models to the analysis of the agricultural sector, in Malaysia and in other countries, has been significant. The models generate quantitative forecasts and enhance the ability of those involved in planning and policy-making in evaluating the effects of policy changes. The models, however, have limitations since models represent an abstraction of a rather complex real world. This lecture therefore attempts to discuss the theory and the reality of the econometric modelling of agricultural commodity for policy analysis and forecasting. Agricultural commodity markets are complex, often with many actors and sectors involved in production, consumption, inventory holding, capacity formation, and trade, and also the competitive nature of the overall market structure. Although the theoretical framework of econometric modelling is easy, the reality can be challenging since the empirical content of the model must not only reflect the essential structure of the model, but also the behaviour of government in imposing policy interventions. Given the enormity of the task, model applications should be improved in the direction of employing newer approaches to market and policy simulation, and forecasting works should embody new probabilistic approaches for evaluating risk impacts. Continued work is also needed on the means of incorporating the effects of changing input prices, such as energy and technological developments on the supply side, and new products and changing tastes on the demand side.