The base-year is the first in a series of spatial equilibrium solutions, one for each year of the projection period. For a thorough treatment of spatial equilibrium models, see standard texts such as Hazell and Norton (1986) or Thore (1992).
The GFPM framework is a primal representation of the market equilibrium, i.e. its variables are quantities (production, consumption, and imports and exports). Prices emerge from the optimization problem (see Appendix B) as shadow prices. Both prices and quantities given by the base-year solution should be close to the input data. Calibration consists of adjusting model parameters so that input data and model output in the base year match. Calibration is important to check for data consistency and to check how reasonable model parameters are. For example, it is often the case that, when plausible manufacturing coefficients are used to derive the required consumption of industrial roundwood in a country, this amount is far greater than the level of consumption reported in official statistics. The current version of the GFPM was built by relying more on statistics for manufactured products and input-output coefficients than on raw material statistics. Consequently, when the base-year is updated it should be remembered that:
· Using the macros (see Appendix I) will speed us model preparation.
· For many countries, industrial roundwood data are inaccurate. In calibrating the base year, it is important to make sure that there is enough wood to support reported production of downstream products (see previous appendix).
· Because the WORLD model output is an input to the regional models, it is wise to have a good WORLD model before investing much energy in fine-tuning the regional models (though some going back and forth is inevitable). See Appendix G for details on the links between the WORLD and regional models.
· Changing the manufacturing coefficients in the base year of the GFPM also results in changes to the manufacturing costs and raw material supply as well.
More accurate price computations can be obtained by increasing the number of steps (menu option 2) in the demand and supply equations. However, this increases the model size and the difficulty of getting solutions. The current GFPM set-up of 12 steps in the WORLD model and 6 steps in the regional models has proven from experience to be a good compromise.