The Food and Agriculture Organisation of the United Nations (FAO) carries-out periodic reviews of the forestry sector and produces long-term forecasts of wood and wood product supply and demand. The production of reliable and timely forecasts is seen as an important aid to planning and decision making in the forestry sector at a national, regional and global level, which FAO will continue to strongly support in the future.
Most projections of future supply and demand are based on a variety of statistical or econometric techniques. Such techniques examine historical trends and changes in supply and demand and attempt to explain these trends by identifying relationships with other variables, such as: forest product prices; the availability of forest resources; and economic growth. Assumptions about future changes in these variables are then used to make projections of future supply and demand.
A major challenge in this type of analysis is always to identify and understand the complex interlinkages between the supply and demand sides of the forest products market. For example, when growth in forest products demand exceeds growth in the availability of resources (and, consequently, supply), the resultant scarcity usually leads to upward pressure on forest product prices. This can, in turn, encourage increased production from existing areas, an expansion of the areas of forest used for wood supply or, in the global context, increased imports of wood products from other countries. However, although the role that prices play in helping to balance supply and demand within and between different parts of the world is well understood in theory, the way in which prices change in response to changes in supply and demand are often very difficult to model in practice and even more difficult to project into the future.
Until fairly recently, most models of forest products supply and demand usually treated the two sides of the market quite separately. This avoids the problem of having to specify the mechanism by which prices change in response to changes in supply and demand. Rather, price is treated in such models as an explanatory variable (rather than as a dependent variable) and projections can be made under a range of assumptions about future prices. The approach used in the current FAO forest products outlook study however, has been to use a price-endogenous model. In such a model, changes in price are determined internally by the model in response to predicted changes in supply and demand (rather than the other way around). This is a more complex way of modelling supply and demand, but is a much better representation of the actual way in which forest products markets work. It also has the advantage that, at the global level, the model will give supply and demand projections that are always in balance.
The use of the price-endogenous modelling framework is not particularly new, but may be unfamiliar to many analysts in the forestry sector. Therefore, this working paper has been produced to try to explain some of the details of what is quite a complex model. By demystifying the model, it is also hoped that this will increase the understanding of how the forecasts presented in GFPOS Working Paper 1 were produced and generate a broader interest in the methodology used in this study.
The paper is also a basic instruction manual for the model. Any researcher who wishes to use the model can request a copy (which will be supplied without any warranty or guarantee) from FAO. The model also requires various other pieces of commercial software, which would have to be purchased from the usual suppliers.
The Global Forest Products Model (GFPM) is the result of a considerable amount of research over several years, by Professor Joseph Buongiorno and his team of researchers at the Department of Forest Ecology and Management at the University of Wisconsin (Madison). FAO would like to express its gratitude to the team, for the way in which they have developed the model to meet the specific needs of FAO. In particular, special thanks must go to David Tomberlin, who came to Rome twice in 1998, to install the model on FAO computers and train staff in the use of the model.
FAO will continue to explore ways in which the quality of future supply and demand projections can be improved through improvements in the collection of forest product statistics and the models used to make such projections. In this respect, we would welcome comments on all aspects of this study from professional analysts and users of this study (contact details can be found on page v of this working paper).
Adrian Whiteman
Forestry Officer (Sector Studies)
Forestry Policy and Planning Division