Economic forecasting is a challenging pursuit. The likelihood of inaccurate predictions is magnified where underlying structural change is occurring. Therefore, it is important to consider the likely future structure of the economy when analysing the impact of potential policy changes. Where structural change is predicted in a baseline forecast the effects of policy changes can widely differ from a status quo assumption about the future structure of the economy. This paper examines methods aimed at improving baseline economic forecasts using a dynamic CGE model. Forecasting can be used to test the validity of such models, as well as to highlight possible improvements, by investigating the discrepancies between the forecast and actual outcomes. The model that is employed in this paper is USAGE – a recursive dynamic, 500‐industry CGE model of the U.S. USAGE generates baseline forecasts by incorporating expert forecasts for certain macro variables and extrapolating historical trends in technology, consumer preferences, positions of foreign demand curves for U.S. products, and numerous other naturally exogenous variables. In instances where important trends either dissipate or reverse, large forecast errors can arise. This paper provides explanations and guidance as to whether these various trends from the period 1992 to 1998 would continue for the 1998 to 2005 USAGE forecast.
It is found that for some commodities, certain important trends should not have been expected to continue, and hence a better forecast could have been generated had all publicly available information at 31 December 1998 (the end of the base year of the forecast) been appropriately utilised. In examining the largest of the forecast errors, macro and industry‐specific commentary is included. It is shown that a much‐improved forecast for certain commodities would have arisen had specialist knowledge of industry trends and conditions been properly accounted for. This sometimes meant nullifying the projection of certain trends. This was the case for USAGE commodities: AsbestosPrd, ComFishing, ElectronTube, Dolls, Theatres, and Recordmedia. More generally, the findings suggest that there is a case to be argued against projecting forward large values relating to import‐domestic preference twist factors in particular.
It is shown that for commodities in the trade‐exposed textile, clothing and footwear industries moderately better results could have been produced by implementing import price forecasts in a way that is more in line with historical trade policy. This was achieved by projecting forward real basic import prices. There are 31 commodities in this space, and 8 of these featured among the 20 largest USAGE forecast errors. However, the key drivers behind these errors were usually the significant underestimation of the impact of import‐domestic preference twist factors, as well as the overestimation of factor input cost savings. In most cases, a lack of convincing evidence (available by 1998) in this sector meant that additional error correction strategies might not have been implemented.
It is concluded that forecasts for commodities in the oil and mining sectors as well as companies that service these cyclical industries typically could not have been improved in the absence of strong convictions (in 1998) about an impending mining “super‐cycle” or extended boom. These USAGE commodities are: AccStrucSMD, PetNgExplor, PetNgDrill, Nonferrores, Copperore. For the construction‐related commodities, such as CutStone, demand was fuelled by virtually unprecedented low borrowing costs. In these instances, it is difficult to conclusively argue that the modeller could have produced a better forecast.
Furthermore, it is noted that where commodities have large import shares (e.g., Dolls, and Luggage), it is always going to be difficult to accurately forecast domestic output in the absence of specialised knowledge given that total supplies of domestic goods will move off a low base. In this instance, the model does a better job at predicting the commodity’s absorption, i.e., all U.S. sales of the commodity irrespective of source.
Moreover, while large improvements in forecast accuracy can be obtained for some industries and sectors, the overall economy‐wide forecast error does not fall greatly due to the sheer volume of commodities. While it is disappointing that the error is not very reducible, it is also reassuring because it implies that the default implementation of the model is quite powerful. In all the twenty worst errors on a relative and/or absolute basis (about 4% of all commodities) were specifically examined to assess the potential for error reduction. However, after due consideration about 7.5% of commodities were in some way directly re‐projected. To generate a large reduction in the forecast error (and hence improvement in model performance) would require an extensive amount of work and probably necessitate the input of numerous industry specialists.
An important contribution to this paper was made by Marnie Griffith. This is available in the form of an appendix, which is in two parts. The first part discusses options for achieving more accurate sectoral‐level forecasts with the USAGE model. This includes ideas such as analysing year‐by‐year trends rather than extrapolating the overall period; a discussion of ‘future studies’ methodologies; and an extensive list of sources of useful information, including organisations, people of expertise and publications. The second part of the appendix provides an examination as to whether the magnitude of China’s rise could have been predicted. This concludes that to the extent that the USAGE forecasts incorporated the extrapolation of previous trends in world prices, this might have been best possible practice. The exception is for the TCF sector, for which a known blockage to trade (import quotas) was removed. However, in 1998 the extent to which this would occur by 2005