Estimated short-run demand elasticities for prices and the end-use indicators for the Group I countries (major markets and producers) are shown in tables 3–10. Estimated short-run elasticities for the export supply equations for these countries are shown in tables 11–18. Estimated long-run price and income elasticities for price and GDP for the smaller markets are shown in Table 19. The significance of estimated elasticities is shown; complete statistical results are available from the authors.
As described above, some equations were not estimated for some products in some countries; equations not estimated are indicated by the entry "ne" in all cells in tables 3–18. In a number of cases, results were affected by missing or weak data; the entry "na" indicates either that data were not available, or that the variable was not included in the estimated model.
A few comments can be made on these results. First, note in tables 3–18 that there are large differences in estimated elasticities among countries. End use elasticities vary, as do price elasticities. Compared to previous approaches to multi-country studies, these differences across countries are notable, and are directly attributable to the fact that most elasticities were estimated on a country-by-country basis (for the Group I countries). In addition, included variables, and model form also varied.
There are some relatively consistent patterns in the estimated elasticities, however. First, as in previous TTS, the results are strongest for the more important products (such as coniferous sawn wood, tables 3 and 11). For this product, both demand equations have the expected signs for the “own price,” and for both equations (domestic demand and import demand) it is important to note that the hypothesis of substitution between imports and domestic sources in consumption is confirmed in nearly all countries in which imports are significant. The results of the examination of the hypothesis of substitution between domestic and export markets are less clear. In addition, elasticities in import demand equations often are higher in absolute value; this is consistent with the general observation that adjustments in trade are more volatile than adjustments in domestic markets.
The evidence of substitution is somewhat less clear for the other products, primarily as a result of weak or missing domestic price data. Nevertheless, where both prices are available, there is evidence of the same substitution that is observed for coniferous sawn wood. A similar pattern of generally higher elasticities for import demand, compared to domestic demand also can be observed. Results are weakest for product groups such as fiberboard, that account for a relatively small component of forest products consumption. Here, the import demand equations are the strongest, primarily because price data are best for this equation. The pattern of higher end-use elasticities for imports, compared to domestic demand still holds.
Estimated elasticities of demand for the countries that are smaller markets are reported in table 19; these are long-run elasticities, calculated using the coefficient on the lagged dependent variable (QT-1) to estimate long-term adjustment. The estimated elasticities are generally consistent with the results reported in ETTS IV, and with results from other studies. (See Buongiorno (1978) and Baudin and Lundberg (1987) for examples of comparable price and income elasticities for paper and board product groups.) In table 19, only one price elasticity has an unexpected sign that is significant (coniferous sawn wood in the lower income group); the remaining price elasticities have expected signs (they are negative), but fewer than half are significant. Weak results for price may be attributable, at least in part, to the use of import price as a proxy for all relevant prices. Income elasticities are consistently positive and, with only a few exceptions, are statistically significant.
Tables 20–22 compare the price elasticities of demand estimated here with those reported in ETTS IV, for three major product groups. Tables 23–25 compare estimated end-use elasticities for these same products. In addition to providing estimates of price elasticities for both domestic demand and import demand (and, in many cases, cross-price elasticities that are not shown in these tables), we also have been able to provide estimates of price elasticities for a larger number of countries. As mentioned above, import demand elasticities are frequently higher in magnitude than domestic demand elasticities, and our estimates tend to bracket the values estimated for ETTS IV.
The comparison of estimated end-use elasticities (tables 23–25) is similar: import demand elasticities are generally higher than domestic demand elasticities, and our results are comparable to those in ETTS IV. The exception is particleboard, where our estimates of end-use elasticities are generally lower than those previously estimated. One explanation for this is the fact that we were able to use data for the period 1980-90, and our estimates are less strongly influenced by the rapid increase in particleboard consumption that occurred in the 1970s.
Overall, the results from the multiple equation approach are quite encouraging. Information is provided on additional factors that influence both consumption and production of forest products--prices of possible substitute sources and markets. This information has a direct bearing on the outlook for forest products markets in Europe, and the implications of this outlook for demands on European forests.
