Joint production and substitution in timber supply: a panel data analysis

Supply equations for sawlog and pulpwood were developed with a panel of data from 102 Norwegian municipalities, observed from 1980 to 2000. Static and dynamic models were estimated by cross-section, time-series andpanel data methods. A static model estimated by first differencing gavethe best overall results in terms of theoretical expectations, pattern ofresiduals, prediction accuracy and parsimony. The results showed thatsawlog supply responded positively to its own price (elasticity e = 0.91 ± 0.07) but negatively to the pulpwood price (e = -0.22 ± 0.06). The pulpwood supply responded positively to the price of both pulpwood (e = 0.53 ± 0.06) and sawlogs (e = 0.20 ± 0.07). Sawlog and pulpwood supply had a common elasticity of 2.04 (± 0.25) with respect to the growing stock, and of 0.30 (± 0.21) with respect to the interest rate. The supply elasticity of substitution of sawlog for pulpwood with respect to their relative price was 0.74 ± 0.04. Policies to raise the annual harvest, which is currently well below the annual growth, should focus on stimulating sawnwood production (thus increasing sawlog prices), because this would increase supply of both pulpwood and sawlogs. Instead, policies to stimulate pulpwood demand (thus increasing pulpwood prices), would give more pulpwood, but less sawlogs.