Production structure and input substitution in Canadian sawmill and wood preservation industry Rao V. Nagubadi and Daowei Zhang Abstract: We use a translog cost function to analyze the sawmill and wood-preservation industry in Canada from 1958 to 2003. The estimated cost function is homothetic and Hicks neutral. Allen elasticities of substitution indicate that sig- nificant substitution possibilities exist. According to Morishima elasticities of substitution, substitution of labour by other inputs is easier than the substitution of other inputs by labour, and substitution of other inputs by materials is easier than the substitution of materials by other inputs. The demand for production labour, nonproduction labour, and electric power inputs is elastic. In contrast to the previous findings of zero or negative rate of technical change, we find technical progress at the rate of 0.57% per annum and a total factor productivity growth rate of 0.54% per annum in the Canadian sawmill and wood preservation industry. Résumé : Nous utilisons une fonction du coût de type translog pour analyser l’industrie du sciage et de la préservation du bois au Canada, de 1958 à 2003. La fonction du coût estimée est homothétique et neutre au sens de Hicks. Les élasticités de substitution d’Allen indiquent que de fortes possibilités de substitution existent. Selon les élasticités de substitution de Morishima, la substitution de la main-d’œuvre par d’autres facteurs de production est plus facile que la substitution des autres intrants par la main-d’œuvre et la substitution des autres intrants par les matériaux est plus fa- cile que la substitution des matériaux par les autres intrants. La demande pour les intrants sous forme de main-d’œuvre responsable de la production, de personnel de soutien et d’énergie électrique est élastique. Contrairement aux conclu- sions précédentes concernant la mesure d’un taux de progrès technique nul ou négatif, nous observons un taux de pro- grès technique de 0,57 % par année et une productivité globale des facteurs de production de 0,54 % par année dans l’industrie du sciage et de la préservation du bois au Canada. [Traduit par la Rédaction] Nagubadi and Zhang 3014 Introduction The sawmill and wood-preservation industry is an impor- tant industry in Canada, contributing 2.06% of the total value of manufacturing shipments in 2003. Previous studies of technical change in the industry in Canada present mixed results. Several studies using data between 1955 and 1984 (Martinello 1985; Nautiyal and Singh 1985; Singh and Nautiyal 1986) report negative or no technological progress in the Canadian sawmill or softwood lumber industry. How- ever, Ghebremichael et al. (1990) estimate 0.4% per year im- provement in total factor productivity (TFP) growth for the lumber industry between 1962 and 1985, whereas Bernstein (1994) reports a TFP growth of 3% per year for softwood lumber industry between 1963 and 1987. The objective of this study is to analyze the production structure and input substitution in the Canadian sawmill and wood-preservation industry. We use a translog cost function approach. The total cost function is estimated simulta- neously with the cost-share equations for seven inputs — production labour (P), nonproduction labour (N), machinery and equipment capital (Q), plants and structures capital (S), fuels energy (F), electric power (R), and materials (M). The output is the sum of all products in the industry, including softwood lumber, hardwood lumber, wood chips, wood- preservation products, wood ties – shingles – shakes, and other products. In addition to the usual Allen elasticities of substitution and own and cross price elasticities of demand, we also compute Morishima elasticities of substitution be- tween inputs. This study differs from previous studies in several aspects. Firstly, previous studies (e.g., Ghebremichael et al. 1990; Abt et al. 1994) cover only lumber and woodchips but not other important segments of lumber production, wood- preservation products, and shakes and shingles. Since chemi- cally treated lumber and wood-preservation products have become an important form of lumber consumed in the out- door applications of home construction and other uses (Nagubadi et al. 2004), we combine the sawmill industry and wood-preservation industry in this study. Secondly, previous studies use aggregated inputs and do not take into account the difference between production la- bour and nonproduction labour, machinery and equipment (M&E) capital and plants and structures (P&S) capital, and fuels energy and electric energy. To demonstrate the differ- ences between these subdivisions of inputs, we use seven in- puts in our analysis. Thirdly, a few studies use value added minus labour cost as the cost share of capital in valuing cap- ital services, which leads to unrealistically high estimates of service price of capital as a percentage (in some years as high as 151%) of total capital stock. In this study, we use the Can. J. For. Res. 36: 3007–3014 (2006) doi:10.1139/X06-187 © 2006 NRC Canada 3007 Received 10 June 2005. Accepted 28 June 2006. Published on the NRC Research Press Web site at http://cjfr.nrc.ca on 20 January 2007. R.V. Nagubadi and D. Zhang. 1 School of Forestry and Wildlife Sciences, Auburn University, AL 36849, USA. 1 Corresponding author (e-mail: zhangdw@auburn.edu).