Improvement of Finnish forest industries’ competitiveness in the world markets through productivity increase as branch and plant level requires the search for appropriate comprehensive productivity indicators and the analysis of factors underlying productivity variations. These were the main objectives of the study. The data was based on the information on individual plants in 1974, obtained from the files of Industrial Statistics in the Central Statistical Office in Finland.
The study uses neoclassical average production functions as the starting point and the theory is expanded to cover factors underlying productivity variation when measured with regard to labour, capital, material input, and total factor input. For the measurement of the latter an index formula is suggested which would not necessarily incorporate neoclassical assumptions as they cannot be assumed valid in the Finnish forest industries. The estimation results of average production functions suggest increasing rate of returns in sawmilling but in pulp and paper production evidence remains inconclusive. The elasticity of substitution is unlikely to be constant and the non-homotheticity assumption cannot be rejected.
The productivity variation is, in general, best explained by a relatively simple model with capital-labour ratio, plant size and output quality as explanatory factors. Further trials with input quality, input price ratio, process characteristics, and the rate of capacity utilization improved the models only marginally in most cases, which may have been partly due to the failure to measure the variables successfully.
The cross-section results are compared with those of an earlier time-series study. The estimation results of average production functions yield somewhat different information in the long and short run. Both cross-section and time-series productivity models illustrate the importance of output level in total productivity.
The PDF includes a summary in Finnish.