Modeling and Forecasting the Demand for Industrial Roundwood in Turkey: a Primary Econometric Approach

Abstract

This study is a primary econometric analysis to explore the factors explaining the changes in industrial roundwood demand in Turkey. The study also includes demand forecasts based on the econometric models proposed herein. We constructed two separate econometric models: one for national demand for domestically-produced saw log, and the other for national demand for domestically-produced non-sawlog industrial roundwood. Models were originally designed in multiplicative form. The original models are then converted into the log-linear form so that the relevant coefficients of the regression equations would immediately reflect the elasticities. Estimation of the model parameters are based on a panel data set of fifteen years (1995-2009) by twenty seven regional forest directorates in the country. In accordance with the maxim of less than the half of the 15 years period of data set, the demand forecasts are made for seven years beyond 2009. In view of the results, the explanatory power of the proposed models can arguably be deemed satisfactory especially considering the lack of earlier studies of this scale and scope. This consequently increases the credibility of the demand projections. Notwithstanding signs of the estimated parameters of the models are for the most part congruent with those expected in light of the economic theory and practice, some intriguing results are obtained. Perhaps most notably, while the sign of the estimated price elasticity of sawlog demand occurred unexpectedly positive, the variation in sawlog demand is explained to a considerable extent by the variation in the price of imported sawlog, hence an expected cross elasticity. Also notable is that the price of imported "fuelwood" holds a positive relationship with the national demand for domestic non-sawlog imdustrial roundwood, which is an expected cross elasticity since virtually all of the imported "fuelwood" is used as raw material for industry (e.g. chip and fiberboard industry). Finally, both models suggest overall boost in demand: yet an upper bound of 4.5 million m(3) for national demand for domestically-produced sawlog, and of 15 million m(3) for national demand for domestically-produced non-sawlog industrial roundwood can be expected by 2016.