TY - GEN
T1 - The Causes of Tropical Deforestation and Contemporaneous Correlation
AU - Culas, Richard
N1 - Culas, R. J (2002): The Causes of Tropical Deforestation and Contemporaneous Correlation, Papers and proceedings of the PhD Conference in Economics and Business, Australian National University, Canberra ACT, Australia. 6-8 November 2002. p. 33
PY - 2002/11/6
Y1 - 2002/11/6
N2 - This paper analyses causes of tropical deforestation in three regions (Latin America, Africa and Asia) covering 52 developing countries over a period from 1991 to 2000. Causes of deforestation are grouped in two levels: direct (first-level) and indirect (second-level) causes. The direct causes come from consumption and export demands for forest products as well as from demand for land use such as cropland and pasture. The indirect causes are due to population, national income, external debt, government policies, etc. A cross-sectional econometric model, recursive in nature, is estimated in two stages. In the first stage, the direct causes of deforestation are estimated using the indirect causes as the explanatory variables. Zellner’s seemingly unrelated regression (SUR) estimation is applied in order to correct for contemporaneous correlation of the equations for direct causes. In the second stage, the SUR estimates of the direct causes are used to regress the deforestation equation. Since the LM test statistics demonstrate the importance of addressing the contemporaneous correlation in the model, the SUR based estimates are then discussed for the effect of different direct and indirect causes on deforestation in three geographic regions. The overall results suggest that the effects of different causes of deforestation vary somewhat across the three regions and that each region needs some specific policy prescription.
AB - This paper analyses causes of tropical deforestation in three regions (Latin America, Africa and Asia) covering 52 developing countries over a period from 1991 to 2000. Causes of deforestation are grouped in two levels: direct (first-level) and indirect (second-level) causes. The direct causes come from consumption and export demands for forest products as well as from demand for land use such as cropland and pasture. The indirect causes are due to population, national income, external debt, government policies, etc. A cross-sectional econometric model, recursive in nature, is estimated in two stages. In the first stage, the direct causes of deforestation are estimated using the indirect causes as the explanatory variables. Zellner’s seemingly unrelated regression (SUR) estimation is applied in order to correct for contemporaneous correlation of the equations for direct causes. In the second stage, the SUR estimates of the direct causes are used to regress the deforestation equation. Since the LM test statistics demonstrate the importance of addressing the contemporaneous correlation in the model, the SUR based estimates are then discussed for the effect of different direct and indirect causes on deforestation in three geographic regions. The overall results suggest that the effects of different causes of deforestation vary somewhat across the three regions and that each region needs some specific policy prescription.
KW - Causes of tropical deforestation, contemporaneous correlation, Zellner’s seemingly unrelated regression, two-stage estimation, policy issues
M3 - Conference paper
SP - 1
EP - 33
BT - Papers and proceedings of the PhD Conference in Economics and Business
PB - Australian National University
CY - Canberra ACT, Australia
ER -
