ANALYSISCan economic causes of tropical deforestation be identified at a global level?

This paper contributes to the literature on the macro-scale economic determinants of the increase in global forest depletion, by performing a regression analysis based on a panel dataset for fifty tropical countries over an eighteen-year period. While the initial findings appear at first to confirm a common causality pattern of selected macroeconomic variables in influencing tropical deforestation, subsequent statistical tests question the significance of the results. More specifically, testing for autocorrelation, which has been downplayed in previous studies, appears to represent a critical issue. When the initial results are corrected for autocorrelation, the significance of the parameters declines substantially below statistically acceptable levels. The implications are twofold. First, regression analyses that seek to explain deforestation at the global level need to be carefully scrutinised and checked to ensure that they meet standard statistical tests. Second, tropical deforestation might ultimately depend upon case-specific factors and further research may render more effective policy suggestions if conducted at a more disaggregated, local level.