A spatial regression analysis of Colombia’s narcodeforestation with factor decomposition of multiple predictors

Abstract

Abstract In the current accelerated process of global warming, forest conservation is becoming more difficult to address in developing countries, where woodlands are often fueling the illegal economy. In Colombia, the issue of narcodeforestation is of great concern, because of the ramification of narcoactivities that are affecting forests, such as agribusinesses and cattle ranching for money laundering. In this study, we use spatially explicit regressions incorporating a factor decomposition of predictors through principal component analysis to understand the impact of coca plantations on global and local-scale deforestation in Colombia. At national level we find a positive and statistically significant relationship between coca crops and deforestation. At the regional level, in two out of four regions, it appears that coca is causing deforestation, especially in the Department of Northern Santander and on the Pacific coast. The spatial models used reveal not only a direct effect but also positive and significant spillover effects, in line with the conjecture that narcodeforestation is not only due to the quest for new areas to expand coca-cultivation, which would determine a loss of forest only in the municipality where coca cultivation increases, but also to the need to launder illegal profits, or create clandestine routes and airplane strips, which can affect forests also in nearby municipalities.