Econometric Estimation of Deforestation Impacts from Roads and Other Drivers
Pfaff, Columbia University, firstname.lastname@example.org
Reis, IPEA/Rio, email@example.com
Robalino, Columbia University, firstname.lastname@example.org
Walker, Michigan State University, email@example.com
Drawing on several papers, we will present an econometric estimation of the effects of a suite of factors that drive deforestation. We will in part focus on investments in roads and we note that road investment and other driving factors may also respond to deforestation (for instance, others show that the climate may respond and so may development patterns determined by private migration decisions as well as locations of various public policies). Further, it seems important to recognize that impacts of interventions may depend on the location and time period in which they occurred. Our analyses reflect all of these issues. Starting with a description of our data, we emphasize its temporal and spatial high quality with deforestation observed for census tracts and for three periods covering 1976 to 2000. The spatial resolution permits much improved statistical controls for unobserved drivers, while the temporal coverage with multiple observations of not only forests but also roads permits analysis of dynamic interaction between forest clearing and the building of roads. Further, deforestation can be examined for changes in driving process across the periods and for an effect of drivers across the census tracts, i.e. spatial interaction in the process. Key results are summarized in analyses that use all locations and periods. While details of any prior development affect their impacts, investments in paved and unpaved roads increase deforestation not only in the period when they occur but also in future periods.
Science Theme: LC (Land Use and Land Cover Change)