Deforestation in the Amazon: Measuring the Effects of the `Priority List'

Abstract

The deforestation in the Brazilian Amazon slowed down recently. Part of the recent slowdown may be explained by increased monitoring efforts that took place in the past 10 years. A particularly important change occurred in 2008. In that year, the Brazilian federal government issued a list of municipalities with high deforestation rates that became subject to more stringent monitoring efforts. The objective of this paper is to estimate the effect of the blacklist policy on the deforestation levels in the Brazilian Amazon. To estimate this effect we face a number of challenges: the selection rule is not purely deterministic; there is no parallel trends (so we cannot exploit Diff-in-Diff strategy); there are relatively small number of treated observations. We therefore give up point identification and apply a partial identification approach. Using the imperfect instruments strategy proposed by Nevo and Rosen (2012), we find tight bounds and significant impacts of the blacklist policy on deforestation rates.