Protected Areas’ Deforestation Spillovers and Two Critical Underlying Mechanisms: An Empirical Exploration for the Brazilian Amazon
To date, the creation of protected areas (PAs) has been the dominant policy in the efforts to protect forests. Yet there is still somewhat limited rigorous evidence about the impacts of PAs on rates of deforestation. Further, most of the existing evidence concerns the impacts of protection within the boundaries of PAs. Much of that existing evidence does not use the characteristics of the protected lands when generating the baselines to which outcomes on protected lands are compared in order to infer the PAs' impacts. Yet even when impact within a PA has been estimated as rigorously as possible, since the total impact of protection involves impact not only inside the PA but also outside the PA even the best possible estimates of impacts within PAs could mis-state total PA impacts. Overstatements occur if there is "leakage" from PAs, i.e., spillovers of activities to forests outside PAs, so deforestation outside is higher than it would have been without the PAs.
My dissertation starts with a reduced form examination of net local spillovers. We follow this with an evaluation of two mechanisms through which PAs could affect forest nearby. In particular we explore two novel angles by considering both migration choices and road building decisions. PA creation could affect the development equilibrium by shifting private and public expectations to lower migration and road building where the PA is established, beyond the PA's boundaries. My dissertation explores implications of such thinking and provides novel empirical evidence for the Brazilian Legal Amazon.
Chapter 1 estimates deforestation spillovers around Brazilian Amazon PAs. Given PA location bias towards regions with low deforestation pressure, we use matching methods to control for observable land characteristics that may confound PAs' impacts. Specifically, we compare 2000-2004 and 2004-2008 deforestation on the land nearby to PAs with clearing of untreated forests similar in key deforestation determinants. We find that some PAs reduce deforestation rates nearby and, consistent with deforestation impacts inside PAs, those local spillovers vary across the landscape. Reductions are significant near roads and cities − not expected if the result is due to insufficient empirical controls but unsurprising if real impacts are arising due to PAs − and around an understandable subset of PAs. This result contrast sharply with most existing analyses of PAs' spillovers where, if anything, 'leakage' (higher nearby clearing) is discussed and observed. Yet we affirm a more general point that local spillovers depend on local development dynamics.
Chapter 2 examines one mechanism for the prior result that PAs lowered rates of deforestation nearby. Given migration's importance throughout the history of this forest frontier, we ask whether dissuading migration could be a mechanism for protection's local conservation spillovers. Examining individual migration decisions among the Amazon municipalities, we find that Federal PAs − previously seen to reduce rates of deforestation near PAs − seem to encourage outmigration from and discourage migration to PA areas.
Chapter 3 examines another mechanism for the result in my Chapter 1. We consider a recent expansion of the unofficial roads networks in the Brazilian Amazon to provide initial evidence concerning whether PAs may affect such investments in development. Specifically, controlling for prior roads − both official and unofficial − we test whether the growth in unofficial roads between 2008 and 2010 is reduced by establishments of PAs. Thus, we examine road growth as another potential mechanism for forest spillovers from PAs. Controlling for relevant observable factors, and using both matching and OLS, we find that having a large fraction of municipal area in PAs − in particular Federal PAs − reduces the growth of unofficial roads. Such impacts can significantly influence regional development patterns.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Masters Theses