This dissertation studies the deforestation in the Brazilian Amazon using semi-parametric methods. In the first chapter, I estimate the demand for deforestation in the Amazon. This demand can be used to investigate multiple policy interventions with the ultimate goal of preventing the depletion of the rainforest. To recover the demand curve, I use a revealed preference approach and exploit the fact that regional variation in transportation costs can be used to infer variation in the private value of forested land relative to agricultural land. By resealing these costs, I am able to value the difference between forested versus agricultural land in dollars per hectare. As a result, from the impact of a change in transportation costs on landowners' decisions, it is possible to infer by how much the value of forested land would have to increase relative to the value of agricultural land to avoid deforestation. The estimates suggest that both Pigouvian taxes on agricultural land and payments to avoid deforestation would have been effective in preserving the rainforest. The results also indicate that large landholders are the most responsive to the programs, which, together with the unequal distribution of land in the Amazon, suggests that these programs are unlikely to reduce local poverty and deforestation simultaneously. Finally, I show that the existing "command-and-control" policy would be too costly for farmers, when compared to payments and taxes, if it were fully enforced.
In the second chapter, I develop a nonparametric estimator for the generalized regression model proposed by Berry and Haile (2009) in which each individual is associated with a group and each group is subject to observable and unobservable shocks. In the previous chapter, the farmers are the individuals and the municipalities are the groups. The motivation for this model is to estimate the effects of group-level observables, such as transportation costs, on individual outcomes, such as farmers' land use decisions, when group-level observables correlate with group-level unobservables. Furthermore, the latter can be indexed by individual characteristics, such as farm size, which allows for more general group shocks than existing approaches. I propose a two-step estimator in which the first step runs a nonparametric regression of individual outcomes (landowners' decision) on individual characteristics (farm size) within each group. It is a nonparametric regression in the presence of common shocks. The second step fixes the individual characteristics (farm size) and runs a nonparametric quantile instrumental variable regression across groups of the predicted outcome obtained in the first step (i.e., the predicted share of deforestation) on group-level variables (transportation costs). It separates the effects of group-level observables from unobservables. I establish consistency and convergence rate of the estimator.
In the third chapter, I consider a nonparametric regression model for cross-sectional data in the presence of common shocks that are very general in nature. I discuss how general the common shocks can be while still obtaining meaningful kernel estimators. Kernel estimators typically manipulate conditional densities, but conditional densities do not necessarily exist in the present case. For this reason, I provide sufficient conditions for the existence of conditional densities, and I show that the estimator converges in probability to the conditional expectation given the sigma-field generated by the common shocks. I also establish the rate of convergence and the asymptotic distribution of the kernel estimator.
|Advisor:||Haile, Phillip A.|
|School Location:||United States -- Connecticut|
|Source:||DAI-A 73/12(E), Dissertation Abstracts International|
|Subjects:||Environmental economics, Statistics, Latin American Studies|
|Keywords:||Amazon, Brazil, Common shocks, Deforestation, Group-effects, Non-parametric, Rainforest|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be