| I study empirically and theoretically deforestation and land conservation policies. I use geographic data to study the deforestation process in space and its implications in policy implementation. I also study how land conservation policies can affect local welfare and income distribution. Finally, I look into a the effects of voluntary forest conservation policies on deforestation.Chapter 1 is the result of joint work with Alex Pfaff. We estimate the effects of neighbors' deforestation actions on individuals' deforestation decisions in Costa Rica. To address simultaneity and the presence of spatially correlated unobservable factors, we instrument for neighboring deforestation using neighboring land parcel characteristics. We find evidence that neighbors' deforestation effects are positive and significant. These findings are important for projecting forest levels and spatial patterns. Policies that promote agricultural development or forest conservation in a specific area also affect deforestation rates in non-targeted neighboring areas. Moreover, positive interactions create conditions in which multiple equilibria might arise. This could justify intervention to reach Pareto dominant land-use outcomes.In Chapter 2, I analyze the impact of land conservation policies on income distribution using a two-sector model. I find that conservation policies can have important distributional effects through changes in rents and wages. I show how aggregate rents rise when protected areas increase despite the reduction of land availability. Simultaneously, real wages decrease in consequence of higher agricultural prices. These distributional changes also affect the efficiency of conservation policies since higher rents lead to deforestation elsewhere. Results suggest that Pareto improving compensation should also be aimed at agricultural workers.Finally, in Chapter 3, I study the Payments for Environmental Services implemented in Costa Rica. I find that the impact of the payments on deforestation rates is negligible. The government's costs of saving an additional hectare of forest are significantly higher than the environmental benefits. This might be due to the low levels of deforestation as a consequence of other factors, to the mechanism in which the program is implemented, or to the government's objective of transferring most of the benefit to landowners. Targeting high-pressure areas lowers the costs of the program. This can be accomplished by a new selection mechanism or by increasing payments only in high-pressure areas. |
