Abstract ID: 592

Protected Area Effects On Deforestation Controlling For Other Drivers

Measuring the effects on deforestation of protected areas can be more difficult than often assumed because the deforestation that would have occurred in the absence of protection cannot be directly observed. Conventional methods of evaluating protected area impact could be biased because protection is not randomly assigned and it can induce spillovers. Following similar work in other countries, we will examine whether impact estimates can be substantially improved by controlling for biases along dimensions that are observable. Observational units are pixels for across the entire Legal Amazon, from satellite metrics. The time period considered is 1976 through 2000, broken into three roughly equal parts. We will apply matching methods to evaluate the impact of protection upon deforestation. This method ?compares apples to apples? by searching for the comparison locations that are most similar to the protected locations in terms of the observable land characteristics. Further, to support development and conservation planning, we can ask whether impact on deforestation varies with observable characteristics planners could use to prioritize. When allocating scarce resources, policy makers have to consider many critical factors, such as ecosystem services provided by a site and the suite of costs for acquiring a site. Holding such factors fixed, we conjecture that impact can be raised by protecting first, in any sequence of protection, the candidate forest less likely to survive without protection.

Session:  LCLUC and Human Dimensions - Social-economic drivers of land-use and land-cover change.

Presentation Type:  Poster