Essays on the Environment and Development

Two essays in this dissertation explore how developing countries’ effort to improve development outcomes and to protect their forest areas affect deforestation. The first chapter investigates the effect of a conditional cash transfers (CCT) program, a rapidly growing poverty alleviation program in developing countries, on deforestation. Although an increasing number of studies measure the effects of environmental programs on poverty, little empirical evidence exists about the effects of poverty programs on the environment. I estimated the effect of substantial and persistent income transfers to poor households in Indonesia on deforestation. To control for non-random selection of administrative areas receiving the CCT, I combine semi-parametric matching methods, which control for observable pre-treatment confounding characteristics, with a difference-in-differences (DID) design that uses a fixed effect, panel data regression estimator to control for unobservable but time-invariant confounders. This study combines administrative data from the transfers program and the Indonesian government, remote sensing data from satellites, and a deep understanding of how the CCT was scaled up across villages and over time. I found that exposure to the CCT decreases annual forest cover loss in a village by an estimated 20\%, on average. Thus, in Indonesia, efforts to reduce poverty can also yield environmental co-benefits. In the second chapter, I examine how a concerted effort to reduce deforestation in a developing country can be successful. Whether countries with tropical forests can innovate in the policy domain to substantially reduce deforestation is an open question. I examine this question empirically by analyzing Brazil\u27s much-lauded effort to fight deforestation. In 2004, Brazil was ranked the third-largest emitter of carbon emissions, driven largely by deforestation in the Amazon basin, the world’s largest rainforest and home to about 10\% of the world’s biodiversity. I seek to estimate the effect of a large anti-deforestation program, the Action Plan for the Prevention and Control of Deforestation (PPCDAm), on deforestation in the Brazilian Legal Amazon (BLA) by combining a synthetic control design and newly available, high-resolution satellite panel data on global forest cover change. From 2005-2009, the PPCDAm successfully avoided deforestation of by 88,841 km2. This avoided deforestation is associated with CO2 storage of 3,521 Mt