This is the authors' abstract. We don't add key points for this paper.
Recognition of tropical deforestation as a major source of carbon emissions has invigorated efforts to reduce deforestation in the Brazilian Amazon.Previous studies throughout the tropics have found that market access, one of the key drivers of deforestation, is positively correlated with deforestation.This relationship has been explained by the effect of market access on farm gate prices for agricultural and/or timber products.However, access to markets also means access to urban centers, which has multiple influences on household decisions.This study incorporates these multiple dimensions of "urban access" into a household production framework, showing that their combined effect on deforestation is ambiguous and could change over time.We test whether the effect of distance to cities has changed over time in a region typical of the agrarian reform settlements established by the Brazilian federal government throughout the Amazon.Specifically, we test for structural shifts in the relationship between distance to cities and deforestation by modeling deforestation over 24 years on approximately 8,800 farm properties.Estimation results for a two-part, "within-between" Random Effects model show that after controlling for forest stock and other drivers of deforestation, the relationship between distance to cities and deforestation changed from negative to positive around 2004.Further investigation suggests that the diminishing role of transportation cost in farm gate price combined with the increasing enforcement of environmental laws forced the effect of distance on deforestation from negative to positive.Regional heterogeneities underlying the general evolving relationship are consistent with observed variation in urban development.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2018) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: