Policy I

Land conversion to agriculture is a defining environmental challenge for tropical regions. We construct a novel panel of land-use change on the properties of municipal politicians and campaign donors in the Brazilian Amazon to assess three channels through which local politics may drive land conversion: (i) leaders’ self-interest, (ii) patronage, and (iii) interest group influence. Estimating event studies around close mayoral elections, we find that winning candidates – and their campaign donors – increase soy cultivation while the candidate is in office, suggesting political connections help landholders overcome barriers to adoption for this high-value crop. At the municipal-level, close election of a mayor with personal landholdings has no effect on land-use or environmental outcomes, but election of a mayor who received campaign donations from landholders increases soy cultivation, deforestation, and environmental violations. Results provide nuanced evidence for each channel of political influence, with implications for the design of conservation policies.

We study the role of subnational institutions in forest conservation in a context in which areas near roads are prone to deforestation. We develop an index of institutionalism to examine the extent to which local institutions can contribute to mitigate the road infrastructure’s adverse effect on deforestation. Using a large dataset from Peru, home to the second largest portion of the Amazon rainforest, we find that a higher value of our index of local institutions is significantly correlated with lower deforestation. However, the effect of our institutions index is not sufficiently large to offset the deforesting effect that closeness to roads has, at least not for relatively short distances to road. These results are robust to different specifications of our institutions index and to the inclusion of a large set of control variables.

In this article, I estimate the treatment effects of having been hit by one or multiple hurricanes, earthquakes, or both on government revenues and expenditures at the municipal level. In the estimations, I employ modern Difference-In-Differences (DID) estimators using 21 years of panel data. Overall, my findings suggest that natural disasters change the composition of revenues and the spending priorities of municipal governments. On the revenues side, natural disasters have a statistically significant change in the composition of federal transfers. Municipalities do not get more transfers, but they get more discretionary funds. On the expenditures side, hurricanes decrease municipal spending on services and assistance, while earthquakes decrease the spending on services. In the absence of additional economic stabilization policies, these changes in the spending priorities imply that natural disasters could be having a large negative effect on the lives of the affected populations in developing countries