Submissions Accepted for Presentation at the World Bank Land Conference 2024

Investments in transportation infrastructure can impact the environment beyond their immediate surroundings. We build an inter-regional trade model to estimate the general equilibrium effects of changes in infrastructure on deforestation. Using panel data on the evolution of the transportation network from Brazil and land use data in the Amazon, we estimate the model and find sizable effects of infrastructure on deforestation. Model simulations show that ignoring general equilibrium underestimates deforestation impacts by one quarter. We also show that this model can be used to evaluate the deforestation induced by individual projects, an essential input for public policies.

How will future climate change affect rural economies like sub-Saharan Africa (SSA) in terms of migration and welfare losses? How can policy enhance SSA's capacity to adapt to this process? I answer these questions with a quantitative framework that, coupled with rich spatial data and forecasts for the future, estimates millions of climate migrants and sizeable and unequal welfare losses in SSA. Investigating migration and trade policies as mitigating tools, I find a trade-off associated with the former: reducing SSA migration barriers eliminates aggregate welfare losses at the cost of more climate migration and high regional inequality. Reducing tariffs attenuates this cost.

This paper studies the e ect of weather shocks on rural land sales and the farm size distribution. Using a unique administrative dataset with transaction-level information and a land registry covering most of Colombia's farmland, we show that extreme temperature events increase the frequency of land sales and decrease the average farm size within municipalities. These results are driven by small farms being subdivided and purchased by previously landless owners, with no evidence of weather shocks leading to the consolidation of small farms into larger holdings. The effects of extreme temperature on land sales are stronger in poorer and more isolated municipalities, where landowners are also less likely to take out land mortgages after a shock. To explain these patterns and explore how they can be exacerbated by underdevelopment, we develop an intertemporal, two-sector model where agents face a subsistence consumption constraint.

We experimentally evaluate whether an interest-based negotiation (IBN) training for community leaders in Liberia improves their ability to strike beneficial deals related to their land and forests. We use environmental assessments, lab-in-the-field, and surveys and find that trainees are 27% more likely to reach a beneficial agreement, and when they conclude deals, their payoffs are 37% larger. Our exploration of mechanisms indicates that the training increases trainees' capacity to identify valuable deals, but does not improve their appraisal of their outside option. We find a reduction (0.27 standard deviations) in the exploitation of communal forestland in treated communities.