Abstract

Abstract 

Most of the empirical literature assessing the impacts of climate change on agriculture has modeled crop yields as a function of the levels or deviations in the growing-period rainfall. However, an aspect that has received little attention in the empirical literature relates to the relationship between the timing of monsoon rains and crop yields. Using a pan-India district-level panel dataset for 50 years, this article investigates two interrelated issues critical to understanding the impacts of weather-induced agricultural risks and their management.

Abstract

Background

We use monthly municipality data for the period 2000-2015 in Guatemala and monthly district data for the period 1992-2019 in Costa Rica. We define relevant catchment areas using water flows to the population centers of the administrative units. Then, we calculate the precipitation inside and outside Pas within the relevant catchment areas, and test how the frequency of floods and landslides is affected by whether rain falls inside or outside PAs. We use a two-way fixed-effects panel data model. For Guatemala, we find no robust statistically significant effects on these types of disasters.