Three essays on economic evaluation of responses to weather-induced risks in Uganda

Abstract

In Uganda, the past five decades have been characterized by increasing temperatures, longer dry seasons, changes in the timing of rainfall with extreme events such as floods and heavy rainstorms, all of which have adverse effects on the livelihood of the rural farming community. Several strategies have been recommended for adaptation and mitigation of negative effects arising from changing weather conditions, including migration, use of weather index insurance, and changes in farm production practices, among others. However, the usability and effectiveness of the strategies are influenced by economic, social, biophysical and farmers’ behavioral factors that are examined in the three essays of this study. Given the importance of weather and labor to rural and agricultural-based economies, the first essay examines the effect of weather anomalies on the likelihood that workers migrate from rural and urban areas. By matching household survey data with weather data, and assuming exogeneity of weather variables, the effects are identified by exploiting the spatial heterogeneity of weather conditions and worker characteristics. The results remain robust to alternative model specifications, all of which show a nonlinear effect of weather anomalies on the likelihood of migration of workers from rural areas. The results show that precipitation extremes reduce the likelihood of labor migration whereas temperature extremes increase the likelihood of labor migration. This research contributes to the burgeoning literature on weather-induced migration, and the findings underscore the need to build resilience for workers. The second essay analyzes the critical temperature for coffee yield reduction and whether the effects for single-cropped coffee farms differ from those that are intercropped with bananas as shade plants. Using panel data for coffee production and weather, I exploit the spatial and temporal variations in temperature and precipitation to estimate the effects. Estimation of random-effects regression models shows a nonlinear effect of temperature and precipitation on the yield for coffee with extreme temperatures greater than 28°C resulting in yield reductions. A sensitivity analysis predicts that increases in temperature results in reductions in yield, but the reductions are less for coffee farms that are intercropped with bananas. The findings can be used to inform policy decisions and research to design interventions that reduce production risks arising from weather changes. The third essay analyzes factors that affect the adoption and renewal of weather index-based insurance contracts. It also examines farmer preferences for attributes and types of index insurance contracts. Given that the use of index insurance is relatively new in Uganda and the market is not yet well developed, the study makes use of data collected through choice laboratory experiments conducted in simulated insurance markets in Western and Central Uganda. Discrete choice models were used to analyze the data and the results showed that the ambiguity of insurance contracts reduces the likelihood of the adoption of insurance. The results also show that farmers have a higher preference for insurance offered through farmer groups, as opposed to insurance offered to individuals. The study contributes to the literature on behavioral and product-specific factors that affect the adoption of index-based insurance.