Predictive mapping of wholesale grain prices for rural areas in Tanzania : A replicable modeling framework for predicting agricultural prices across time and space
Our understanding of small farm decision-making in developing countries is often critically constrained by sparse information about the input and output prices faced by farmers operating diverse landscapes with heterogeneous market and accessibility characteristics. We present a methodology for pred...
| Autores principales: | , , , , , |
|---|---|
| Formato: | Informe técnico |
| Lenguaje: | Inglés |
| Publicado: |
EiA
2024
|
| Materias: | |
| Acceso en línea: | https://hdl.handle.net/10568/162729 |
| Sumario: | Our understanding of small farm decision-making in developing countries is often critically constrained by sparse information about the input and output prices faced by farmers operating diverse landscapes with heterogeneous market and accessibility characteristics. We present a methodology for predicting local market prices over time and space, using relatively sparse pooled observations on crop commodity market prices at different locations and times. We show prediction results for wholesale prices for grains (six different cereals and beans) and potatoes in Tanzania. We find that pooling observations on prices for different commodities improves prediction for any given commodity, because of spatiotemporal covariance in observed prices. We discuss how our modeling framework could be used to design relatively low-cost monitoring systems for enabling regularly updated, national-scale spatial price maps. |
|---|