Nontraded food commodity spatial price transmission: evidence from the Niger millet market

Using an Augmented Factor Vector Autoregressive (FAVAR) Model, this study analyzes spatial millet prices transmission in Niger. Our results did not find condition for millet markets integration existence. However, the Granger causality tests and impulse response functions from the estimated short-term dynamic as FAVAR model revealed the existence of leading markets whose millet prices affect a maximum number of other regional millet prices, while some regions seem to be isolated from trade or information flows. Furthermore, the significance of a shock depends also on the characteristics of the region where it originates in terms of millet demand or supply, indicating that the region to target and where the price shock originated matter for the policies' success.