Accepted 20th July, 2019.
Over the past few years, policymakers and researchers in Kenya have implemented different interventions to improve the efficiency of the sorghum subsector. However, marketing barriers persist, while poor market information access remains a major constraint. This study evaluated the integration of sorghum markets in Kenya and tests for causality using bi-monthly data from five markets for the 2011 to 2017 period. Johansen cointegration approach and Granger causality tests were employed to test for market integration and causality respectively, using data from five markets; Kalundu, Kibuye, Homabay Town, Soko Posta and Gikomba as the reference market. The cointegration results revealed that there is existence of a long run relationship between Gikomba and all the other markets. This implied that sorghum prices converge towards equilibrium in the long run. Granger causality tests revealed presence of three unidirectional, one bidirectional and three independent price relationships. The independent relationship implied marketing inefficiencies. The study recommends the strengthening of market information systems as this will enhance market integration.
Keywords: Market integration, Unit roots, Johansen cointegration, Granger causality.