Measuring spatial market integration between urban and rural food markets in South Africa

Abstract

This study focused on measuring spatial market integration between urban and rural food markets in South Africa. The study was influenced by continual debate on the issue of high food price increases among the basic agricultural foods which have impact on food price differentials that exists in market locations. Higher food price differentials between market locations can have negative implications for households’ livelihoods. In South Africa the majority of lower-income earners live in rural areas, which implies that the disposable incomes of those households are affected differently by higher price differentials. To ensure that there are lower price differentials between market locations, spatial market integration is required. The main objective of this study was to measure spatial market integration between urban and rural food markets. This objective was achieved through the modelling of secondary price data of the 2.5 kg package of super maize meal. Data used in the study was collected by the NAMC and Stats SA during the period of November 2006 to July 2012. The study used a co-integration model, together with the Vector Error Correction Model (VECM) and the Autoregressive Distributive Lag (ADL) model. The study revealed that out of the nine markets which were measured for spatial integration, six of the rural food markets were co-integrated with urban markets. Estimation of the six co-integrated markets with the VECM revealed a long-run equilibrium relationship between urban and rural food markets. The VECM results further showed different speeds of price adjustments to the long-run equilibrium which was faster (65% on average) in five markets and slower (40 %) in one market. The fast speed of price adjustment to long-run equilibrium relationships suggests that transaction costs have significant effects on markets linkages. Markets also showed different time of adjustments which was between three and five months. Short-run dynamic effects were found in only three markets and could not be established in the other six rural food markets. Price information flow, transportation costs and transaction costs are seen as bottlenecks that prohibit integration of markets in the short run. With regard to price relationships, the study found statistically significant differences between urban and all rural food market mean prices. This suggested that the price of 2.5 kg maize meal was generally high in all rural markets, as compared with urban markets. Markets with very high price increases were those located in the lower production potential areas of the maize commodity. The Impulse Response Function (IRF) was employed to establish the effects of negative and positive price transmission shocks from the urban to the rural food markets. One standard deviation was put into the model to establish the response of markets to shocks. The markets tested revealed a period of between two and eleven months for the negative and positive impulses to be cleared in all the markets. A high degree of spatial market integration was found when negative and positive shocks did not exhibit price diversion from the long-run equilibrium relationship.