Testing for Food market integration: A study of the Vietnamese paddy market

With its increasing integration into the world economy, agricultural exports and rural incomes in Vietnam have increased substantially in recent years. At the sub-national level, however, there are concerns that not all regions and categories of agricultural producers have and will benefit from the ongoing liberalization of agricultural markets. Vietnam's elongated geography and lack of spatial market integration pose special problems in this regard. Accordingly, this study aims to answer three interrelated questions: (a) whether there is spatial integration between paddy markets in the North and South of Vietnam; (b) whether there is spatial integration in paddy markets within the North and within the South; and, (c) if within-region integration is stronger and faster than between-region integration. The empirical model we develop to answer these questions, uses estimates of transfer costs to generalize the well known model of spatial market integration due to Ravallion to allow for the possibility of threshold effects. A sequential testing strategy is developed which progressively tests for market segmentation, the number of thresholds, long-run market integration, common dynamics/informational efficiency, and (a strict version of) the 'Law' of One Price within an error-correction framework. When the unrestricted version of this model is estimated using monthly paddy prices for eight markets between 1993 and 2006, we find weak evidence of market integration between paddy markets in the North and South of Vietnam with an absence of threshold effects. However, there is evidence of both threshold effects and stronger forms of spatial market integration for paddy markets within the North and within the South, with at least 60% percent of price changes being transmitted between markets within one month whenever price spreads exceeds their upper or lower thresholds. The extent and speed of price transmission within regional paddy markets is generally faster in the South than the North of Vietnam. However, the instantaneous version of the 'Law' of One Price, which requires full price adjustment to occur within a month, only holds for a few regimes and market pairs. Three main policy implications flow from these results. First, since there is limited evidence of integration between paddy markets in the North and South of Vietnam, national level policies cannot be relied upon to stabilize or support paddy prices.Second, since there is evidence of spatial market integration within the Red River and Mekong River deltas, paddy markets within these regions can be relied upon to transmit price signals between deficit and surplus areas relatively well. Third, since the speed and extent of price transmission is relatively rapid within the North and within the South of Vietnam, the private sector trade can be relied upon to transfer rice and paddy between markets in an efficient manner. Problems might, however, emerge if large demand-supply imbalances were to emerge between the North and South, as transfer costs would prevent private sector trade taking place. In these circumstances, the public sector might need to intervene, in a consistent and market friendly way, to ensure adequate food supplies in the short-term.