Conventional tests for food market integration ask often misleadingly, whether prices in different locations move together.
This paper develops an alternative methodology, the parity bounds model (PBM), which uses information on transfer costs in addition to food prices to assess the efficiency of spatial arbitrage. Monte Carlo experiments using data generated by a point-space spatial price equilibrium model show the PBM to be statistically reliable. An application to Philippine rice markets demonstrates that the PBM detects efficient arbitrage when other tests do not.
