Delineating Spatial Markets Using Multivariate Time Series

Recent advances in multivariate time series modeling techniques have produced statistical concepts and methods that may surpass many of the economic and antitrust market delineation procedures currently in use. The appropriateness of cointegration tests and vector error correction models for market delineation are developed in the context of a spatially linked oligopsony and then employed to determine whether a single economic market exists among spatially linked North American pork producers. Institutional detail and other time series information are also used to explore antitrust market issues.