Estimating the degree of market integration

Abstract

Existing tests of spatial market integration are commonly based on statistical criteria without an explicit link to an economic model of price determination. This article proposes new measures of market integration defined directly in terms of a well-known spatial price determination model and develops an econometric methodology for estimating these measures. Due to the intractability of the conditional density function of prices, we use indirect inference to estimate the model parameters and market integration measures. The methodology is illustrated with simulated data and is applied to soybean price data for the United States, Brazil, and the EU.