Economic Theory and Econometric Methods in Spatial Market Integration Analysis

Abstract

The study of Spatial Market Integration (SMI) has been of great concern for agricultural economists, with the Takayama and Judge Price and Allocation Model (TJM), in which prices are bounded by the Spatial Equilibrium Condition (SEC), being the core economic theory. The SEC implies that no profits are made from trading goods among spatially separated regions and is a weak form of the Law of One Price (LOP). Indeed, the LOP is regarded as perfect market integration and a perfect equilibrium. Another theoretical concept in SMI is market efficiency, which can be understood as markets being cleared, hence there is an optimum allocation of the resources, which leads to the correct pricing of the goods. In theory, when trade occurs among regions, the excess supply and demand signals are transferred to the prices of the goods among trading regions, so that prices move together among the regions; nevertheless it is important to point out that price co-movement does not necessarily lead to Pareto efficiency, so that co-movement can occur for markets in disequilibrium. Following such theoretical foundations, most of the empirical research in the field of SMI deals with prices. The early work done in the field has dealt with price correlations and regressions; it often provided weak support in favour of the LOP. Later, with the development of the concept of cointegration, new econometric techniques provided support in favour of the LOP; however, such methods suffer from neglecting the role of the SEC by depicting the equilibrium as a linear relationship such as the LOP. Such a limitation served to support the use of non-linear techniques, namely the regime-dependant Threshold Vector Error Correction Model (TVECM), on which the inner regime, often referred to as the neutral band, is analogous to the SEC. The apparent compatibility between the neutral band and the SEC has served to popularize the use of the TVECM in empirical SMI research. Nonetheless, economic theory still suffers from an unclear definition of market integration and little attention is given to its foundations and implications in the price formation process. Following this concern one can question to which extent the TVECM is the correct instrument when little attention has been paid to economic theory. With the following thesis the author aims to contrast economic theory and econometric techniques used in Spatial Market Integration in order to evaluate whether or not the TVECM is the correct specification for SMI analysis as is it claimed or assumed in the literature. Chapter One introduces the Takayama and Judge Price and Allocation Model (TJM), which serves as the theory for Spatial Market Integration. It also introduces the TVECM and the standard econometric technique used on its estimation. Then, using the TJM, artificial prices are generated. For the simulations the true parameters are known, hence if the TVECM is the correct specification, the estimated parameters from the TVECM have to be unbiased with respect to the true parameters. Chapter Two starts off with an introduction to the econometric concept of cointegration and the testing procedures of linear cointegration and Threshold Error Correction. The main aim is to test whether the data which is economically integrated in equilibrium serves to econometrically test for threshold error correction. Chapter Three addresses the incompatibilities found in Chapter One and Two between pure equilibrium data and the TVECM. Following such concern some modifications to the original Takayama and Judge Price and Allocation Model are proposed in order to obtain prices beyond the SEC. The rationale of the processes which violate the SEC is based on economic theory. Following the same procedure as in Chapter One, the new models are used to generate prices (Monte Carlo simulations) for which the true threshold value is known. Then those prices are used to estimate the threshold parameter(s) under the TVECM. The last Chapter is a summary of the major findings regarding the compatibility between the economic theory and the econometric methods. The purpose is to point out the importance of improving vague and ambiguous definitions in economic and econometric theory; for that plausible alternatives are reviewed. Along with the lack of sound theory is the fact that empirical applications often do not support theory; this is exemplified with two studies conducted in Mexican and US maize markets. Nowadays Spatial Market Integration analysis has a main role in research and policy making, thus the people conducting such analyses have to be more aware of the theoretical implications in order to draw the conclusions of their empirical work properly.