Measuring Transaction Costs in Plural Formed Marketing Channels: An empirical investigation of franchise units in the oil industry

Abstract

Previous empirical research has supported the predictions derived from transaction cost economics that asset specificity, uncertainty, frequency and complexity entail vertical integration (David and Han 2004). The underlying assumption is that integration creates the most efficient organizational formation. Given this assumption from transaction cost theory integration caused by market failures due to asset specificity lead to efficiency. This test focus on the ability of the principal company to control opportunism and to reduce transaction costs through vertical control. Therefore, the empirical question raised here is whether dimensions of costs can be contract related. Contract related transaction costs unlike production costs are associated to the incentives defined by the contract. This problem has barely been studied in previous empirical analysis. However, the theoretical question makes it crucial to explore a homogeneous setting like a plural formed franchise system where third variables also including asset specificity can be kept relatively constant. My intention, therefore, is not to test traditional hypotheses derived from transaction cost theory, but to explore dimensions of transaction costs and to test the prediction from the theory that costs associated to the bilateral exchange are related to the incentive system given by the contract. The test reveals how transaction costs are related to aspects of the bilateral contract. The bilateral contract is dimensionalized into structural variables like centralization and formalization and a variable describing the interactive process. The empirical setting is an oil company (Shell) and its plural formed franchise system in the Norwegian gasoline market, representing standardized technology and products, and trademark specific assets equally distributed among dealers. Both dyadic and unilateral data are used to test the hypotheses. The results point out the importance of formal rules and procedures and the scope and magnitude of interactions as efficient instruments of bilateral contracting. Centralization, though, is related to both control costs and freeriding costs.