Paper info: ON THE ANALYSIS OF VERTICAL INTEGRATION an Alternative to the Fisher Body Case
ON THE ANALYSIS OF VERTICAL INTEGRATION an Alternative to the Fisher Body Case
Place of Publication
The paper was published at the 16th IMP-conference in Bath, U.K in 2000.
Does asset specificity explain vertical integration? The debate among transaction costtheorists draws on a case more than seventy years old. The present paper offers a moderncase that contradicts the strength of the relationship between asset specificity and verticalintegration. Instead, it points at the shortcomings of one version of transaction cost theory anddiscusses reasons why a co-operative relationship can be superior to vertical integration.One of the issues often debated within transaction-cost analysis concerns the reasons forvertical integration. When those of us who are not transaction-cost theorists compareinterpretations based on that theory with arguments derived from exchange, resourcedependence,or network theories, we tend to disregard that there is not one and one view onlyamong transaction-cost theorists themselves. When we discuss whether firms incorporatecritical functions through vertical integration, instead of organising them through various cooperativerelationships with other firms, if market transactions are deemed unfeasible, thenthe transaction-cost alternative of explanation is predominantly based on Williamson´s works(Högberg 1998; Högberg 1999).In 1987 there was a conference during which transaction-cost economists discussed, amongother things, the reasons for vertical integration. The proceedings from that conference werepublished in Williamson and Winter (1993). The conference proceedings capture the essentialdifferences between two groups of transaction-cost economists when it comes to theexplanation of vertical integration.Transaction costs can be defined as the costs of organising resources across markets(Demsetz 1993, p. 161). They pertain to the negotiating of and the securing compliance withthe conditions under which transactions are carried out between firms. All other things beingequal, the higher the transaction cost, the greater the impetus to remove the transaction fromthe market. This can be done by removing it entirely from the market and performing it inhouse - the case of vertical integration. There are also intermediate forms such as long-termcontracts of a more or less incomplete nature. This means that not all issues are regulatedonce and for all in the contract but kept open in the sense that the contract specifies theconditions under which renegotiations can be triggered as well as the procedures for suchrenegotiations.Williamson extended classical transaction-cost theory. His contributions were to have greatinfluence on how transaction costs are defined and how they can be linked to forms ofregulation. However, two important building blocks in his theoretical fundament have causedcontroversy when it comes to their role in explaining vertical integration. Asset specificitywas introduced as a (the) determinant of transaction costs. The specificity of an asset dependson the range of alternative ways of using that particular asset. The difference between thepresent and the next best use of an asset owned by one of two transacting firms represents thevalue that can potentially be appropriated by the other one and, hence, a transaction cost (cf.Klein, Crawford and Alchian 1978). The other dimension is the behavioural assumption ofopportunism, ?a deep condition of self-interest seeking that contemplates guile? (Williamson1993, p 92). For example, asset specificity could make it possible for a supplier to act in anopportunistic manner that under certain circumstances could not be checked effectivelythrough long-term contracts. Instead, the buying firm would, provided it has the resources,resort to vertical integration by acquiring the supplier or by setting up its own production ofthe parts in question. What matters in the extension of the classical analysis, is not so muchthe cost of negotiating a contract as the (potential) cost of enforcing the contract and,especially, the cost of its failure. That cost is defined as a function of asset specificity andopportunism.Coase, and other ?classical? transaction-cost theorists, criticised what they considered to bean over-simplified analysis of the issue. They did not subscribe to the view that assetspecificity plays such an important role for determining whether a firm would integratevertically or not. Demsetz (1993, p 166) stated that he was ?not so sure that it costs (much)more to detail the terms of a contract when asset specificity is involved than when it is not....Truth is, it is not a predictably significant variation in transaction cost that motivates thevertical integration solution offered by these authors (primarily Klein and Williamson ? myremark). It is the presumption that losses are greater if an agreement fails when assetspecificity is involved than when it is not?. Evidently, Demsetz did not subscribe to theextended definition of transaction costs made by Williamson and others. Coase (1993)admitted that there would be a risk for opportunistic behaviour as a result of specificinvestments, but concluded that he had been more concerned about it than the businessmenhe interviewed already during the thirties. He had found that the implementation of long-termcontracts is commonly accompanied by informal arrangements not governed by contracts andconcluded that the propensity for opportunistic behaviour is usually effectively checked bythe risk of losing a business or the business reputation of the opportunistic party. Coase(1993, p. 72) stated: ?Vertical integration is an important subject but as I said earlier I thinkwe will be able to understand it better if we treat it within the context of a morecomprehensive theory?.