Levels of Asset Specificity in Relational Contracting

Abstract

Standard property rights theory (whether static or dynamic) assumes assets are specific, but once this assumption is in place, the level of asset specificity has no bearing on the make-or-buy decision. While there are good reasons to doubt the universality of transaction cost economics' prediction that the more specific the asset, the more likely is vertical integration to be optimal, this is an issue that cannot be addressed within the existing property rights framework. In this paper, I provide a simple and straightforward way of incorporating these different levels of asset specificity in a relational-contracting property-rights model of the firm, and show that they matter even in a static model. Using the Outside Option principle in bargaining results in a model in which the integration choice is affected in a non-trivial way by realized asset specificity. For instance, I find that vertical integration is the efficient response to widely varying supply prices as long as the level of asset specificity is low enough. When asset specificity is high, on the other hand, the incentives of the upstream party are best aligned with those of the downstream party and incentives may be best provided under outsourcing. In general, under bargaining with outside options, the first best is easier (harder) to achieve with outsourcing than with integration when the degree of asset specificity is high (low).