The End of Cost Allocations As We Know Them

Abstract

Cost allocations have befuddled managers around the world for over a century. The need for objective, actionable accounting data is frustrated not just by the difficult theoretical issues in distinguishing among various classes of cost, but also by informational asymmetries and agency conflicts between operating managers and corporate managers. Despite the obstacles, there are good reasons for allocating costs. Properly structured, allocations or other transfer pricing schemes can serve the same purposes as market prices, i.e., communicating useful information about the relative scarcity of resources within an organization, thereby helping to balance supply and demand for those resources. Most transfer pricing schemes, like Activity-Based Costing, attempt to solve the problem through costly refinement of standard cost measurement models. This paper offers a contrasting model that begins with the recognition that post-hoc allocations inevitably involve subjective assignment of costs to activities, and that whatever objectivity can, in principle, be brought to bear on said assignments are undermined by internal incentives. The author offers, instead, transfer prices that arise from negotiations between two parties to an internal exchange under a particular set of decision rules reinforced by a performance measurement and reward system that gives both parties positive incentives to reveal their reservation prices for the goods or services being exchanged. For multiple parties, the author has created a simple, but sophisticated transfer auction process to facilitate investments in shared facilities or services that lead, prospectively, to efficient allocations of operating or capital costs.