Diana Angelis

Applying Insights from Transaction Cost Economics to Improve Cost Estimates for Public Sector Purchases: The Case of U.S. Military Acquisition

ABSTRACT This article uses Transaction Cost Economics (TCE) to help characterize, explain, and ultimately reduce the cost growth that plagues many of todays major investments in military capabilities. There is mounting evidence of a systematic bias in initial cost estimates of new weapon systems purchased by the U.S. military. Unrealistically low cost estimates result in cost overruns. Fixing cost overruns can substantially impact public budgets and military readiness. Cost estimates serve a dual function: first, as an integral part of the decision-making process to evaluate military purchases/investments, and second, as a baseline for future defense budgets. In the first case, underestimating costs can result in too many new weapon program starts and excessive investments in those systems. In the second case, unrealistically low cost estimates result in overly optimistic budgets. Budgets planned on the basis of optimistic cost estimates create the illusion of more resources available than actually exist. Two factors are often blamed for unrealistically low cost estimates: bad incentives (psychological and political-economic explanations), and bad forecasts (methodological explanations). While briefly exploring the former, the focus of this study is on cost estimating methodology. Conventional public cost estimating techniques focus on the production costs of public purchases (input costs, learning curves, economies of scale and scope, etc.). The goal of this article is to improve cost estimates by expanding conventional cost estimating methodology to include TCE considerations. The primary insight of TCE is that correctly forecasting economic production costs of government purchases or acquisitions is necessary, but not sufficient. TCE emphasizes another set of costs—coordination and motivation costs (search and information costs; decision, contracting, and incentive costs; measurement, monitoring, and enforcement costs, etc.). This study encourages public officials and cost analysts to capture these costs and to understand key characteristics of public-private transactions (uncertainty, complexity, frequency, asset specificity, and market contestability) to generate more complete and reliable cost estimates and improve public sector purchases.

Does computer-based training impact maintenance costs and actions? An empirical analysis of the US Navy’s AN/SQQ-89(v) sonar system

The United States Navy decided in the early 2000s to replace traditional, instructor-led schoolhouse training with computer-based training (CBT). While employing CBT may produce gains in knowledge acquisition and lower costs for repetitive, low-skill work, there is a lack of empirical evidence whether these benefits exist for more highly skilled Navy operations. Anecdotal evidence suggests that CBT failed to sufficiently prepare new sailors for sophisticated systems’ maintenance and operation. To determine the validity of this evidence, we examine how CBT has affected the AN/SQQ-89(v) sonar. We empirically analyse whether the Navy’s introduction of CBT significantly altered fleet maintenance costs, actions and training requirements, by assembling a unique data set of ships, locations, personnel, maintenance costs and maintenance actions. Controlling for the Navy’s plan to man the system, the number of authorized billets and the number of personnel on board, we find that CBT adversely impacts costs, actions and maintenance hours for the sonar system.