Close Enough! Exploring the Consequences and Motivations Behind Estimates in Capital Budgeting

Abstract

Analytic tools used for decision support require high quality data inputs in order to arrive at precise estimates. Using an extensive database from US government procurement contracts, we document that an over-representation of accepted contract bids end in a series of zeros. These ending-zero contracts show systematic differences in contract outcomes. They require more extensive revision and are less likely to have embedded options exercised, indicating weaker contract performance and requiring additional resources for renegotiation. We find support for the presence of ending zeros being associated with uncertainty in the contract solicitation. Ending zero contracts are more frequent for certain services, particularly research and development, and less frequent when the contract refers to standardized goods or services. We explore whether the presence of ending-zero bids and their associated weak outcomes is driven by small contracts, potentially unsophisticated bidding firms, or particular government agencies. Our results suggest that the specificity of the contract solicitation itself and the information it provides may be a driving factor and that government allocation of resources could improve with additional detail provided in contract solicitations.

Keywords: Capital Budgeting, Contract Incentives, Analytics, Decision Support, Procurement

JEL Classification: G31, G40, H57