The Mystery of Zero-Ending Contracts

When researchers pored over U.S. government procurement contracts as part of a study on lobbying, they noticed something strange: an inordinate number of contracts for goods and services seemed to end with a series of zeros — many with five zeros or more. 

Rounded contracts showed up so frequently, in fact, that the researchers figured there was a larger story that was being overlooked, some dynamic that could explain the seeming lack of rigour in pricing such large purchases and projects. They decided to dig deep and find out what that was. 

The researchers considered several possible explanations. Yes, people tend to find rounded numbers appealing and easier to process cognitively (notwithstanding the penchant for marketers to end their prices with 99 cents). Would a bid for government business appear more attractive if it ended in a series of zeros?

Alternatively, do so many bidders use unsophisticated budgeting practices or rely on executive intuition that would lead to such tidy estimates? 

Yet another potential explanation — and one the researchers suspected from the start — is that bidders prefer to make round-number submissions when they are unsure what they are bidding on. In other words, are solicitations that lack sufficient detail or are confusing themselves to blame? 

In a study called “Close Enough! Exploring the Consequences and Motivations Behind Estimates in Capital Budgeting”, the researchers looked for an answer. They took a close look at the preponderance of zero-ending government procurement contracts and examined how well these perform compared to contracts that appear more precisely costed out. 

Government procurement may seem to be a dry issue but there is plenty at stake. Governments in Canada spend almost $200 billion per year on just about any type of good and service supplied by Canadian and foreign businesses. The U.S. federal government alone spends more than US$650 billion. Any improvement to the procurement process would pay significant dividends to public sector agencies and taxpayers. 

How was the study designed? 

The study relied on data from USAspending.com on contracts entered by the U.S. government and firms for goods or services. The data sample covers 2004 to 2017 and includes nearly three million contracts valued at greater than US$10,000. 

What did the study find? 

• Procurement contracts with bids that end in a series of zeros are over-represented among all contractor bids accepted by the U.S. federal government. At least 35 per cent of bids end in at least one zero. A little more than three per cent of all accepted contracts end in four zeros. 

• While controlling for firm, government agency, year and other factors, bids ending in a series of zeros are far more likely to require revision and renegotiation and less likely to be renewed. 

• Research and development contracts are significantly more likely to end in a series of zeros and more likely to require renegotiation. In contrast, contracts that result from a standardized system of solicitation (known as Simplified Acquisition Procedures) are far less likely to end in zeros and show no significant evidence of underperformance. 

• The prevalence of zero-ending contracts is likely due to ambiguity in how contracts are solicited rather than to the size of the contract and the characteristics of bidding firms or government departments. 

What do I need to know? 

The high cost of economic policy uncertainty, arising from recessions, pandemics, poisoned politics and confusing tax regimes is well documented. This study of U.S. government procurement is a variation on the same theme. 

Uncertainty in how goods and services are solicited by the federal government — in the form of ambiguous or loose directions or incomplete information — has significant downstream effects. Contracts can be poorly costed out based on intuition rather than rigour and require revisions and renegotiation.

This is not a story about contractors wilfully flying by the seat of their pants. The researchers found that businesses bidding on the same product category within a given year sometimes submitted round-number bids and other times did not. “This observation suggests that our results are not driven by unsophisticated contractors but instead by contracts that are more difficult to price or lacking key information,” the researchers write in their working paper. 

To illustrate, the researchers offer two contrasting examples. A solicitation for forklifts includes details on minimum and maximum net horsepower, travel speed, electronic shift control, adjustable volume back-up alarms, cup holder size and other requirements. By contrast, the solicitation for research services weighs in at 24 pages but is vague on scale and scope. 

The phenomenon of individuals providing round values when they lack sufficient information has been seen elsewhere. One study from 2020, for example, found that more than one-third of the CEOs in its sample received round-number compensation (that is, evenly divisible by 100,000 or 10,000). This most commonly arose when board members provided weak oversight of compensation and thus faced more uncertainty in estimating executive pay. 

The implication from this study is that governments would do themselves — and bidding firms — a favour by developing more detailed solicitations for goods and services upfront, not only for forklifts but for R&D-related projects. That would result in bidders having the confidence to rely less on executive intuition or other cognitive shortcuts and more on data analytic tools and budgeting systems — as well as fewer people needed to monitor contracts and clean up the mess. 

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

Authors: Matthew Boland (Sobey School of Business) and Lynnette Purda (Smith School of Business) 

Published: Working paper available for download