Cost avoidance is not the same as cost savings

When talking about the government, some people love to talk about cost savings and efficiencies. People in government procurement talk about cost savings and efficiencies occasionally but, much more often, people in government procurement talk about a metric called "cost avoidance," which is not the same thing as cost savings. Cost savings is an intuitive concept: if the government is spending less money by doing X instead of Y, there is a cost savings. Cost savings happens when you used to spend $1 million per year, and now you're paying $800,000 per year.

Cost avoidance, on the other hand, is a metric that describes the difference between what was paid for something and what could have been paid for that thing, otherwise. For example, if I paid $10 for a book at a bookstore, but I saw the same book at a different bookstore for $12, I didn't save any money; I still paid $10. But I avoided a $2 cost, or a 20% cost avoidance.

You might be able to guess that, like other metrics that can be gamed, cost avoidance is a bit of a slippery concept. For example, suppose a vendor quotes you at $1.2 million and you negotiate them down to $1 million. Congratulations! That's a $200,000 cost avoidance.

And, apparently, when it comes to the Transportation Security Administration's Screening Partnership Program, which provides security screening services at commercial airports, cost avoidance is the law of the land.

Seriously! According to a recent GAO protest, federal law requires that vendors must do two things to win a contract for security screening:

  1. "[P]rovide its employees with compensation and other benefits that are no less than the level of compensation and benefits that would be provided to federal employees."
  2. Have a price "equal to or less than the cost to the agency."

These things might seem in tension. After all, how can you pay your employees more and still cost less? This question was at the heart of American Eagle Protection Services Corporation's protest to the GAO.

Before you get carried away, TSA indicated one "straightforward" way to pay your employees more than the federal rate and still cost less:

TSA responds that the statute does not require private screeners to be compensated identically to their federal counterparts based on seniority; rather, TSA explains that the statute only requires that private screeners be compensated at a level not less than the minimum level provided to federal screeners.

In other words, you don't actually have to pay screeners more than they would make if they worked in the federal government. You just have to pay screeners more than the theoretically lowest paid federal screener. Cost avoidance in action!

GAO agreed with TSA. Sorry, screeners.

But the protest didn't end there. Namely, American Eagle (also known as AEPS) argued that TSA's solicitation "imposes excessive risk on the selected contractor" because it didn't share enough information about how to actually manage its costs:

AEPS contends that the RFP imposes excessive risk on the contractor. First, AEPS argues that the RFP does not reasonably limit the amount of work that the contractor may be expected to perform at a fixed price. Second, AEPS argues that the RFP does not provide adequate and meaningful historical data that allows offerors to factor risk into their proposed pricing. Third, AEPS argues that the FCE operates as a price ceiling, which prevents offerors from effectively capturing risk as part of their proposed pricing. Finally, AEPS argues that the cumulative impact of these factors creates an unreasonable risk for offerors.

TSA and, ultimately, GAO disagreed. In the government's telling, "the RFP contains discrete parameters that limit the scope of work. To illustrate, the RFP provides the total number of security lanes that the contractor may operate, the requirements when such lanes must be opened, and the airport's hours of operation. By providing this information, the RFP provides an upper and lower bound to the amount of labor that must be provided."[1]

And, it turns out, because of the statutory constraint that the government only have cost avoidance, the vendor's upper bound of risk only needs to be lower than the government's upper bound of cost:

While we recognize that the FCE imposes a cap on all costs, including the ability of offerors to factor in risk of unanticipated labor costs, we do not think this creates an unreasonable risk for offerors. Instead, as the agency points out, the FCE effectively disqualifies offerors if their risk tolerance indicates that the proposed price should be higher than the government’s cost of providing the services through federal employees. Although the protester may complain that this position is harsh, it is well settled that there is no right to a government contract.

Undoubtedly, the government might have more information than industry about volatility and the distribution of passenger volume. But the government isn't really trying to save on costs here. It's only trying to avoid them.

[1] A few years ago, while traveling for a wedding in Tuscany, I read a graduate-level textbook on "Queueing Theory."[1] At the time, I believed that queueing theory was basically everywhere and that many difficult problems could be effectively framed as queuing problems. To a certain degree, I continue to believe that! Queuing theory is a surprisingly useful thing to understand. I know what you're thinking. Does this guy know how to party or what? But also, after reading this book and the protest, I'm somewhat sympathetic to American Eagle's complaint that TSA didn't provide daily passenger counts.

Subscribe to GovContrActually

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.