Dropping degree requirements from federal contracts opens the door to AI-enabled services

Congress recently sent to the president’s desk the Skills-Based Federal Contracting Act, a bill that stops agencies from writing minimum education requirements into contract solicitations unless they justify them in writing, directing them to assess skills instead. 

The bill has been described as a jobs measure. It is also a change to what the government specifies when it buys services. As an economist who builds AI products, I think that change matters more than the jobs.

When the federal government buys services, the contract usually describes the people who will do the work. Many labor categories name a degree and years of experience, and the contractor bills for those people by the hour. This gives the buyer some assurance about quality, and it sets a floor under the price, because no one can deliver the service for less than it costs to employ people who meet the specification.

The act removes the degree from that list. A degree requirement is also, in effect, a requirement that a certain kind of person does the work. Without it, the government is closer to buying the work itself, and that work can now be done with far fewer people. A small team using AI tools, with someone accountable for the output, can often meet the same standard at a much lower cost. When that kind of delivery is allowed, the floor under the price drops, and it drops a long way.

This small change could have large unintended consequences.

First, the case for paying by the hour gets weaker. If a firm using AI can produce in two hours what used to take 10, a contract that pays for hours rewards the slower method. Part of what justified the rate for each labor category was the degree reinforcing it. Take the degree away and the rate is harder to defend. An agency that wants the lower price will need to buy a defined piece of work at a fixed price, and say clearly how it will judge whether the work is good enough. The act does not require agencies to do this, but it makes that easier.

Second, the act changes the terms of competition. Many large incumbents are built around benches of junior staff billed by the hour, and degree requirements suited that model. With them gone, evaluations can give more weight to how the work will be done and what it will cost, which opens space for firms whose delivery relies on AI and small teams. Whether those firms win depends on the other barriers they face, including past performance and clearances, and on how quickly the large firms change their own delivery models.

Third, the effects reach beyond the United States. Other governments have already moved further on credentials. The United Kingdom and Australia price much of their digital and professional services work against skills frameworks, with little reference to degrees. But they still buy people by the day or the hour, through panels and rate cards built around roles and seniority.

It is that habit of paying for time that AI makes harder to defend, and the degree was only one part of it. The United States is one of the world’s largest buyers of services, so other governments pay attention to how it buys. If U.S. agencies show they can buy a defined piece of work for less, at a quality they are satisfied with, governments that have already dropped the degree will have to decide whether to keep paying for time. Firms that sell time to those governments will have to rethink how they price their work.

The act was written to open federal work to people without degrees. Its larger effect may be to open it to firms that use AI to do the same work with far fewer people.

Laurie Nicol is an economist, founder of Tendl, an AI company that assists companies in bidding, and author of The Venture Economist.