As the federal government moves to become more technologically advanced and automate certain kinds of work, potentially hundreds of thousands of federal employees could soon require reskilling, according to William Eggers, executive director of the Deloitte Center of Government Insights.

“I was over at [the Office of Management and Budget] the other day, and their belief is that they’re going to need to reskill upwards of 300,000 federal employees or more over the next few years alone,” Eggers told attendees Sept 11. at a Professional Services Council event.

Part of that reskilling comes from a push toward automation, which could eliminate thousands of hours of federal work, according to Eggers.

“It turns out that federal employees spend about 4.3 billion hours doing a variety of tasks,” said Eggers, explaining that feds spend the most time documenting and recording information, assisting and caring for others, communicating with supervisors and monitoring and controlling resources.

“If you look at a lot of that, you’re talking about hundreds of millions of hours of kind of paperwork processing. Four out of five of the most labor-intensive activities now done by the federal government have a pretty high level of automation potential.”

If the federal government upped its productivity by 200 percent, it would free up 1.1 billion hours at a value of $37 billion, according to Eggers.

Shifting from “low-value” to “high-value” work is one of the components of the President’s Management Agenda and would include the development of automation software to address administrative requirements faced by agencies.

“You are also in some ways fundamentally changing the process of how work gets done,” said Jenn Gustetic, visiting fellow at the Harvard Kennedy School of Government, at the same event.

But Eggers warned that a traditional change management approach would not be the best strategy for addressing this newly freed-up time.

“What we need to do is move from thinking of jobs as activities to jobs as a series of behaviors of solving different problems. What kinds of problems are we trying to solve, and what are the ideal human to machine parings? Otherwise, if it’s just activities, we’ll automate away a bunch of stuff and we won’t optimize what that work is,” said Eggers.

However, Joseph Klimavicz, chief information officer at the Department of Justice, said that the government’s traditional form of training for federal employees — putting them through an online training course — wasn’t going to cut it.

“What we need is something that people want to do that is keeping their skills current,” said Kilmavicz. “The other thing is that if somebody wants to go get an advanced degree, I’m willing to pay for it.”

Max Everett, CIO at the Department of Energy, said that the government needs to think about other kinds of learning, such as mentorships and trading the workforce in and out of the private sector.

“Most of us need a great deal more help than we think of to learn different things. And so we’re going to have to have different styles of that,” said Everett.