Still workin': Job training has a golden population the money can’t reach.

A version of this piece first published on November 4, 2025.

American workforce programs are built in ways that make it hardest to help a population that providers say is the best investment for these programs. The conversation among business and political leaders is shifting away from them, even if the fixes needed aren’t dramatic.

There are two populations in which political and business leaders most want to invest workforce dollars. One is young people—especially teenagers—who aren’t exposed to enough job types and skills opportunities. The other is the population I wrote about last year, the swaths of college-educated “knowledge” or other white-collar workers whose fields seem likely to be replaced by artificial intelligence.

Yet, increasingly left out of these conversations is a population that some workforce providers think is the best bet in workforce development, which I have taken to calling the “golden population” for jobs programs because of how providers describe their return on investment.

This population is a little older than the age range we call “youth” in workforce development, typically starting around age 25. These workers have some or no college, meaning on paper, employers might view them as having fewer “skills.” Many work in fields like retail, the food service industry, and entry-level roles that don’t require a college degree in sectors like healthcare and logistics. They usually aren’t making a living wage, but they may be closer to it than other workers.

Why do providers rave about them? If they have managed to get these workers into a jobs program, the workers have plenty of motivation to succeed and skills to do it from their purportedly unskilled employment.

They also have a degree of clarity in what they want to do that is different from the “baggage” employers dislike in recent college graduates. The golden population workers are a few years removed from K-12 education, meaning they have had time to forget the bad lessons they might have gotten from it, including signals about their capability and value that have more to do with rigid approaches to education than their ability to perform. They also have seen that there is more to the world than the types of jobs associated with college, or they may have encountered people in industrial or trades jobs that are making the kind of living they want.

How many workers could be in this population? What I have been told about this population is not uniform, but there are enough clues to reach a rough estimate. I asked my Robot Research Assistant to blend a variety of data sources to calculate the number of American workers who are earning between $20,000 to $40,000 per year, which is enough to approximate a “living,” but not a living wage in any state in the country.

In all, The Robot estimated around 56.5 million workers based on Census data. Providers have told me that the golden population is typically aged 25 to 34.1 After blending of data sources, The Robot’s best estimate is that age slice makes up about a quarter of that 56.5 million, or 14 million to place a nice round number on it. That’s nearly 13 times the estimated hiring needs in traditional manufacturing, semiconductors, aerospace, shipbuilding, biopharmaceuticals, data centers, and energy production, with the caveat that workers may not have geographical access to these jobs, which the Trump Administration recently identified as “high-priority industries.”

So, why isn’t this “golden population” filling these jobs, even amid a lot of kvetching about how America isn’t producing enough workers for these fields?

One small reason: they are shut out by how we build workforce training in the United States. And policymakers and business leaders are looking past them again even as they fret about growing retirements and shrinking birth rates diminishing the availability of qualified talent in key fields.

There are a lot of uncomfortable conversations we could have constructively about how American employers (and society) treats people who haven’t “figured it out” until their late twenties. We also could talk a lot more about how broken hiring is and whether these “unskilled” workers really can’t do some of these jobs. But as political leaders complain of the “inefficiency” of workforce spending, I think it’s important to zoom in on why these workers don’t get the help they need now.

For the sake of illustration, let’s say that the only thing standing between a 26-year-old working in a stable customer service role and a job in manufacturing is a credential, a piece of paper saying that the worker has had some sort of training that means they can safely and capably do the gig.

As recent research from Pew suggests, most likely that worker is going to have to pay out of pocket to get that credential. If that worker is making a less than the living wage—meaning they’re not self-sufficient—affording that is likely going to require sacrifices they might not be able to make.

It’s not impossible for some of these workers to get help from federal workforce programs, but it’s definitely harder and their options might be limited. As I often say, Congress has legislated these programs into a phone booth. One part out of fear of the wrong low-income person getting help, one part out of silly concerns about how helpful “welfare” should be, and one part a belief that the neediest people can be fixed if we just give them a little job training.

The most consistent streams of American workforce funding are the federal dollars Congress allots to states under the Workforce Innovation and Opportunity Act. Of those formula dollars, two streams—Dislocated Worker and Adult—are the most likely funding sources for a worker in this age range.

Because this worker has a job and isn’t losing it, they won’t qualify for Dislocated Worker funding, which generally applies to unemployed workers. WIOA Adult funding also probably won’t cover this worker. They may not be able to show they’re skill-deficient—a category that includes people who struggle with problem-solving, reading, and other skills—or they may have too high of an income.

Even if the Adult dollars could serve this worker, the credentialing program may not be cheap enough to get WIOA coverage in any meaningful way. Congress has programmed WIOA dollars to prioritize getting workers into the first available job, which is typically the type of lower-paying work the golden population is in now. Additionally, it’s hard to to pry from WIOA supportive services funding for things like childcare and transportation. Those services could be necessary to a golden population worker even starting a program.

In other words, the golden population worker is both too needy and not needy enough to get the help they need from public sources.

We can—and should—do many things, but I think a key issue is that many of our policymakers and political leaders come from wealthier backgrounds where college is all but guaranteed and work with a decent paycheck is a default expectation. As I told Rachel Lipson last fall, I think there is a relatability gap between the people who make decisions about funding these programs and the people who actually could benefit from them the most.

I think that’s a massive liability on our leaders’ part in building effective workforce strategies. I don’t think they see the golden population, and I think they struggle to understand the challenges facing workers who the system didn’t work for or who might have stumbled into adulthood. Accordingly, they’re missing a group of workers who could move into the jobs they’re worried about going unfilled—and who might not need that much investment to do it.

Fixing that is hard in the current environment of flaming tennis balls lobbed at every political problem regardless of advisability. So here’s the “easier” fix: Congress should expand definitions of who can be served by federal workforce programs to help this golden population get some assistance that helps them fulfill their potential. If focused on credentialing and guaranteed job opportunities, this could be a politically viable approach that could upgrade the outcomes of WIOA, which tend not to be great for participants—golden population or not.

Greetings, August people. I’m back with a regular edition of THE MONEY on Thursday, with a surprising new grant development from the Trump Administration and early news on the House fate of the current spending bill. See y’all then.