Don't cut federal jobs money, spend it on people the smart way.

The real way to make federal workforce spending more efficient in the age of DOGE.

The issue

It is too hard to spend the little money America puts into workforce development on things we know actually help people get good jobs. This flaw makes the money an even riper target for our new reactive Government Efficiency overlords.

Wait, what?

A few months ago I talked with a worker in the Phoenix area who is the perfect success story for a federal workforce program. She reminded me of something that lawmakers seem to forget when they talk workforce.

Kimberly Rush¹ spent two decades in food service jobs without sick leave or paid time off before she entered a Department of Labor-funded workforce program. She got trained at a community college, then hired by a technology and manufacturing company, then promoted to an even better salary and flexible work schedule.

This is awesome, but what her story really reminded me of was that it’s damn hard work to get a good job if you do not come from means—and often it takes some extra help to do it.

During the program, Kimberly was working, going to her oncology appointments, and attending classes at a community college. One day she needed sleep so bad she loaded an airplane pillow full of ice and dozed in her car in Arizona heat.

“I found myself just going on empty, physically, spiritually, emotionally,” she told me.

One of the things that helped the most? The DOL-funded program supporting Kimberly gave her a debit card for gas.

“Having just that card for gas and groceries, it gave you a little bit of a cushion,” she said. “I don’t have to worry about this right now.”

In my experience, when workers get the support Kimberly did, they tend to succeed. But in my decade in the federal government, I found that federal systems default against covering the type of support that got Kimberly ahead—when evidence and pure logic would tell us that it should be an essential part of every workforce program.

Explain yourself

For the unfamiliar, that debit card was something workforce programs provide called a “supportive service” or “wraparound service.” These services cover the costs that can prevent someone from entering or completing a training program.

Because you still have to pay your rent even if you’re trying to get a Commercial Driver’s License, and your kid generally can’t just hang out when you’re learning about welding. Even if the kid has more experience welding than you.

Strangely, research suggests that paying for things that help people through jobs training tends to help people actually get jobs. The program funding Kimberly’s training is outside DOL’s main pot of workforce dollars and usually pays for these services without too much bureaucratic consternation.²

For that main federal pot of workforce money, though, supportive services are often a maybe expense—and one Congress has framed as a last resort.

To help the unfamiliar, I’m going to lump together a building-full of programs for the sake of making a point: the pot I’m referring to comes from the Workforce Innovation and Opportunity Act and other interconnected statutes that tell DOL and the Department of Education to run workforce programs. To keep things really simple, let’s say Congress puts about $6 billion per year into this pot³ for the entire country—or about 1/7^th of the cost of an emotional purchase of Twitter, a metric I mention for no purpose whatsoever.

Under WIOA, adult workers who need supportive services must be “unable to obtain such supportive services through other programs providing such services[.]”⁴

Let me explain why this is a bad standard with a theoretical based on another interview I did last summer with a single mom in Alabama.⁵ She could get her young son into Head Start for childcare, but Head Start closed at 2 p.m., a time when she was, of course, working.

If she was a federal workforce trainee in the program above, a reasonable reading of the law would be that she is “unable to obtain” childcare through other programs and could get help with care that fit her training hours.

A much stricter but also OK read would be that the mere existence of that Head Start program means she could get childcare from other programs. Therefore, as a trainee, she could not receive help even though childcare available couldn’t possibly work.

(FYI: I would expect a restrictive reading of the law for the next four years.)

So if we won’t spend on things that help workers, what does Congress want the money spent on instead?

In effect, bodies and paper.⁶

Workforce is severely underfunded in the United States. By not investing more, the philosophy flowing down from the Hill has been to train as many people as possible with as little money as possible. Oh and document what you do through extensive reporting systems hardwired into the law,⁷ which create overhead that can cause even needy grantees to swear off federal workforce dollars, in my experience.

In my opinion, this policy thinking was reinforced in a failed bill in the recent lame duck session, which would have reauthorized most of the programs in the pot. This bill required 50 percent spending on training, and based on my read of publicly available drafts, it would have carried forward the last-resort approach while capping some supportive services at a ridiculously low percentage for participants who need them the most.⁸

A good example of the attitudes that flavored that bill and current law is in this 2022 excerpt from Virginia Foxx, then ranking member on the House Education and Labor Committee. She is speaking on an earlier reauthorization bill that, had it passed, would have increased access to supportive services.

This bill . . . expands supportive services, which risks turning the workforce system into a welfare program. . . . Barely one-third of participants in adult and dislocated worker programs exited WIOA in 2020⁹ with employment connected to the skills education they received, despite pouring nearly $2 billion in taxpayer funds into services.

This is a literally backwards way of looking at the problem, but look, I understand the concern. Kind of.

People can make poor choices with money, especially if they never had it before. It’s a risk to give people money for virtually anything, and you don’t want a program to snowball into spending on everything but actual training and jobs. That’s why it feels so much safer, politically, to prioritize training, which conceptually has fewer of the frustrating human risks that bleed in from lifetime after lifetime of diminished expectations.

It also must be acknowledged that $2 billion is a lot of federal cash for the entire country—almost a third of the $6.2 billion loss SoftBank posted after the bankruptcy of WeWork, a company with the business model of “Free beer in the kitchen!” and “Don’t ask about how we’re gonna keep paying for the free beer in the kitchen!” Which, of course, I mention here for no reason at all.

That said, when we’re talking about training humans…

These are necessary risks. It’s hard to effectively get humans trained and hired without buying those humans what they need to get trained and hired. Like offering tools to apprentice on the job site, providing care for their kids, or giving them an affordable way to get to the places they need to go to train.

So?

If the government doesn’t spend on those things, people are less likely to be successful—and you’re more likely to spend federal resources on those folks later. Including dollars from the federal workforce system you’re trying to keep so spendthrift now.

Don’t mistake me, there are several awesome training providers—including states and localities—that help a lot of people with these constraints. God bless them.

But as we enter the age of Government Efficiency, our big ol’ federal pot of workforce money has a good chance of being shredded for not being efficient enough. As the Foxx statement and a mess of other interpretations say, the numbers are mixed. So if they’re looking to dismiss the system as massive failure, they have the material.

Even if they never notice that material is sitting around because Congress never put it together in a way that effectively covered the people who needed it most.

What do we do about it?

In public policy, you get what you pay for, but you get what you don’t pay for, too.

If Elon Musk and his Government Efficiency homies gave me a ring, I would recommend they rethink, not indiscriminately cut, how we spend federal workforce cash to give the things that work a better shot.

They’ll do none of that, of course, but I think some of these are good ideas for any workforce program.

Build training that works by spending on what people need.

If I had my druthers (and I frequently don’t), I would spend at least 20 percent of every program on what humans actually need to train and hop from one class to another: childcare, housing, food, etc. Training is very important, but if you invest in the things that make training work, you’re less likely to have to provide future federal help to the folks you’re helping now.

Focus on quality, not quantity.

It makes no sense to try to move as many bodies as possible through the workforce system because, well, we are only willing to spend so much. If Congress is not going to invest the money to serve all the people well, then we need to do the absolute best we can with the money at hand.

Be slightly less precious with workforce dollars.

As a policymaker, I have seen and heard enough from the field to think that, by requiring mounds of data and tracking, we are capping the benefits the government gets as an investor in workforce programs. Venture capital takes risks with its money, why can’t government when it invests in actual people who need it on the things that will help them get ahead?

Unburdening the money just a tiny bit will help providers get the job done and maybe, just maybe, give Congress a reason to feel good about spending those dollars.

Mildly unfair parting thought

Call me crazy, but it feels like we don’t want to invest in these workers because they are lower income and we don’t value them that much.

I say this as a class-jumper myself who sits here, in a fancy chair you can’t see, because a jobs program dropped me off at a newspaper when I was 14. It took a lot of damn hard work to get to here, and it was a long while and many fancy chairs before people really felt comfortable investing in me. I know several folks who were not as lucky as me, and I think about them a lot and what could have happened if they got a little more help.

Last year, DOL made available more than $1.2 billion for good jobs for the entire country like the ones that helped Kimberly. I don’t know what will become of that money based on the past week, but I hope at least a few of those dollars give a little more help to the people who need it.

Even if that total amount was $575 million fewer dollars than what investors poured into six fabulous months of Quibi.

Which I mention for no reason at all.

Coming soon

Thanks for reading the first edition of JOBS THAT WORK, which hopefully was muuuuuuuucccch longer than most of these newsletters will be. But the supportive services question is foundational to a lot of problems in American workforce policy, and there was plenty to unpack and plenty I left in the pack, where it may well gather ants.

Disagree with what I said? Including the part about ants? Need to helpfully remind me I lumped a lot of separate programs in one pot to make my point? Email me at nick@jobsthat.work.

Last week was bad! No reason! Let’s talk the next two weeks!

• Well, FRIDAY was supposed to be the first edition of our grant listings but OMB circulated a memo last night halting all federal financial assistance. Cool! My plan is to still do grant listings but have some analysis of what we know by Friday morning (including any TROs) and explain what I would do if I was a grantee reliant on a particular federal program.

• NEXT FRIDAY, I expect to publish a fuller survival guide for federal grantees. This could be delayed if developments suggest that it will be outdated as soon as it hits the street.

• NEXT TUESDAY (hopefully?): The Trump Labor Department’s “The Lady, or the Tiger?” choice between Trumpian immigration and Trumpian apprenticeship. Spoiler: there is no lady.
  
  ROWR.

1. I originally spoke to Kimberly for a story on the Department of Labor’s website about changes made to grants to help people get good jobs. The Trump Administration removed this story last week.

2. And now to answer question I get a lot these days: is this doomed? Kimberly participated in the Women in Apprenticeship and Nontraditional Occupations Program, operated by DOL’s Women’s Bureau. The Women’s Bureau is one of the more effective agencies in the federal government for the amount it is funded. It has done tremendous work tracking the childcare cost crisis keeping parents out of the workforce.

Project 2025 calls for significantly remaking the “politicized” Women’s Bureau to focus on “disentangl[ing] the influences on women’s workforce participation and [understanding] the true causes of earnings gaps between men and women.” Ending the 105-year-old Women’s Bureau would take an act of Congress; House Republicans unsuccessfully tried eliminating it in 2023 without explanation.

If the Bureau survives, I expect significant danger to the WANTO program. Ending it also would take an act of Congress, and doing so would sunset by far one of the country’s best programs based on return on investment. WANTO has never had more than $5 million granted in one year, but regularly produces results like Kimberly’s.

3. A disclaimer for the folks who swim deep into these waters: I am combining statutes and funding lines to offer the roundest number for unfamiliar readers. In my estimate, I include the Workforce Innovation and Opportunity Act, Older Americans Act, and dollars appropriated specially for Registered Apprenticeship, among others. If you put the money from all these laws in one pot, you have about half the money you need for your own emotional buy of Twitter at its current valuation.

4. WIOA sec. 134(d)(2). To assuage people who swim deep in these waters: yes, WIOA funds several different programs and this provision pertaining to adult and dislocated workers does not cover all of them. My experience is that this provision tends to cross into programs where it doesn’t necessarily apply because it is politically and administratively safe.

5. As you probably gathered if you clicked the link, I talked to Diana, the mom, while researching an op-ed co-authored by my old boss Julie Su and Birmingham Mayor Randall Woodfin.

6. Just to let you know what type of sacred oath I’m abandoning: I tracked misuses of public dollars as an investigative reporter, then I was a belt-and-suspenders-over-another-pair-of-pants-with-a-belt-and-suspenders kind of government attorney when it came to the money.

7. I need to have a few more conversations before I feel comfortable letting out my WIOA data take, but to give you a sense of what is collected, it’s things like wages two quarters and four quarters after trainees finish a program. At a high level, I think those indicators tend not to measure what really matters, which is whether trainees get jobs that they can actually stay and thrive in.

8. The linked bill caps supportive services at 5 percent for workers returning from incarceration, who typically have nothing and need the extra support to avoid sliding back into jail. To put it politely, that cap is utterly ridiculous.

That said, many of the leading organizations in workforce development endorsed the bill. Despite its flaws, I kind of wish it had passed to insulate the workforce system from the cut-first-ask-questions-later-oh-no-fire approach I expect to dominate the Hill this spring.

9. 2020 of course being a year when the global pandemic made it more difficult to train people and the jobs market cratered like the value of Theranos, which raised $1.4 billion in funding from investors based on fraud, connections, and what seems like the greatest bland pitch deck ever made.

Which I mention for no reason at all.