We're talking about apprenticeship from the wrong angle.

America's main quality control method for apprenticeship is seen as a barrier to scaling it, but that's not what it was built to do.

The issue.

Critics complain that the main government-blessed form of apprenticeship is not growing fast enough to fill employers’ needs. But that’s not what Congress built it to do at all.

Wait, what?

Politically, expanding apprenticeship in the United States is always a good idea, yet also always behind schedule. As lawmakers issued their list of demands to new DOL leadership in hearings over the past few weeks, thee seemed to be some befuddlement that apprenticeship isn’t a more common way into a good job. Especially in light of some lawmakers’ increasing belief that overemphasis of college is the biggest problem in American workforce development.

This always seems true for Registered Apprenticeship,¹ the regulatory program run by the Department of Labor that targets certain (underfunded) resources toward programs that show they meet certain quality standards. Set up by a 1937 law, Registered Apprenticeship deals with a lot of tense politics and exaggeration of its features. For newcomers, there are four important things you should know:

1. You don’t have to register an apprenticeship program with the Department of Labor to operate, but you may get some federal and state help for your program if you do.

2. While DOL and others often refer to a “National Apprenticeship System,” we really have more of a patchwork apprenticeship “confederation.” Half the states handle day-to-day registration and management themselves, having met certain deputization requirements set by DOL. In the remaining states, DOL registers programs directly. These two types of schemes are more connected than they used to be, but not always operated the same because, well, DOL and states don’t necessarily share the exact same stakeholders and goals.

3. The general consensus is it takes a bit too much work to set up a Registered Apprenticeship program, especially from DOL’s understaffed and overextended Office of Apprenticeship. Programs must meet basic labor standards, such as paying at least the minimum wage, ensuring worker safety, and avoiding discrimination against certain groups. But you also have to show DOL you have proper training methods and there are some extensive Obama-era efforts meant to diversify Registered Apprenticeship, which historically has had trouble with employment of women and people of color.

4. Unions run a great many Registered Apprenticeship programs, making the concept inherently political. To state it in broad terms for the uninitiated, Registered Apprenticeship is viewed suspiciously by anti-union employer groups and steadfastly protected by organized labor.

Despite all this political freight, Trump Team appears to have done a gentle about-face toward support of Registered Apprenticeship. The first Trump Administration spent most of its time on a failed effort to invent a Registered Apprenticeship alternative, Industry-Recognized Apprenticeship Programs, or IRAPs, that I think turned out on balance to be a more complicated, vaguely privatized version of Registered Apprenticeship. So it was a big deal last month that the new Labor Secretary Lori Chavez-DeRemer showed support during her confirmation for Registered Apprenticeship.

Separately, now-Education Secretary Linda McMahon tweeted in November in support of a more rigorous form of apprenticeship by asking everyone’s favorite workforce question: why can’t we all be more like the Swiss?²

Yet, I doubt the Administration’s apparent backing will quiet critics and skeptics of Registered Apprenticeship, who say that the reason apprenticeship is behind in the United States is that Registered Apprenticeship is too hard and it doesn’t move enough people into jobs quickly enough.

For example, the following is from the summary of a Rand Corporation report from December 2024 on Registered Apprenticeship and construction employment:

Although registered apprenticeship enrollment has doubled over the past decade, anticipated labor demand is more than double what current workforce development pathways are likely to provide. . . . If the United States focuses on apprenticeships as the main strategy to meet labor demand without addressing capacity and noncompletion, the country will fail to meet the growing construction workforce demand.

Why doesn’t apprenticeship grow more easily? Why is it so hard? Apprentix, a company that helps with Registered Apprenticeship compliance, blames an “antiquated, unnecessarily complicated system.” Something similar was said by a leader at Multiverse, a U.K. tech company focused on apprenticeship that entered the U.S. market to great fanfare in Apprenticeship World, lost $50 million, then exited abruptly to great sorrow in Apprenticeship World:

“What we heard from employers was a need for training that can be deployed rapidly . . . and can deliver immediate results,”³ says Tim Smith, Multiverse’s vice president of corporate affairs.

“Ultimately that requires a level of flexibility in duration and content that isn’t fundamental to U.S. registered apprenticeships,” he says.⁴

This all seemed to feed into the original Trump Administration view of Registered Apprenticeship. For example, this is what former Secretary of Labor Alexander Acosta, really the father of Trump I’s IRAPs initiative, said in 2017 to justify the need for an alternative system:

Acosta said “0.3 percent of the workforce” use the registered apprentice program and lamented that it’s “not scaling.”

I think about this Acosta quote a lot because The Great IRAP Adventure taught me what Registered Apprenticeship really is. As part of my work on a rule that got more than 300,000 comments, I had to spend many hours with the legislative history of the National Apprenticeship Act, that 1937 law that authorizes the federal registration of apprenticeship programs.

What I learned is that Registered Apprenticeship is not really a “model” of workforce development—it’s a consumer protection device. It’s not built to “scale” because it was meant to protect consumers by giving them a signal to recognize safe apprenticeship programs that train workers well.

To me that raises a question: why is a voluntary consumer protection device to blame for there not being more apprenticeship programs?

Explain yourself.

The NAA is also known as the Fitzgerald Act because it was championed by Connecticut Rep. William J. Fitzgerald, a Democrat and former superintendent at a foundry. He held a series of hearings in April 1937 on the need for more protections for apprentices, calling out exploitation by employers and fraudulent programs that don’t provide real training. A lack of quality training meant too few skilled workers, he said.

A federal mechanism for recognizing quality training would address this issue, or as Fitzgerald put it on the floor of the House:

[The bill is] throwing a cloak of protection around the boys and girls and setting up standards and protecting them and guaranteeing that when their time of service in a trade has expired, they will come out full-fledged mechanics.

It’s amazing to read conversations from 88 years ago and see how little progression there has been in some regions of the political conversation about apprenticeship. This was all well before the overemphasis on college that current political leaders seem to portray as the big underlying problem in workforce development. Yet, back in April of 1937, long before the ubiquity of college due to the GI Bill, witnesses at hearings also seemed to think apprenticeship should work for more young people and that livelihoods were left on the table because of the lack of widely available, adequate training programs.

Even the management-labor conflict is eerily similar. One of the House witnesses at the Fitzgerald hearings, Julia O’Connor Parker of the National Youth Administration, observed that unions seemed to be better at apprenticeship then, too. Her comments imply that the bill was not really needed for union programs because their apprentices were treated well and trained well.

[Non-union workers] need this protection . . . and these standards of apprenticeship will do a very great good in improving industrial standards, I think, by providing a recognized standard of true apprenticeship with its definition understood.

I pulled these quotes from legislative histories found by the lovely staff at the Library of Congress, but I think they are incomplete. My recollection of my original encounter of hearing transcripts—one I hope is not apocryphal because it was two kids and a pandemic ago—is that there was Senate testimony by THA GAWD of Labor Secretaries, Frances Perkins. I wasn’t able to find that testimony for this piece.⁵

But one thing I recalled that I found in my revisit of these materials is that there was a definite limit on how far Congress was willing to go. Congress would add a voluntary requirement that held up certain standards of apprenticeship, but that was the beginning and end of the solution. That seemed to be a carefully negotiated point of the bill.

On the eve of the pandemic, there was an attempt to reauthorize the NAA that, in my opinion, would have killed IRAPs by making everything explicitly about registration. Beyond that part, the bill put actual investment toward creating a true national apprenticeship system beyond the patchwork we have now.

Yet, I have my doubts those investments would ever have become law in the last two administrations. Because if we couldn’t spend real money on apprenticeship during the New frickin’ Deal, I doubt it would have happened in the last five years of American politics.

So?

We probably shouldn’t blame a voluntary consumer protection device for apprenticeship not growing in the United States.

Instead, we should look to what employers and government can do to grow apprenticeship instead of devolving the quality of Registered Apprenticeship to somehow spark growth.

So what do we do about it?

Let’s get three big points out of the way:

1. No, I don’t think we need an alternative to Registered Apprenticeship. We have perfectly good government-blessed apprenticeship at home.

2. No, I don’t think we need fewer protections and training standards for registered apprentices. That’s the point.

3. But I don’t think we can handwave away the intervening nine decades of conversation about Registered Apprenticeship since the NAA passed.

We do need to reset the conversation, though. Because right now people are complaining that Registered Apprenticeship isn’t easy enough to be an entrypoint, when Registered Apprenticeship should be a goal employers want to achieve with their programs.

Here are some ideas for doing that:

Treat Registered Apprenticeship as a Michelin star, not a permit to open from the Health Department.

Instead of asking for Registered Apprenticeship to be worse, employers should figure out workforce programs that fit their needs then bring them into Registered Apprenticeship to signal quality to consumers. That fits with what Registered Apprenticeship was really built to be—a voluntary marker of quality—and not what some people treat it as—a threshold to starting any workforce training program.

The Michelin star analogy is imperfect, but it’s also the best way I can think of describing where we need to get. I’m a barbecue enthusiast and I have had some of the best meals of my life in ramshackle structures that look like they will give you tetanus if you hang out in them too long. I also have had fabulous meals at slick Michelin-rated restaurants that dress up the shack food for a broader and less-daring swath of consumers. Still, there are definitely people who prefer the Tetanus Hut.

But people don’t blame the Michelin star for not opening more restaurants, and they shouldn’t blame Registered Apprenticeship for holding back apprenticeship expansion in the United States. If employers need programs that are easier to start than Registered Apprenticeship, they should just start them, not try to make Registered Apprenticeship worse to make it easier to get into (presumably to tap into limited public resources).

So if you’re an employer, go build your Tetanus Hut if you want, and don’t register it if you don’t want to. But if it turns out workers like the union apprenticeship equivalent of French Laundry—because they put in the work to deliver quality and also it doesn’t give them tetanus—that’s the risk you’re taking by not securing the proper bona fides.

Maybe Registered Apprenticeship isn’t the problem with employers not finding enough skilled workers.

I want to revisit that quote from Multiverse, with the acknowledgement that I might be missing some context. Something occurred to me as I was writing about it earlier.

“What we heard from employers was a need for training that can be deployed rapidly . . . and can deliver immediate results.”

Apprenticeship involves hiring a human being to do the job right now and then watching the results get better immediately because they’re learning more skills.

As I wrote about skills-first hiring a few weeks ago, skills-first hiring won’t fix your hiring if you don’t actually fix your hiring. If having someone do the work right that moment and train to meet your specifications isn’t the answer, then I think you have other problems to solve.

Rub some AI on the apprenticeship registration process.

I do think employers have a point that the registration process is too slow. And while I dislike the answer to everything being “rub some AI on it” … AI really could be a solution here.

The federal government should invest in training generative AI to help review Registered Apprenticeship applications and help connect employer training programs to existing apprenticeship standards and requirements to reduce back and forth in the application process.

If DOGE is going have tech entrepreneurs and software engineers scrutinizing DOL, I think this is a place where they can invest—not cut—and get a much more efficient result.⁶

Build the interstates of apprenticeship instead of trying to get a bargain on it, Congress.

Congress is still trying to do apprenticeship on the cheap, spreading under $300 million a year across the whole country to try to turn a consumer protection device and a patchwork of state-federal partnerships into a “system.”⁷ If lawmakers want apprenticeship to grow—and Registered Apprenticeship to move faster and more evenly—they have to build its interstates by investing in data infrastructure, personnel, and greater incentives for employers to get into the game.

And if you’ve got that Old Neutral Feeling and find yourself asking

Well, the Swiss set up public infrastructure that meaningfully furthers apprenticeship.

And until our lawmakers put the investment into apprenticeship, well, maybe we shouldn’t ask why it’s not growing fast enough despite all that political backing.

We already know the answer.

Hello, valued customer

Sorry for this one being late. I’m sick plus I spent most of the past few days unpacking what’s going on with workforce money thanks to Congress and the Administration. Three notes:

1. I am overwhelmed by the feedback on the grants coverage in the past week. Thrilled JTW has found an audience so early.

2. Expect a price hike next week for the Founding Reader paid tier to reflect some changes to my schedule for the 45-minute counseling sessions. If you want the current rate, subscribe this week.

3. I would love your feedback on whether you prefer newsletter delivery in the morning or afternoon. The back-end metrics show some interesting patterns and I want to better serve these free pieces to you at the right time.

Card subject to change

You may have noticed I didn’t publish a piece on Chavez-DeRemer’s confirmation as Labor Secretary, mainly because the callback of DOL probationary staff rendered much of it moot (for the moment). I’ll have a final #LCDWatch for Substack Notes and LinkedIn, but the short of my point is that if the Trump Administration wants to show the Chavez-DeRemer pick means it is committed to workers, you kind of need a non-gutted DOL to do it. Also, don’t revise all the workforce money to avoid sending workers into good jobs and quietly kill quality pre-apprenticeship. That would help too.

FRIDAY: Grants listings plus I delicately explain why I think a Trump DOL move to put more unrestricted money into apprenticeship incentives—something I actually like—could be repeating a mistake from the first term.

NEXT TUESDAY: TBD based on the shutdown! Crap!

1. It’s time for Messaging Nick to peek in here and be picky: it’s “Registered Apprenticeship,” not “registered apprenticeship.” It’s a formal designation, a concept, and a brand, so we capitalize. This is a DOL thing decided long ago that has been lost in the folkways of the workforce development sector.

Separately, my preference is “Registered Apprenticeship program” not “Registered Apprenticeship Program,” but the abbreviation “RAP” and “RAPs” is so ubiquitous that I’m fine with it.

2. (I appreciate all two of you who get this reference.)

3. I cut some text here about adapting to changing technology to hone in on what I think are the most relevant portions of this quote.

4. For what it’s worth, I think that this last comment by Multiverse shows that they didn’t scope the U.S. market correctly, not that Registered Apprenticeship has some inherent flaw that can be blamed for Multiverse losing millions of dollars here. Happy to have a conversation if you think I’m wrong, Multiverse! I’ll write about it for this newsletter! Email nick@jobsthat.work. Seriously!

5. If you have it, though, nick@jobsthat.work. Since I recall her talking to Hugo Black in it, this isn’t some privileged DOL document.

6. They’re going to cut.

7. Weird how nearly all of my pieces end with pointing out that we don’t spend enough money on workforce development in the United States. I wonder what it means?