'Preventative care' workforce development is a good idea. Magical thinking is not.

Don’t try to use one problem to solve another.

The issue.

Policymakers have been describing more of a “preventative care” approach to workforce that helps keep people employed amid AI’s remake of the labor market. It’s a good change, but there’s some fantastical thinking in its suggested deployment that tries to use one workforce problem to solve another.

Explain.

In a year of big workforce changes, a very significant one might have fallen beneath your radar. In talking up training for “incumbent workers,” many higher-level policymakers and decisionmakers are describing what I call a “preventative care” approach to workforce: retraining people who are in jobs now that could go bad due to AI or other technological change.

In an ideal case, this would be checking the symptoms—your job’s skillset and role—and treating them, if necessary, with training or a transition to a role that is less likely to meet technological oblivion. This is in contrast to our current workforce funding systems, which are largely built for people who are out of work or will be very soon.

Being more proactive against job loss is a good, and overdue, shift in how leaders think about workforce development—even if it’s not borne meaningful changes just yet. But I find myself frustrated by some of the notions we’re shifting toward.

I have heard a very spreadsheet-y idea that because we have an oversupply of office workers, we can just shift those workers to industrial or trades jobs about to be hit by a wave of retirements. Here’s a good example from the Trump workforce blueprint that doesn’t specifically mention office workers.

Meeting critical industry workforce needs will require not only engaging the unemployed but also supporting the voluntary redeployment of workers currently employed in other sectors. In some cases, the only viable path to closing talent shortages in high-priority industries will involve workers leaving current jobs to pursue new opportunities that better align with national economic needs.

Emphasis mine. I connect this dot with the office worker redistribution theory because getting people—especially men—out of office jobs is a whole thing with the second Trump Administration and its supporters. There is a belief in MAGA circles that an education system aimed toward college and societal pressures have funneled too many workers into office jobs when they really should be working a line in a factory or operating a backhoe. Some flavors of Dudes Who Like Tweeting About How Mad They Are About Things and/or Stuff think getting men out of offices will increase birth rates. Neat!

Moving beyond the more politically relevant thought of whether men in cubicles are adequately sexy, my issue with the office-worker-to-lathe-person theory is that it tries to take an emerging workforce problem—AI job loss affecting white-collar fields—and use it to solve another—generations of barriers to many workers getting good jobs in manufacturing and trades, among other job types.

I’m not sure about that for a lot of reasons, but the main one is that it looks there are other options available that make more sense. Like better serving this population also described in the Trump blueprint.

An increasing number of Americans are disengaged and disincentivized from returning to work, with more than 21 million Americans aged 25 to 54 not participating in the labor force. At the same time, America’s workforce is aging, with millions of experienced workers retiring and fewer younger workers entering the pipeline to replace them.

I don’t want to fall victim to Spreadsheet Brain myself, so I will acknowledge upfront that not all 21 million of those people can accessibly enter open jobs that need filling. This population is not monolith, and the reason they’re not engaged in the employment market are complex and not easily fixed by a job, as my former colleague Rachel Lipson and I talked about earlier this month.

That said, by the pure math, 21 million workers is nearly 16 times the number of openings in “high-priority industries” identified in the Trump Administration’s workforce blueprint: traditional manufacturing, semiconductors, aerospace, shipbuilding, biopharmaceuticals, data centers, and energy production. That’s also not counting the millions of Americans unemployed, underemployed, or working underpaying jobs who could fill these roles.

So why are we saying “the only viable path” to filling worker shortages is by prying people from jobs they already have—and might kinda like?

Drastic medicine.

Here’s one reason why: the earliest places I heard this idea was part of a notion going around Republicans on the Hill that people already with stable jobs are easier and cheaper to train and place. Constructively, the problem with that thinking is that it’s nonsense and signals a lack of actual interaction with workers.

Reskilling is hard work and even office workers with decent incomes are likely to run into the same problems that have kept workers from lower tax brackets from doing it successfully. For example, if a worker has night classes, they might need to find nighttime childcare to cover their gaps. We’re in a childcare affordability and availability crisis, and that’s largely for regular work hours.

The type of reskilling suggested here also is drastic and effectively starting over for many workers. That doesn’t really line up with a realistic (or cost-efficient) way of helping office workers survive AI job loss. To go back to the idea of “preventative care,” a smart approach would be to diagnose workers’ vulnerabilities to AI job loss and train them to go into jobs that aren’t disappearing. To really follow through on the analogy, it’s not jumping from bloodwork that finds your cholesterol is a little high to a heart transplant.

I do think the threat of AI taking many Americans’ jobs is real, but I also see the demand aligning behind very-human skills needed for successful deployment of AI. If there’s something really dramatic coming, political and business leaders need to be more tangible and honest about it with workers they think need to go somewhere else. If there’s not, they need to think about retooling old jobs to make them AI integrated or transitioning workers to jobs that emerge from workforce changes driven by AI—which we need leadership and collaboration to define, not more fretting.

So what should we do about it?

First, I really do like the idea of “preventative care” workforce thinking because it not only helps the worker but the employer as well. My go-to strawman here is a bank that lays off accountants but complains they can’t find any good financial analysts who get the company. Many workers don’t want to start over, and as employers are having a billion AI products thrown at them, I think they will do themselves a favor by not throwing out people acclimated to their business.

Second, more business and political leaders need to go talk to and listen to workers. I won’t pretend that I’m on a worksite every week, but I can confirm that I made better policy in my last year at Labor because I was talking to workers about their experiences. I think there is a huge relatability issue that’s sending some political and business leaders far afield.

Obviously, the Trump Administration has a very employer-heavy approach to this work. Employers’ insights are important, but they shouldn’t be the only thing driving this conversation, particularly since some employers tend to get caught up on the Polaroids of missing technical skills and not a fuller view of where the labor market is headed and how to get workers there.

Finally, a new workforce problem can’t be used to solve an old one. That very much feels like what political leaders are doing when I hear talk about how we can simply fill industrial and trades job needs with office workers who may lose jobs due to AI. Both problems need to be addressed.

And if leaders really think the most viable option is to get office workers to go into manufacturing, the trades, or other jobs that will effectively reboot careers, those leaders are going to need to sell that idea to those workers instead of assuming desperation will do the work for them.

Card subject to change.

This is a topic shift from what I promised last Tuesday, but I didn’t feel like this thinking shift—and the flaws in some of the approaches to it—had fully been aired out. I’ll be back next Tuesday to talk about the golden population of job training participants.

I’m on the road this week, but minding the shutdown and other developments. Friday may be a little late, but I’ll see you then. Spookily.