Ladies and gentlemen: Johnny Cash (also walking a line)

There’s a labor market economist who posts on LinkedIn named Gad Levanon. He writes a lot about AI and the travails of young adults. I find some of his stuff annoying,1 but almost all of it worthwhile: it pushes my thinking and forces me to reconsider my priors.

Today he posted about how this is a historically strong labor market, in relative terms, for non-college graduates. This section in particular caught my eye:

Part of the non-graduate strength is probably supply rather than demand. Immigration enforcement has thinned the pool of less-educated workers over the past year and a half, and the construction and factory-building boom has soaked up much of what remains. Data centers, after all, are built by electricians, so the same technology that may be cooling desk work is hiring hard hats.

Emphasis mine. Now, I think on balance that immigration enforcement has been a pretty significant net economic and cultural negative for the country. Our labor force has stopped growing, the “breakeven number” of new jobs required each month to keep the unemployment rate flat has fallen off a cliff, and long-term we’re facing the risk of significant shortages in those same sectors where native-born Americans are presumably having an easier time finding jobs right now, like construction and healthcare.2 Those shortages will jack up prices, lower supply of services, and generally reduce quality of life, particularly for low-income households. There’s also the less immediately quantifiable loss in entrepreneurial energy, the opportunity cost of brilliant international students electing to enroll in other colleges, and the diminishment of our standing in the eyes of a world that now sees us, understandably, as having elected a government of white supremacists.3

All of this strikes me as a wildly, obviously unacceptable trade-off for helping native-born non-college grads have an easier time finding work. But is it really working? If so, what does that mean in political terms?

I started by asking Claude for an overview of the labor market outlook for white non-college workers. The word in its response that stuck out was “deterioration.” The unemployment rate was slightly down,4 but the employment-to-population ratio was down by considerably more. As for earnings versus inflation—purchasing power, in other words—the only group by education that saw real wages outpace inflation was those with less than high school. This was the same group that saw the biggest drop in labor force participation.

When it comes to earnings, however, the picture is more positive. White workers in all education categories below bachelor’s degree saw real earnings growth, as did Black and Hispanic Latino workers in most categories—the exception being Black workers with some college/associate degree, who saw a modest decline.

Now, has this mattered politically? That’s a question we can only answer by inference, but the answer would seem to be between “not enough,” and “no.” Overall, Trump’s approval has declined among white non-college voters, the group that pretty much won him the presidency twice. He’s under 50 percent with them now, according to most polls. And while it seems almost certain that most of those voters still prefer him and Republican candidates for lower offices to the Democrats, they’re not as enthused about voting this year. If this population has seen employment gains, apparently they’re not significant enough to outweigh whatever else is going on.

Also important to note: The economic data here lag the polls. Trump kept net positive approval among this group until winter 2026, when Bibi Netanyahu Sith mind-tricked him into attacking Iran. So it’s possible that the spike in gas prices wiped out whatever purchasing power gains that workers with less education had accrued in 2025, and/or more of them simply became alienated over the realization that Trump’s priorities weren’t theirs.

But even if those gains have been wiped out, that they happened at all seems notable for two reasons. One, it is under-covered—to put it mildly—among mainstream media outlets like the New York Times and NPR, which at this point I’m guessing mostly reach college graduates.

Two—and this might not be intuitive to my fellow liberals—this is, in itself, a good thing. We want Americans to prosper, and this group really does have a legitimate gripe that both parties have made policy choices over the last 50 years that served to screw them over. It might not be how we would have drawn up trying to help them. It might not be, and I clearly don’t think it is, worth the tradeoff. But I think having the economic priorities of working class voters more central to our politics is probably more of a good thing than bad.

I also could have gone with “All the Pieces Matter.” But who could say no to these clean-cut young go-getters?

Over the past few months, I’ve become involved in multiple projects that take a holistic view of challenges across the young adult employment landscape. This is very welcome, in that it builds upon the work I did in New York City government at the intersection of youth development, career readiness, and labor market entry—and also quite intimidating, because there’s a lot of scary stuff happening in this area, and fairly vast unknowns to navigate. What will AI implementation or the normalization of remote work mean for entry-level white collar employment? How are young people and their families thinking about higher education and training alternatives? Do we have the career guidance infrastructure to help young people first identify their interests, then pursue them with intention?

There’s more. What about regional variation, including the different dynamics at play in large versus small metro areas, or rural communities? How much space is there now, or should there be, for career connected learning within K-12 education? What are the implications of the “loneliness epidemic,” or mental health challenges generally, or drug decriminalization, on young adults’ career readiness?

My understanding of the hard sciences is that researchers give a lot of thought and attention to interaction effects. Your health isn’t just a function of your appetite and consumption habits, or exercise, or genetic makeup; it’s all of the above. Economics as a field takes this approach too: economic growth (or its absence) is downstream from a multitude of factors that might include everything from weather patterns to how well world leaders get along.

In the areas of policy where I’ve spent most of my time, I don’t think anyone would actually argue for singular and direct causal relationships; I can’t imagine any serious person claiming that, say, someone’s performance in high school is the only determinant of whether or not they get a job afterward, or even how well they perform in college. Functionally, though, that’s kind of how we operate. We assess one single program, or input, and draw sweeping conclusions as to its value without considering all the other potentially confounding factors.

My favorite example of this is probably the landmark study of NYC’s Summer Youth Employment Program (SYEP) released in 2015, right around when I started at the Center for Youth Employment. As thoughtful and responsible researchers, the authors noted the following effects:

What they didn’t find, as hinted at in #2, was a connection between SYEP participation and stronger subsequent labor market outcomes over the long term.1 An alternative way to frame that might be “SYEP participation doesn’t help kids find jobs or earn good wages.” If you want to stick a “researchers find” in there, fair enough.

But I’d argue that’s not really much more true—in the sense of saying something meaningful—than my asserting that despite eating healthy and sleeping well, I’m not a professional basketball player. Yes, it’s technically accurate. But it’s powerfully not the point, and as the kids say, absence of evidence is not evidence of absence.

I believe that people fundamentally respond to incentives, and pretty much every incentive for professionals runs toward simplifying complicated things. The more factors one considers, the less control one seems to have. The more there is to account for, the greater the risk of being exposed as incorrect or ignorant.2 As the problem becomes more complicated, the solutions feel further out of reach.3 There’s even risk of being blamed, explicitly or implicitly, for things that are not only not your fault, but that you might never even have touched in your career.

I’m not immune to these considerations, nor to the urge to throw up my hands and declare certain problems beyond our ability to meaningfully address. It is super tempting, and not irrational, to point a finger in the direction of capitalism4 and more or less call it a day. But that’s not really the job, and if anything I personally feel like I have more of a responsibility to try to look at the entire landscape, since I’m not permanently accountable to any one employer.

I remain more pessimistic than otherwise about AI.5 But one thing I’m pretty sure it’s helpful with is managing complexity. The other night I started a framework and rubric for state-level career readiness ecosystems, going back and forth with Claude for an hour or so on what turned out to be 14 indicators over three domains. It was a synthesis of the AI’s ability to search vast realms of materials with my sense, developed from experience, of what inputs mattered most. I’m excited to see what happens when it comes into contact with reality, and perhaps eventually to create something beautiful and valuable out of all this messy complexity.

This guy’s whole thing was to work where the money was.

Last year I worked on a project for a client in which we tried to map out exactly how and for what purposes New York City spent money on job training and employment programs. Perhaps the most striking finding was how relatively little NYC invested in opportunities to support the more than 900,000 working-age adults employed part-time but seeking full-time jobs, or employed full-time but under or barely over the poverty line, and often in jobs with high turnover or limited advancement prospects.1

From a certain perspective, one could argue that this reflects the system performing as designed. Federal Workforce Innovation and Opportunity Act (WIOA) funding remains geared toward placement into relatively low-wage, low-quality jobs, with minimal focus on what happens next. As I’ve written before, I see this as a manifestation of our mostly unconscious2 view that people mostly should make their own way in the labor market—a choice that sacrifices vast economic potential on the altar of “individualism,” and (please hold onto your monocle and keep your smelling salts at hand) tends to reinforce existing patterns of privilege and power that are strongly coded by race and class.

Now, I don’t want to overstate the case here, or assign blame where empathy is a more appropriate response. Local and state system administrators and policymakers are constrained by ever-declining federal investments—in real dollars, federal workforce spending has fallen by 77 percent since 1979, even as the workforce has grown by 50 percent—as well as policy that requires “priority of service” for public assistance recipients and others with particularly acute needs.3 This of course helps explain why most of the placements are in low-quality jobs, and why so few dollars are left over to do anything else.

I’m more curious why places like NYC, which does have some capacity to make investments aligned to its policy priorities that complement federal and state funds, haven’t really leaned in here—at least not in a fully intentional and strategic way. The indispensable ALICE (Asset Limited, Resource Constrained, Employed) report for New York, released earlier this year,4 found that 58 percent of households in the five boroughs did not earn enough to meet basic costs of living. The large majority of those households include at least one working adult, and many have two or even more.

If I were, say, a Democratic Socialist public official who wanted to show that I meant what I said about putting the power of the government to use on behalf of the working class, I would see in those 900,000 low-income working New Yorkers an enormous target of opportunity. How many of them might be within reach of middle-income jobs and real economic security with just a bit more support—a chance to finish the college degree they stopped short of completing, or access to funds that help them overcome emergency expenses related to childcare or transportation or medical issues, or just a source for career guidance and financial planning? How many might have built skills or developed networks through their past and current positions that could translate into higher-paying opportunities than they have right now? What would it cost for my administration to help them, and who—among businesses or philanthropy or even state government—might want to share those costs?

A lot of this, of course, is doable with assistance from AI. In that spirit, I’d like to welcome Claude back to Disillusionist! I asked for a starting strategy on identifying a share of current low-income workers for support to advance into middle-wage jobs, and got back a four-step plan graphically presented here:

First you’d disaggregate the 900,000 to determine categories of need; then you’d consider where there’s leverage to engage; then you’d determine the mix of interventions most likely to enable advancement; and finally you’d recalibrate the policy/funding ecosystem to render this progress sustainable.

Within that third bucket—the most important as far as actually helping individual workers achieve economic mobility—Claude also noted a Bloomberg-era initiative that I’d forgotten about called WorkAdvance, which attempted (on a much smaller scale) something along these lines, and suggested three major takeaways:

This is, indeed, the stuff—and none of it really fits well with federal policy as it’s currently constructed. Yes, you can do some or most or all of it under WIOA, but with limited dollars and competing priorities, you probably won’t. If it’s going to happen, it needs to start with local public of philanthropic investment. But the strategy is pretty straightforward, and the fruit hangs relatively low.

Look: sometimes you put a term (“middle man”) into a search engine, and what comes back is a sweaty middle-aged dude who looks like he left his office job in a big hurry, now covered in blood and holding a severed appendage. In those situations, I take the hint and just include it in my story about something utterly unrelated.

Hola amigos. I know it’s been a long time since I rapped at ya, but I’ve gotten pretty busy (a good thing), plus a few ideas I had for posts either felt less like fully developed concepts than multiple children stacked up inside a trench coat,1 or would have been in violation of my general guideline (not quite a rule) that Disillusionist should focus on workforce and related issues, rather than my all-too-emo venting about politics.2

Instead I’m going to write about intermediaries, which if done wrong could look not just like biting various hands that have fed or are feeding me,3 but actually going after them with a chainsaw. So I’m going to try to do it right… and keep it high level. Any specific examples will be fictional.

Let’s start by trying to define our terms and establish a little history. As I remember it, workforce intermediaries got their glow-up in the 2000s, when people like Rich Kazis at Jobs for the Future and Bob Giloth at the Annie E. Casey Foundation4 set out their role and potential value. Analysts argued that a wide range of organizations could function as intermediaries: nonprofits, chambers of commerce, workforce boards, and so on. The main idea that Bob and other thinkers had was for workforce intermediaries to connect supply- and demand-side stakeholders—training providers and businesses—in ways that would facilitate job placement and career advancement for workers from low-income households and communities.

I’m not sure that any of the New York-based workforce intermediaries exactly fit that description—in part because the sheer scope and complexity of the place, and the way that the interests of individual stakeholders are in tension with those of the system. Most focus more on supply-side than demand-side issues, and in some ways function more like industry associations for the provider community (albeit with employer partners as well) than intermediaries as Giloth and Kazis described them.

If this sounds judgmental (and I really hope it doesn’t), I don’t intend it that way. Every intermediary has a value proposition. For the purpose of this post, I’m much more interested in the inputs that help to determine their effectiveness, both actual and perceived.

I think there are three. And they are all really, really tough—much more so than I expect anyone who hasn’t worked for an intermediary (much less more than one) is likely to appreciate.

1. Mission Clarity. What’s so difficult about that? you might ask. Surely any new organization starts by declaring why it exists and what it aims to do. And yes, that’s generally true at the outset. The challenge begins to emerge the first time different stakeholders—staff, Board members, external partners, philanthropic funders, government, the general public—disagree on what the words of the mission mean, how to pursue it, or both.

   There’s a lot of room for disagreement under the broad heading of any goal. If your objective is, say, to ensure that every high school student in New York City has a paid work experience before they graduate, people can and will disagree about the definition of “paid,” “work,” and “experience.” Then you’ll get into argument about tactics: are we supply- or demand-side focused? Which stakeholders should be doing the paying—does it count if the dollars aren’t from businesses? Shouldn’t the public school system be handling all this anyway? If so, why aren’t we pushing them harder to take this on?

   This isn’t an insurmountable problem, but it requires more or less constant attention. If you have an active, engaged, and thoughtful Board with open lines of communication to the staff—think of it as a working connection between the brain, the heart, and the hands (and note I’m not saying who’s who)—that doesn’t make disagreements disappear, but it does render them much easier to resolve. Communication with external stakeholders is a big deal as well. The goal must be to minimize how often anyone asks the common snarky question that plagues intermediaries: “What do they do, exactly?”

2. Field Buy-in. This is related to the mission issue, in that all the intermediaries’ counterparties must understand both what the entity is there to do and how it does it. But it goes further, in the sense that partners also must believe that the intermediary is effective in pursuing its mission. There should be a shared understanding of what’s been accomplished, as well as how this materially benefits the counterparties.

   Sometimes this is straightforward, if not easy. A funder will attach deliverables to a grant (or contract). A membership driven organization will articulate benefits that they either actually deliver, or don’t. Where it gets murkier is in less quantifiable impact. Are you influencing what’s on the agenda, or isn’t? Do other actors in the space take your group into account, either in terms of jumping on board in support of your efforts, or moving in because your failure or inactivity has left a void? Ideally an intermediary wants to be held in esteem, regarded as a powerful and capable actor. Short of that, being feared also has its benefits.

3. Resources. Finally, you have to keep the lights on. The economics of intermediary work are inherently unstable, as organization leaders must constantly consider the next grant cycle, emerging “market” opportunities, larger trends and so on. But if you eat everything in front of your face—if you chase every potential check—eventually you’ll poison yourself. This ties directly back to mission clarity, and mission creep… and what you don’t understand until you’re faced with it is how seductive and easy it is to justify a little stretch.

   In the worst case, you can actually lose by winning. Going back to our imaginary universal-paid-work-experience intermediary, what if there’s a grant solicitation from a national education funder to connect classroom teachers with externships that will help them incorporate “real world competencies” into their pedagogy? It’s arguably in scope, and the grant is for six figures—it would increase their budget by, say, a third. But the funder is a stickler for performative accountability, and suddenly the small staff is spending hours every month on compliance activities, with an opportunity cost of engaging with schools and businesses to set up work-based learning experiences, or conducting research that might move policymakers, or even raising other funds that might be less encumbered. Over time, mission clarity deteriorates, and buy-in declines.

I don’t have a grand concluding point here. (Remember, I haven’t done this in awhile.) But given all the time I’ve spent with and on intermediaries, and how necessary and valuable I find them to be—particularly in a complicated and “noisy” setting like New York—it seemed worthwhile to set out some thoughts.

Whether it’s human nature or a sort of performance of open-mindedness, there’s little that policy analysts seem to enjoy more than renouncing past views now considered to be misguided. So it is with “tracking,” the practice of categorizing students into different categories related to their presumed future prospects. If you’re of a certain age, you’ll remember that some kids in high school were always in the Honors or AP courses, others were in general courses, and some who “worked with their hands” or were considered “not college material” were shunted into vocational tracks.

This was usually more than a little racist, not to mention staggeringly presumptuous: imagine someone in the 1950s or the decades that followed thinking they understood the potential of a 14 year old well enough to dictate their future course. It also devalued vocational education, at least until its rebranding as career and technical education really took hold, and at the other end probably pushed a lot of young people into “a college track” when that might not have been their preferred option.

Tracking seemed dead and buried, but it’s come up recently in some surprising contexts. I was more or less accused of it in a recent LinkedIn comments thread a couple weeks ago, and then I heard NYU’s Angie Kamath, who’s probably our greatest living New York-based workforce thinker and doer,1 talking about it on the most recent episode of the NYC Workforce Drop podcast.2

As to why this might be on people’s minds, it could have something to do with how much more information is suddenly available to us to help support students and other young people.3 It’s now possible to get a fairly complete picture of a young person’s academic journey, including attendance and performance, as well as their extracurricular and work-based learning experiences, and assessments from teachers and other responsible adults very quickly and easily. The most advanced application of this in New York City is the Portal, developed by New Visions for Public Schools.

Yesterday in City & State, New York City Council Member Althea Stevens4 offered praise for the Portal and condemnation of a Mamdani administration proposal to cut funding for it:

The city runs extraordinary programs through the Administration for Children’s Services, Department of Youth and Community Development, Department of Homeless Services, our public schools, CUNY and dozens of nonprofit partners. Each of those agencies is full of people who care deeply about the children they serve. The problem has never been the people. The problem is that the systems they work in cannot see each other.

I see the cost of this every day in my district. A young person in foster care whose school doesn’t know they have a caseworker. A family in a shelter whose children’s attendance pattern is invisible to the staff who could help. A high school senior eligible for a CUNY College Now course no one ever told her about. A teenager who participated in a Summer Youth Employment Program last year and whose school has no idea the experience could count toward a graduation pathway. Tens of thousands of these gaps, every year, all of them invisible until the consequences arrive – a missed graduation, a delayed referral, a benefit that went unclaimed, a young person who ended up in a system no one wanted them in.

…

I am arguing that when a partnership with a mission-driven nonprofit produces what the Portal has produced – connecting agencies that have never been connected before, at a fraction of what an internal build would cost, with philanthropic dollars covering much of the bill – the city should treat it as exactly what it is: critical public infrastructure worth protecting and expanding.

The City Council and the Mamdani administration will soon make important decisions about where to invest in the children and families of this city. We will talk about youth employment, afterschool programs, foster care, mental health and the universal childcare promise the mayor has made central to his agenda. None of those investments will work as well as they should if the agencies delivering them can still not see each other’s work.

I’m genuinely curious why the administration is proposing to cut funding for the Portal. Years ago, I heard that some people inside the Department of Education were uncomfortable with an outside organization having built the tool and wanted to develop their own. That struck me as foolish and wasteful at the time, and still does. But it’s sadly representative of a phenomenon that’s fairly common within government:5 a sense that nothing can be trusted unless it is owned and controlled.

The most valid objection I can imagine—which brings us back to tracking—is that there’s real downside risk in having all that information available so easily. For one thing, a constant and entirely valid concern within city government is inadvertently exposing undocumented immigrants to federal authorities that would gleefully deport them. But it’s also disturbingly easy to imagine our would-be tech overlords taking all of this information and feeding it into an algorithm that would dictate the life paths of countless young people whose families have limited economic resources.6 If you’re thinking about what technofeudalism might really look like, this sure seems to fit.

I fear the tech broligarchy enough, and trust them so little, that this objection resonates with me. But in the end it sort of feels like inventing fire and then not using it because one might get burned. It’s within our control to protect data privacy and dictate the purposes for which it can be used.

If the Portal enables a provider with NYC’s Summer Youth Employment Program to learn that a newly enrolled participant took a sequence of courses in a high school CTE auto maintenance program, and that provider has a relationship with a local auto repair shop looking for extra hands, that’s probably a better result than sending the participant to work at the nearest CVS. Even better if the profile includes notes that the participant has challenges related to transportation, or needs improvement on their communication skills, or is exceptionally creative—these are all pieces of information that can lead to a more rewarding customized work experience.

Ultimately, information in the hands of well trained and conscientious adults enables us to create more valuable experiences for young people that build their skills, networks, and resilience. It’s a potential Great Equalizer, a tool to help young people and their families make meaning from disparate experiences if and when they might not have the resources—money, time, connections—to do so on their own.

As I remember well from every time I tried to build system infrastructure within government, it’s not costless or frictionless. Humans need to enter the info and make sure it’s valid, and systems must be compatible and build in necessary safeguards. Someone needs to maintain it, and adults must be responsible for understanding data as a tool to find answers rather than the answer itself.

But it’s a powerful tool, and to leave it on the shelf either from concern about misuse or as yet another sacrifice on the altar of “labor market individualism”—the idea that everyone should figure it out on their own, without any recognition of how that just perpetuates generational inequality—would be a terrible and immoral choice.

Maybe if I found five files scattered through these boxes, and compiled them…?

Recently, a client asked me to pull together a briefing memo with talking points and a fact sheet on the Workforce Innovation and Opportunity Act (WIOA) in New York State. The talking points were simple enough: I’ve written a lot about WIOA, mostly on here, to have a coherent point of view on both the law as it exists now and as I think it should be revised. As for advocacy priorities, the immediate objective is to resist further cuts and harmful policy changes, and longer-term, to meaningfully reinvest in and reimagine what workforce policy can do.

Compiling the fact sheet, however, was a total bastard—because the way the program is organized, it’s almost literally impossible to answer seemingly simple and straightforward questions like “How much money did New York get to run WIOA programs in 2025?” and “How many New Yorkers did those programs serve last year?”

You know: the questions with which not only any funder or elected official, but any mildly thoughtful layperson, would begin a conversation about the program.

I began this odyssey of profound annoyance with a seemingly simple question for my research assistant:

Claude tried to be helpful, as it always does.1 After explaining that “2025” isn’t an operative concept because WIOA Program Years run from July of the named year until June of the following year,2 it produced reasonably precise estimates of funding for Program Year 2024 for services offered under Title I (job training, employment, and related programming for Adult, Dislocated Worker, and Youth customers) and Title II (Adult Education and Literacy).3

But for Title III—the Wagner-Peyser Act , which dates back to the New Deal—we started to see some ambiguity. And it kind of got worse from there:

I maybe shouldn’t confess this given the solid possibility that my client will read this post, but I did not “pull the full PDF tables” from the Federal Register.

Ultimately, the estimate was between $540-560 million. That Title IV—which, over my 26 years in this field, I would guess I’ve talked about with people for, total, less than two hours, and can’t remember ever reading a serious brief or report regarding—is the second largest category, feels grimly apt. It’s our version of the medieval map where there’s a huge section with a drawing of a dragon surrounded by question marks.

Still trying to be helpful, Claude noted that if I required “a harder number,” I could peruse the state DOL’s WIOA Obligation/Expenditure Summary, which apparently reaches print “12-18 months after the program year”—meaning that it’s not out yet for PY 2024— “plus ACCES-VR’s [the state agency that co-administers Title IV] annual report for Title IV actuals.”

This, too, I did not pursue.

But the “how much?” question was a virtual straight line compared to the maze of madness that awaited when I asked “how many?”

Claude referred me to this site, maintained by the state DOL, which I’ve used before and is quite good—for what it is, which is a point-in-time or longitudinal summary of WIOA performance broken down by four of the six WIOA program Titles (all three of Title I, and Title III), quarter of exit, and demographic subpopulation (gender, race/ethnicity, age, educational attainment). You can ask for statewide totals, but since there’s no distinct user count, and even the definition of “participant” is super-muddy (see below) you will not receive satisfaction.

As this went on, I began to vent to a sympathetic figurative ear:

Well, fine, I thought. I can’t get a total statewide one-year participant count for all six WIOA titles. But surely I could get… something, just for the three Title I programs?

Nope.

I find there’s a grim satisfaction when I ask Claude a question it’s unable to answer. I like that there are still, apparently, limits—until I think about it for a moment and realize that the limits aren’t intrinsic to the AI, but rather to the sources it can glean.4

I noticed in the above answer that those nationwide “participant” figures looked extraordinarily low. While $5 billion—the total federal WIOA allocation, more or less—is shockingly inadequate compared both to historical precedent and current need, it’s not nothing. Taken on its face, those numbers would suggest we served 562,799 people across the three categories (Adult, DW, Youth); divide $5 billion by that number and the per participant outlay would be $8,884.17, which seems quite robust… and which a conservative Congress would never tolerate.

But—and you probably know where this is going by now—it is an extremely small subset of total “participants.” It counts only individuals who received staff-assisted services through the WIOA-mandated Career Centers. By contrast, the number for the Wagner-Peyser Employment Service—Title III of WIOA—is 2.56 million.

The NYS Department of Labor reported two more statewide figures for PY 2024 that I temperature checked for relevance: about 196,000 Career Center appointments, and around 275,000 user satisfaction surveys sent out. Neither really work: the “appointment count” does not capture unique individuals (and, I think, might not capture how many appointments were kept), and the survey number includes both those sent to businesses (around 38,000) and methodological murk around who was sent a survey. It also got a very low response rate, suggesting additional challenges.

Ultimately, Claude concluded, “The true unduplicated headcount of New Yorkers who received any meaningful WIOA-funded service in PY 2024 is almost certainly in the hundreds of thousands—but that number simply isn’t published in a clean, citable form anywhere.”

This seems like the final failure of a closed system. Call it “toxic compliance”: a regime that prioritizes preventing “waste and fraud” over serving workers and businesses. In this light, how can we be surprised that funding continues to spiral down, and that our work remains marginalized even as questions of employment become increasingly urgent?

This is actually making me (very slightly) more sympathetic to the administration and others who propose even deeper cuts: in the face of a law and set of programs that can’t articulate its value, the only rational responses are to fix it, zero it out, or effectively shrug at it by reauthorizing the status quo in hopes that it’s doing something helpful that we are simply failing to capture.

I honestly wonder how many workforce professionals, advocates, and administrators are aware of this fundamental obstacle to telling a high-level story about our work and its impact and value. As I’ve written in a different context (but also about WIOA), maybe it’s like criticizing water to a fish.

If your whole advocacy strategy is to try to communicate the human value of your services—stories of people in a given congressional district who get off public assistance and overcoming homelessness or addiction and carve out careers as medical assistants or customer service reps—then at best you’re going to stave off harm. But ultimately, without the capacity to answer basic questions—and ideally, more advanced ones like “do these services help people achieve long-term economic security—in the end, all you’ll have achieved is to lose more slowly.

The author at graduation, too tired to think about his career

This feels like a strangely shameful thing to admit for someone who’s made his career in workforce development, but when I was growing up and thinking about going to college, I don’t believe I thought much at all about how it would help me find a job or a career.

As we’ll see below, from the perspective of student motivations today, this seems odd, maybe even irresponsible: my family wasn’t wealthy, the school I went to was very expensive, and I wound up taking out very substantial loans.1 But to the extent that I thought about my undergrad experience as career preparation, it was that for most of my senior year, I planned to go to grad school to pursue a doctorate in history and eventually become a college professor.2

More than thirty years after graduating, I realize that college absolutely helped shape my ultimate career path, in two ways. Majoring in history and political science, with their heavy doses of required reading and analytical writing and class discussion, all helped me build and refine the skills that I’ve mostly used ever since. Even more important, the closest friendships I made there helped set the course of my career after coming out of grad school: one of my close friends flagged the think tank job opening that was my first role in policy (I’d previously helped connect him to a job working with me at NBC Sports).3

But again, that wasn’t why I went, and that’s not what I mostly value from it. I was thinking about this earlier this week, while reading a piece at Inside Higher Ed finding that most college graduates feel good about their choice to go:

Seventy-one percent of all surveyed alumni—and 83 percent of those who graduated before 1996—said their education was worth the price they paid for it. Asked if they would do it again, 93 percent said yes.

More than three-quarters of bachelor’s degree holders and 52 percent of associate degree holders said their degree was “critical” or “important” to reaching their financial goals. Eighty percent of bachelor’s degree holders and 55 percent of associate degree holders said the same regarding their career goals.

It’s not a newsflash to note that top-line survey results can obscure as much as they reveal. Are the respondents also satisfied with their careers? How does this break down by demographics (other than before/after graduating in 1996)? Did choice of major matter? What were the motivations or intentions of those respondents when they went to college?

The Gallup/Lumina Foundation report from which IHE pulled these stats doesn’t get into any of that, but we do have a high level sense of student motivations for going to college, and how they’ve changed over time. The gist is that between 1967 and 2018, far more students cited being financially “very well off” as a reason for going to college (42% to 82%) and far fewer identified “develop[ing] a meaningful philosophy of life” (86% to 46%).

My research assistant helpfully adds that this isn’t just about rising materialism; as everyone knows and complains about, college itself became vastly more expensive over those decades. Why this was the case is beyond my immediate interest here, but it seems noteworthy that based on that same Gallup/Lumina report, “satisfaction” wasn’t inversely proportional to cost: if anything, the opposite seems to be true.

Again, it would be helpful here to understand why they were satisfied—and for that matter, whether the satisfaction level further rose at the most “elite” schools, and if I’m really being research-greedy,4 whether the “well-off” group or the “meaningful philosophy” group were, on balance, more satisfied. But we’re stuck at the level of (pun unavoidable) educated guesses.

On balance, “satisfaction” with college, while still high, has declined over time. Here’s the share of respondents agreeing that “their education was worth the cost,” broken out by year of graduation. Strong agreement is in dark green.

It doesn’t feel like a huge stretch to guess that some relationship exists between the decline in satisfaction and the shift in student motivation from desiring a “meaningful philosophy of life” to prioritizing career success and economic security. But there are a couple other things to keep in mind as well.

The first is that today’s students look a lot different than those from the late 20th century—and that there are a lot more of them.5 Undergraduate enrollment in the United States has more than doubled since 1970, and the bulk of the increase has been among students who are:

• Age 25 and above (by 2010, this group accounted for more than a third of undergraduates)

• Non-white (45% of all postsecondary enrollment in 2022, up from 15% in 1976)

• Lower income (51% of low-income high school graduates were in college by 2014, up from 31% in 1975)

So we have different students, whose backgrounds help explain the shift in motivation: if you’re worse off financially, and/or already of working age, of course you’re more likely to care about making enough money to feed your family. And there’s nothing at all wrong with that.

The second thing is the experiential gap between the people who run colleges—administrators and faculty—and their students. Claude had nothing for me on how many faculty followed the “traditional” path from high school straight into undergrad and then on to doctoral programs. But it did find that college faculty “are up to 25 times more likely to have a parent with a PhD.”6 The people running colleges and universities clearly thrived within the classical higher education model, and it’s understandable that they would be reluctant to change it.

But eventually I think they either will have to do so, or face closure. Let’s posit that the implicit value proposition of college is twofold: you’ll have a holistic experience that will “prepare you well” for the rest of your life, and you’ll accumulate assets—not just a diploma, but also skills and a network—that will significantly bolster your employability and earning power.

Colleges seem to have delivered on this promise, more or less, through the second half of the 20th century. The GI Bill and other factors drove significant expansion even before A Nation at Risk turbocharged enrollment, and the huge increase in graduates made their presence felt in a labor market characterized by deindustrialization and the shift toward a knowledge economy. We know that college graduates are healthier, happier, and more connected to their communities, in addition to an earnings premium that still sits around 80 percent above what a high school graduate makes.

Between falling graduation rates and declining user satisfaction,7 though, the model is showing serious cracks. It’s to the point that I’m starting to wonder whether it’s time to split that value proposition into its two component parts.

A few years ago I worked on a project looking at non-college postsecondary options for young adults in New York City. At one of the most highly regarded programs I talked to, the executive director described how his organization was building a one-year version of the college experience without the usual trappings: no dorms, no quad, no sports, no degree as such, just pure education to prepare learners for a great career. I had two reactions: one, that this made sense and was more cost-efficient, given their great outcomes, than anything else I knew of; and two, that without all those things, I didn’t think you could call it “college.”

But there’s no reason why “college” has to serve both aims.8 If your goal is economic mobility, there should be an affordable solution. If your goal is holistic development, let’s solve for that. And of course it shouldn’t be, and wouldn’t be, all one or the other: a curriculum of reading instruction manuals, versus one solely comprised of Joyce and Wittgenstein.9

The counterargument will be that this reinforces class distinctions, will stigmatize students pursuing the more purely vocational options, et cetera. It’s a valid point. But the status quo, with more than a third of bachelor’s degree students failing to complete their degree (and significantly lower rates for older students, Black and Hispanic students, and part-time students) within six years, and nearly two-thirds of associate degree students not finishing within three, isn’t doing much to inaugurate the classless society either.

We could (and should, and to an extent are trying to) shift the college model to boost retention and completion, which largely would involve making changes to better support those non-traditional learners. That will always entail overcoming resistance from everyone with an emotional or financial investment in the status quo.

Or we could just figure out a better model. It could be an embrace of European-style apprenticeship, or something like the “Big Blur” concept that Jobs for the Future unveiled a few years back—essentially to combine the last two years of high school with the first two of college, overlaying explicit skills training and work experience.

Creating something new might minimize the stigma risk… though it also could be that as AI restructures entry-level employment and further underlines our need for workforce talent in blue-collar and care sectors, societal views around what work is most valued and prestigious might change anyway.

If “Career College” became a thing, that wouldn’t eliminate the value of the other function, which I think we could shorthand as “individuals maturing within a community.” To go back to my own case, I can’t really imagine that I would have made a successful transition into full-time employment at 18: I was arguably able, but I surely wasn’t ready. Over the four years in between, I started to figure out who I was and how I could function in the world, at the same time I was reading hundreds of pages a week and writing papers and all that.

Are there other structures and means by which we could help young adults “find themselves”? Of course. Maybe it’s service corps, or faith or community groups. The search for those entities is likely to hit the big obstacle of our collective/unconscious/sadistic (choose one or more) choice to have people make their way on their own. But as the cost of that path becomes clearer, I’m holding out hope that collective action can gain ground.

Probably the greatest policy researcher I’ve ever known was the late Dr. Mark Levitan. He was far more brilliant and accomplished than I could ever hope to be, but was a role model for his fearlessness, creativity, and idea that the work of his life pretty much stayed the same, whether he happened to be operating out of a nonprofit research organization, academia, or municipal government at any given time.1

Mark’s best known achievement was changing how New York City officially measured poverty, but to me he was always the guy who wrote “Out of School, Out of Work… Out of Luck?”, a groundbreaking analysis that raised the issue of “disconnected youth”—young adults who were neither in school nor working—to a place of prominence in New York City that it’s never really yielded.2 Mark pegged the number at 170,000, or roughly the population of Shreveport, Louisiana.

Released in early 2005, just as Mayor Bloomberg was starting his re-election run, the report helped inspire the Commission on Economic Opportunity, which later became the Center for Economic Opportunity,3 and which identified 16-24 year olds as a priority population for program development and investment. It also helped inspire JobsFirst NYC, which has advocated and incubated programs on behalf of high need young adult New Yorkers for 20 years now.

The issue of “disconnected youth” hit the policy sweet spot for me: the intersection of a moral offense (no decent society should shrug at hundreds of thousands of idle young adults, with all the implied despair and wasted potential), an economic hinge (they can add to the tax base, or detract from it), and a system failure (who’s responsible for this mess anyway?).

After reading Mark’s report a bunch of times, I started working on a related project, which really only spoke to the middle of those three issues. My report assessed Boomer retirements and projected workforce demand in 2006, presenting these masses of undirected young New Yorkers as a potential solution to the city’s pending need for nurses, carpenters, auto mechanics, and various other occupations.

I’m deeply embarrassed about this report now, in part for the banality of the analysis but even more for how it blithely regards young people as anonymous economic inputs; more on that below.

I got to take another run at it about fifteen years later, as our office staffed the de Blasio-era Disconnected Youth Task Force from 2019-2021.4 The final DYTF report is a bit easier to take, in that I seem to have figured two things out over the intervening years.

One, it doesn’t use the term “disconnected youth,” with its implied value judgment, but instead describes “out-of-school/out-of-work” young adults (OSOW).5 Two—and now, finally, we’re approaching the point!—it begins to tell the story in a more nuanced way than whether or not a young adult is in school and/or working.

The report includes some pretty solid research on the long-term effects of being OSOW, both on individual life outcomes and community finances. It usefully bursts the myth of OSOW young adults as a bunch of high-school dropouts: by 2018, nearly three-quarters had a high school diploma or equivalency, and one in eight had a bachelor’s degree or higher. It also shows how the number of OSOW young adult New Yorkers had gone up and down over the years, spiking through the Great Recession but steadily declining through the de Blasio years thanks to solid job growth and a rising high school graduation rate.6

In terms of my writing and thinking about the issue, though, the chart above represents the biggest step forward. It suggests that there’s a far larger universe of young adult New Yorkers at risk of becoming OSOW, which is true as far as it goes, and helpful when thinking about the issue from a system level. Indeed, I use the graphic in the narrative of the report to make a point about the importance of prevention strategies.

Where it still falls short, though, is in missing the point that’s obvious when you think about three-dimensional, flesh and blood humans rather than statistics: a young person can remain “connected” and still be in a world of hurt.

Is the point of life to be enrolled in school or employed in a job? Is that all we want for our own friends and loved ones? Of course not! But when it’s other people and their kids—kids who might not look or sound or think like us or our kids, who don’t share our interests and tastes and concerns—it is astonishingly easy to fall into that trap without really thinking about it. To decide that our obligation ends at the point we’re making reasonable efforts to ensure that young people remain “connected.”7

Yesterday I attended an event for an initiative in the city focused on better serving young adults with barriers to education and career success, at which the organizers shared some updates and led breakout group discussion. The final prompt was to share what we hoped to see for young people in NYC within five to seven years’ time. One of my table-mates suggested that the goal of the system should be to support all young people toward first understanding and then making intentional choices to pursue a life path that would make them happy.8

I do think this has to be the core motivation—or at least the single biggest input. What I suspect we really want to solve for is the Venn diagram overlap between “gives me joy” and “pays the bills.” But it has to start with the agency of young people: helping them define themselves as humans and as economic actors, and then making sure they have the tools and supports to define and then pursue their ambitions.9

Noted career-changer Roland “Mr. Prezbo” Pryzbylewski

I’m not sure where the following ideas fall on the spectrum running from “duh” to “wow.” Maybe they don’t plot neatly along that axis at all.

These ideas seem obvious to me, grounded in common sense and logic as well as evidence and a simple ability to look at the world as it is now and seems on track to become. They also would represent a comprehensive break with the main thrust of federal workforce policy for coming up on 50 years—a period during which, as we’ve established, the field’s funding and relevance consistently declined even as the world changed in ways that really should have spotlighted the value of job training and employment services.1

Just as a super-quick refresher, the status quo workforce system as governed and funded by the federal Workforce Innovation and Opportunity Act (WIOA):

• Prioritizes short-term job placement (“rapid attachment”);

• Does not emphasize job quality;

• Does not incentivize building long-term or comprehensive relationships with businesses (and views them only as “employers”—a function, not a full description); and

• By virtue of being underfunded, offers little to displaced professionals or high-skilled blue collar workers who need retraining, or the businesses that might wish to hire them.

Worse, WIOA as practiced in recent years looks less and less viable going forward, as a number of the most common training and placement options are projected to dry up as AI implementation continues. If the law is reauthorized along the lines of the most recent proposal, A Stronger Workforce for America Act of 2026, the most likely outcome is a continued slow bleed of funding, stretching local areas as they try to pivot toward evolving demand.2 ASWA Redux perpetuates all these problems, and if passed would add the opportunity cost of meaning another five years at least3 until we got another bite at the apple.

So let’s maybe try this instead.

1. Support career success, not just job placement.

2. Deliver “employment-plus” services that enable workers to thrive on the job.

3. Build relationships with businesses, not transactions with “employers.”

4. Shift the culture of workforce development from compliance to excellence.

5. Embrace the intersectionality of workforce as a field.

It’s worth looking at these five ideas all at once, because as we’ll see, they turn out to be pretty interdependent. But let’s take them one by one as well.4

1. Support career success, not just job placement. Rapid attachment—the basic premise that the first and most important task of publicly funded workforce services is to help connect someone to a job, any job, and what happens after that is mostly up to them—has been the core concept of the system since 1982. If that approach ever made sense, it doesn’t anymore: the path from the mailroom to the corner office has long since been sealed off. A new emphasis would maintain job placement as one objective, but understand it as only the beginning of the story. Through deeper engagement with the organizations that employ workers, workforce providers could better understand the challenges to retention and advancement, and deliver services that improve those outcomes. This approach benefits both the worker and the business, which saves money on turnover costs and sees its employee become more productive over time.

• What it might look like: Supporting a worker over the entirety of their career arc could take several different forms. Most obviously, it could include a more active sustained engagement with jobseekers following placement: not only tracking their retention per current WIOA metrics, but providing intentional follow-up services including career advice, peer supports (to help workers build networks), financial empowerment, and support services. It also could come through deep engagement with businesses to help them determine and address common causes of turnover, reducing related costs, and defining internal advancement pathways. Another aspect could be to incentivize higher-paying job placements, either immediately or over time: with stronger data tools, providers that place and support jobseekers who eventually reach a self-sufficiency income can be recognized and rewarded.

2. Deliver “employment-plus” services that enable workers to thrive on the job. Perhaps the biggest factor driving high turnover rates in many job titles is uncertainty in an employee’s life outside the workplace. A worker who has recurring concerns about the reliability of their childcare or being able to get to their place of work is far more likely to miss shifts or otherwise struggle on the job, and risk getting fired as a result. Just as the workforce field is realizing that wraparound services are vital for jobseekers in training programs, the next step is to ensure that they remain available beyond the point of placement. The childcare shortfall alone is an enormous drag on the labor force, which WIOA does somewhat north of nothing to address:

• What it might look like: Closely aligned to Premise #1, the idea here is to create the enabling conditions for workers to stay on the job and advance over time. Ultimately, this means a stronger safety net that leans into an ethos of care and strengthens the nonprofit/social services sector. Operationally, this could include grants that attach to newly placed workers—vouchers for childcare or other supports—and/or working with providers, employers, or outside actors who can negotiate a higher volume of support services at a lower price point.

3. Build relationships with businesses, not transactions with “employers.” All too often, workforce providers approach businesses with the same transparent intentionality of Beavis and Butthead talking to girls. It’s purely transactional: they ask about current job openings, or present a menu of services, without considering the organization’s deeper circumstances. These interactions should instead be grounded in efforts to understand the potential business partner’s objectives and pain points holistically—viewing their talent strategies and processes in the same context that businesses themselves see them. This mindset shift is necessary if we are to evolve beyond placement and toward embracing worker career success and shared prosperity as system priorities. It also might lead to fewer figurative slaps in the face.

• What it might look like: This change is the simplest in theory and perhaps the hardest in practice, because it involves culture, norms, and motivations. At least since JTPA, the incentives of federal workforce policy have been such that providers conduct “employer engagement” with an eye toward the outcome, most often sourcing candidates for job openings on the adult side or securing commitments to host interns or seasonal workers on the youth side.

  But it’s one thing to assert that the field should consider businesses in the entirety of their operations rather than through the lens of their function as “employers,” and another to actually do that, unless and until the incentives change. This is why new policy that creates incentives for that shift, in the form of both a wider range of worker outcomes (retention, advancement over a longer timespan) and new metrics (say, bonuses for sustained engagement with high-road employers) is absolutely necessary.

4. Shift the culture of workforce development from compliance to excellence. After decades of dwindling resources and circumscribed goals, public administrators and providers in the workforce ecosystem have adopted a scarcity mindset and a deficit focus. Particularly in the face of federal threats to already-scarce dollars, looking to avoid further harm is a rational response. Yet, this mindset contributes to the vicious cycle of diminishing value and growing irrelevance. The field must embrace bravery and creativity—and policymakers must make it safe for practitioners to be bold on behalf of workers, jobseekers, and businesses.

• What it might look like: In current workforce policy, “compliance” functions like one of the greatest lines from The Wire: “No one wins. One side just loses more slowly.”5 As the system continues to feed low-wage workers into low-wage, high-turnover jobs, providers meet their numbers, low-road employers enjoy outsourced HR, workers continue to churn at the bottom of the labor market, and Congress progressively starves the program of funding—but “more slowly.”

  A different approach would look much more like what Rachel Lipson called for, in one of the best pieces written about workforce last year: “treating human capital like research and development.” In R&D, “failure” doesn’t mean you lose funding, much less that your narrow-margin provider ceases to exist; it’s part of the process, from which you learn and move on. Understanding workforce as an investment in shared economic prosperity rather than the policy equivalent of an OTC painkiller for structural poverty, our whole perspective would shift in ways that, as Rachel puts it, might “enhance national competitiveness, [and] expand opportunity for Americans historically left out of the innovation economy.”

5. Embrace the intersectionality of workforce. Work is a nearly universal human experience, in that almost everyone is, has been, or will be employed. Workforce development sits at the junction of economic development and education, and touches issues ranging from housing and social services to criminal justice and healthcare. This vast reach offers enormous potential to form coalitions and pursue creative solutions. Recognizing this intersectionality as a strength to leverage, rather than an obstacle to navigate around, can build momentum and supercharge advocacy.

• What it might look like: If, as Nick Beadle has it, “Everything is workforce,” that suggests both a challenge and an opportunity. Policy issues without an obvious institutional home or population focus tend to get orphaned: everybody’s problem is nobody’s priority. Add in that “workforce development” is not as immediately comprehensible in terms of what it is and what it does as “education” or even “social services.”6 But the opportunity is that there’s almost nothing that doesn’t have workforce implications. A unifying story—around, say, maximizing dignity, security, and mobility in every job—can suddenly cohere a vast coalition of active and passive supporters.

Now, as we’ve covered previously, how much money is available for workforce programming probably matters much, much more than the particulars of legislation. The smarter WIOA defenders I talk to likely would point out that there’s little if anything in the above that the current law straight up wouldn’t allow.

Which, fair enough. But I also think that what’s in the legislation—who it’s understood to serve, and to what ends—has a lot to do with congressional appetite to fund these programs. A couple other relevant points are that, one, it doesn’t seem like many (most?) WIOA administrators are fully aware of the flexibility they have, and two, given their limited dollars and accountability obligations, even those who are aware of their options might not be in a position to pursue them.

I haven’t really asked for comments to any of these posts—not that I wouldn’t always love and cherish them—but if anyone has feedback on these five ideas, I’d love to hear them. This post, and pretty much everything I do here, should be understood as an effort to articulate my thinking. That’s always and rightly an iterative process, and life has proven me wrong about things with sufficient frequency and intensity that I’ve gotten relatively comfortable with changing my mind.

A principal who’d probably agree with this principle

Hey folks! This is Part 1 of a two-part series on federal workforce policy. It presents my current thinking on why this system is the way it is. It does not do what I’d previously planned, which is really dig into down the whats and whys and wherefores of present (much less past) policy. I couldn’t get past the reality that there are many, many people who do this on the regular better than I could, including but not limited to The Century Foundation’s outstanding analysis of the Workforce Innovation and Opportunity Act from last summer, and ongoing coverage by the National Skills Coalition, and Nick Beadle’s Jobs That Work newsletter. There’s a high-level summary toward the end here, but I encourage you to dig in deeply at those sites.

Part 2, coming early next week, will present five principles for rebuilding the system. This is a conversation I’m excited to follow and hoping to find ways engage in over the next year-plus, as initiatives like NSC’s A New Promise of Work raise the salience of these issues, platform new thinking, and inject some needed energy and urgency.

Once in awhile I’ll be working on a document, making comments to something a colleague shared, and out will pop an idea I subsequently realize must have been rattling around in my head for a long time but hadn’t quite found its way into words.

I had one of those earlier today, in the context of a possible project focusing on career readiness for young adults:

We could talk here about the mostly unexamined/unconscious premise at the root of this: that while society has some interest in public education, subsequent labor market outcomes are left to the individual, who is obligated to navigate systems and make meaning [from their experiences] on their own. Based on our policy choices, one could reasonably conclude that we value this principle so much that it justifies otherwise-needless wasted economic potential and human misery.

One way that I understand the history of social policy in the U.S. is as an ongoing argument between individual versus collective responsibilities. Individualism got off to an enormous head start, both because the Western world until the Enlightenment had very little sense of community outside of religious groups, and not much more for a long time after that, and because there was just so much space in North America.1

Since then, collective responsibility gradually has caught up—in part on values grounds, and in part owing to economic efficiencies. For example, three hundred years ago, there was very little public education to speak of; if you weren’t born into a wealthy household, you might or might not be literate,2 but you almost certainly weren’t “educated” as the word would have been understood then.

But as the labor market began to evolve beyond the simple categories of “farmer, soldier, or artisan,” the value of literacy to society rose, and communities began to make provision for education. It was simply more efficient for a community to employ teachers who could work with groups of children, and to figure out how they would pay for this, than for everyone to do so on their own. By the early 20th century, public education through high school was nearly universally available in the United States.3

As K-12 schooling came to be understood as the government’s responsibility both in terms of providing it and ensuring quality,4 education took its place along with old age insurance, unemployment support, regulation of markets in the public interest, and much more as markers of a civilized society. The through-line of these innovations and interventions was that a small-d democratic government had real stakes in making sure that a universal foundation of opportunity, and a safety net for those who struggled, were both in place.

Once they reached adulthood and had to work, however, most people were still on their own. While the norm was no longer sturdy yeoman farmers taming the vast wilderness, the expectation remained that Americans would find their way through the labor market.

There were exceptions, particularly during the Great Depression, as I wrote about last week: government employed millions through the Civilian Conservation Corps, and then substantially took over the economy to fight World War II. Direct public sector employment diminished after that, but early federal workforce policy, at least through the Comprehensive Employment and Training Act (CETA), seems to have carried forward the New Deal concept that government had some role to play as an employer.

All that began to break down in the late 1970s, as public skepticism of government spending and programming—really, of that idea that we had collective responsibilities—came to dominate American politics, culminating in the 1980 election of Ronald Reagan. I don’t think job training and employment services were a top target of the Reagan Revolution… but it wasn’t an enormous conceptual leap from Reagan’s (deeply misleading) demonization of so-called welfare queens to the conclusion that job training was a dodge for parasites.5

At the least, we took pretty severe collateral damage from the idea that “government is the problem.” Back by popular demand, it’s the table showing actual workforce spending from 1979 to 2024 vs. what we would have spent if funding had kept place with overall federal spending growth!

I used to argue that one goal of the workforce system was to give jobseekers and workers the tools they needed to navigate the labor market on their own. I don’t repudiate this, exactly, but I now consider it incomplete. As the labor market gets more challenging and inhospitable—not only for workers, but for employers as well—stronger and bolder interventions seem called for. This is because the premises used to justify the decline, including that workers once employed could make their way toward career advancement and higher wages through union support or employer investment in skills training, are no longer valid.

And yet those premises continue to guide the Workforce Innovation and Opportunity Act (WIOA), as they did its two predecessor laws: the Job Training Partnership Act (JTPA), and the Workforce Investment Act (WIA). Having covered this before, I won’t go through it all again, but to summarize, I think it’s fair to say that WIOA:

• Prioritizes short-term job placement (“rapid attachment”)

• Does not emphasize job quality

• Does not incentivize building long-term or comprehensive relationships with employers

• By virtue of being underfunded, has little to offer displaced professionals or high-skilled blue collar workers who need retraining6—which also means that the law holds limited value for businesses seeking talent in those categories

These realities are the result of choices, active and passive, that have accumulated since Congress tossed CETA aside in favor of JTPA’s “you’re on your own” approach in the early 1980s. I don’t think it’s a coincidence that funding has trended down since then—or that evaluations of JTPA and WIA in particular found that the programs mostly didn’t work all that well, which of course further sapped Congressional interest in funding it.7 Federal workforce became, to quote an ugly cliché, “a poor program for poor people.”

With the spending counterfactual in mind, today I asked Claude a different question: if we had funded workforce programs to keep pace with overall spending growth since 1979, what might that have meant for American workers?

Claude kind of hedged in response, pointing out that such a modeling exercise requires a bunch of entirely hypothetical assumptions about service delivery mechanisms, effectiveness rates, and exogenous factors like unionization decline that were happening at the same time.8 But in the end, drawing upon research literature that probably would have taken me a solid month to get through, we got some data:

Statistics aren’t people. But the idea of Americans getting a cumulative additional 9.8 million years of life (based on research finding that losing a job roughly doubles short-term mortality risk and keeps it higher for years to follow), and 2.1 million of us who were actually incarcerated avoiding that fate… the mind slides off of those numbers.

The wealth gains are, of course, quite significant as well—even before we consider how greater economic security might have factored into democratic participation, social connectedness, housing and food security, educational attainment… pretty much any input you can imagine into a life worth living.

“Fine,” says the hard-headed government actuary. “People would have had more home-cooked meals and holidays with their families and other Kodak Moments. What would it have meant for the public fisc?”

Well, I asked that too:

Claude doesn’t just run these numbers, but includes commentary as well—in this case, the observation that even the most conservative fiscal ROI scenario suggests that every unspent workforce dollar would have yielded a return of $1.80, “making the funding gap look less like a savings and more like a compounding loss.”

The question now is whether we’re going to continue digging the hole deeper, or start to climb out. The first principle that undergirds the decision to disinvest in workforce development—that when it comes to jobs and careers, everyone is on their own—has imposed a staggering cost. I presume that few if any people understand this cost. Maybe that’s for the best, as I can’t imagine how anyone could argue with a straight face that it’s been worth it.

We probably shouldn’t use this rather creepy image for a revised Corps model.

One big theme I’m seeing right now in policy circles is concern about labor market entry and early-career progress. I’ve written about this here and here, as have the always-great Martha Ross at Work Shift and Nick Beadle at Jobs That Work. Not ten minutes ago I read Gad Levanon on LinkedIn with an analysis of how young four-year college degree holders are contending with historically higher-than-average unemployment. Nirit Cohen at Forbes is on this too. This was an underlying theme of the brilliant Normal Gets Us Nowhere series. The New York Times now almost seems to have a dedicated beat for this subject.

Everyone’s diagnoses more or less converge: young people, both college graduates and not, are struggling to access career-track employment opportunities. AI is probably a driver, particularly in white-collar fields. In blue-collar industries, the challenges are a bit more varied—high tariffs, interest rate uncertainty—and perhaps temporary. Bound up in all of it, I’d add as a close and longtime observer, is a dearth of adequate and effective career advising: too few young adults and their families are fully aware of the options available to them, much less well enough informed to make comparisons as to costs and benefits.

What I haven’t seen much of, beyond the general sentiment that “we have to do something!,” is a specific plan of action. The exception, and what I want to drill down on, is from the first piece I linked, by Brookings Institution researcher Martha Ross. She suggests that we look to Service and Conservation Corps programs:

Service and Conservation Corps are local organizations that engage young adults, typically ages 16 to 30, in service projects addressing a variety of community needs, including land and water management, infrastructure, and disaster response. Corpsmembers build and maintain trails, restore habitats, respond to wildfires and floods, plant trees, weatherproof homes, and more. These are paid, hands-on work experiences in which corpsmembers provide real benefits to their communities while gaining valuable skills.

And corps projects are natural stepping stones to careers in related fields. My colleague Joseph W. Kane and I identified more than 80 relevant occupations with median wages of nearly $30 per hour—more than four times the federal minimum wage.

…

Corps-related occupations have relatively low barriers to entry—about half of workers in those occupations have a high school diploma or less, and workers often gain job-specific skills through on-the-job training or apprenticeships.

Of course, these programs are the modern day successors to the New Deal-era Civilian Conservation Corps, which employed around three million young men ages 18 to 25 on public improvement projects between 1933 and 1942. At a time of 25 percent unemployment, Corps members were paid modest (very modest) wages, most of which went home to support their families, in addition to free room and board. Some of the CCC infrastructure was shifted to explicit military preparedness after World War II broke out and President Roosevelt accelerated rearmament.1

Martha frames the modern Corps concept as a potential solve for the 4.2 million Americans ages 16 to 24 who are neither in school nor working, sometimes referred to as “disconnected youth” or “opportunity youth.”2 I don’t disagree, but 1) I have a problem with this typology generally because “connection” is a spectrum rather than a binary,3 and 2) I think the benefits of a Corps approach might be no less applicable to young people who are technically “connected” to school and/or work, but might not perceive themselves as firmly on track toward career success or economic security.

As Martha notes, some of this exists already. She points to a few programmatic examples in the piece, and cites a return-on-investment analysis for corps programs that found $7 in value for every $1 invested. Beyond the benefits of rebuilding community infrastructure and pathways to careers that she names, I think there’s a strong argument to be made that working in community with other young adults would help participants build transferable skills (collaboration, communication, creativity, critical thinking, et al) and networks.4 Perhaps it’s also a partial antidote to the loneliness epidemic that is especially pervasive, and especially damaging, among young adults.

So my first thought reading the Work Shift piece was, “Yep, makes sense.” The thought that immediately followed was “Never gonna happen at scale, because it’s too Big Gummit.” At some later point, though, I started reconsidering that conclusion.

I came of age in the 1990s, when a Democratic president asserted that “The era of big government is over,” and then went out of his way to prove the point.5 As I wrote a couple months back, federal workforce policy mostly got out of the direct employment business in 1982, when the Job Training Partnership Act replaced the Comprehensive Employment and Training Act. Even something as innocuous as the School to Work Opportunities Act, passed in 1994 to expand options for local areas looking to more closely connect education to employment paths, drew criticism from right-wing Republicans as sinister government incursions into the free market.

So maybe there’s not much appetite for the idea of mass spending on public sector employment. AmeriCorps, probably the biggest single federal program in this vein, received $1.26 billion in funding for Fiscal Year 2024. For context, that’s between about half a day and two and a half days of spending for our current paused war against Iran, depending on what estimate you use for the costs.

But then again, maybe there is. One effect of Trump’s rampage through American public life over this last decade is that he has severely, perhaps mortally, wounded small-government conservatism. Even during the Biden interregnum, Democrats got much less scared of both industrial policy and large price tags. If they manage to regain control of the checkbook, I could imagine then getting behind a big plan that could address urgent needs related not just to infrastructure, but perhaps also to public health and childcare—while creating new paths into occupations and sectors with strong demand going forward, and easing the pressure of AI implementation on a labor market that is clearly not ready for it.

With these thoughts bouncing around in my head, I worked with Claude to build out a ten-year “C-Corps” program in NYC with tracks in “Care” (childcare), “Conservation” (or construction—but the focus is built environment and sustainability), and “Community” (public health). These are all areas of need, offering employment pathways into which participants could seamlessly transition upon completion of their program experience.

My pilot would scale up to 15,000 young adults per year, 5,000 in each track, with appropriate staffing and support services. That last is important, because we’re prioritizing high-need populations including justice-involved young adults, those in foster care, and young New Yorkers with disabilities. Participants would be compensated at minimum wage ($17.00 in NYC) for 25 hours per week for a full year,6 plus completion stipends of $5,000 each, as well as 8 CUNY credits awarded through Credit for Prior Learning. I worked out specific, complementary roles for employers, government, community-based organizations, philanthropy, and higher education (CUNY). Here’s a rough graphic:7

The program would serve 120,000 young adult New Yorkers over that span, and a reasonable estimate of its likely ROI yields $2.30 for every dollar spent.

The total cost of this over ten years? $7 billion. Or between 3.5 and 14 days of our Iran War spending. Think that might be a politically resonant argument?

It’s coming right at us!

I have promised to fully unpack my thinking on the Workforce Innovation and Opportunity Act (WIOA),1 the federal law that sets the basic premises for and substantially funds workforce development in the United States. Thinking about how to do this, I had an idea for a series of three posts. The first one would be on what the system is and does: the organization of the law, how it evolved from predecessor legislation,2 the metrics and outcomes by which it judges performance, and its culture and mindset. The second would feature a set of research questions that I think policymakers should look into as they consider potential changes to WIOA, and highlight a few state-level systems from which a new federal policy framework might take inspiration. The final post advances five basic principles that I think should inform our policymaking around workforce development.

I might or might not follow through with this plan.3 But while I was noodling how to lay out my thoughts, it occurred to me to ask a more immediate question—one prompted in part by some recent super compelling research that came out of the Brookings Institution and Opportunity@Work: how exposed are the most common job titles that people access through WIOA services might be to AI?

The answer is… a lot. The table below ranks occupations by volume of WIOA-paid training, noting “AI exposure”—the extent to which the technology is likely to impact work done under that occupation—in the second column.

This presents the top 14 occupations by WIOA training volume, as of Program Year 2021 and drawn from USDOL sources. “Boosted” means that AI is actually increasing demand, largely owing to the construction of data centers and the related need for electricians and welders and HVAC techs. “Low exposure” means that AI is largely irrelevant to current hiring and projected demand.4 “Evolving” is a bit more ambiguous, but I understand it as projecting a likelihood that AI will change the description and day to day tasks of the role, and thus the preparation and credentialing required, but not demand itself. (For example, new diagnostic tools and information resources will alter how licensed practical nurses deliver care, hopefully for the better, but won’t replace them.)

Four occupations among the top 11—CDL/heavy truck driver, customer service representative, office admin support, and medical billing/coding—are at either “high” or “critical” risk in terms of exposure to AI. They combine to serve an estimated 31,000-54,000 trainees per year. That is, per the note at the bottom of the table, potentially more than 20 percent of all participants in WIOA-funded adult and dislocated worker training.

If you look at average wages, these are probably about median-quality WIOA placements. Their training costs are on the low side, compared to others in the table—in fact, with the exception of CDL, they’re all the cheapest.

So, from the perspective of a WIOA contracted provider or local administrator, these four job titles have represented a strong value proposition. Strong (past) demand, modest upfront expense avoiding undue strain on the painfully inadequate local WIOA budget, average wages that won’t lift someone into a high tax bracket but, depending on where you live, might come close to meeting expenses.

And now those jobs are going away. Anyone who counsels a jobseeker into training in one of those four sections should do so with the warning that they are signing up for a high likelihood of coming right back to seek services again in a year or two or three.

Will they give that warning? The imperatives of the program from the perspective of the provider or local administrator—the need to hit volume goals for job placements—are in direct tension with the long-term best interest of the participant seeking training. Claude makes the point:5

If that’s going away, how shall the system respond? I’d suggest the answer is to emphasize the sectors where we see positive projected hiring demand and relatively higher average wages and limited exposure to AI in terms of risk. Among the top 14 listed above, eight occupations fit these three criteria.

So here are eight occupations, all relatively AI-resistant, all offering median annual pay between $42,000 and $64,000 per year. (I omitted Nursing Assistant/CNA and Home Health/Personal Care Aide, which are job titles with enormous demand that offer poverty wages. Again, those jobs are perhaps best understood as a volume play for system actors. But until they pay a living wage, they shouldn’t be a priority option.) Note that the median wage across all occupations, per Claude, is $49,500.

I should note also that I asked Claude to show growth, not total openings—which is the more relevant number for planning purposes. The next phase of the analysis would include that, plus perhaps the median age of workers in these occupations or share of workers over age 50. One of the more notable reports I wrote as a think tank researcher6 looked at this issue of demographics-driven replacement,7 and it’s… I’ll go with “interesting” that so many of the same fields stick out for projected demand, even through all the economic twists and turns of the last two decades.

More to come, but I think the takeaway here is that WIOA as currently structured and targeted will become less and less viable as AI continues to transform the labor market and take out so many of the go-to occupational targets for trainees, providers, and local administrators.

The alpha and omega of American consensus

The Trump administration just released its “budget topline” for Fiscal Year 2027. The news coverage is focused on the mammoth $1.5 trillion request for new military spending,1 which I guess is necessary when you deplete the armaments inventory at a few million a pop to dismantle orders-of-magnitude-cheaper drones after starting a war of choice based on, so far as anyone can tell, vibes and/or boredom.

But I’m a workforce guy, so what I notice—with a little tickle of dread that some minion of Russell Vought must have felt that this merited inclusion in a four-page press release—is this:

The MASA proposal2 first emerged last year. As I wrote about here, it would have merged 13 distinct federal funding streams into a block grant, while cutting their combined funding by about $1.5 billion. Nick Beadle and others noted that it didn’t find much traction in Congress last year, when Trump’s political position was quite a bit stronger than it is now and the midterm elections were an abstract concern rather than a rapidly approaching hazard. So I’m not super concerned that it will become law in 2026.

Still, it illustrates a larger point that I’ve been thinking about off and on since Trump’s second term began: this administration breaks with a bipartisan consensus of more than 90 years in not really prioritizing economic growth. I wouldn’t say they don’t care about it at all; they’d rather have it than not, I’m sure. They just care about other things much, much more.

From Franklin Roosevelt through Joe Biden, including Lyndon Johnson and Barack Obama and Richard Nixon and Ronald Reagan, every presidential administration (and every session of Congress) has shared two overarching objectives. The first is that the American economy should grow, and the second is that the benefits of that growth should be, to a lesser or greater extent, equitably shared—or at least should provide some floor below which even the impoverished could not easily sink.3

I don’t think you can understand much about modern American history outside of this basic concept. The evolving dynamic between those two objectives informed everything from the Marshall Plan to the civil rights movement and Great Society to Reaganomics and Obamacare.

The parties obviously had different favored approaches to driving growth: investments versus tax cuts, spending on social services versus block grants and local control. Just as obviously, they would favor one of the goals over the other based on partisan preference and changing conditions. But economic growth and a baseline of economic security were always the landmarks by which they navigated.

Not the Trump administration. Whether because of Russell Vought’s anti-government zealotry,4 Elon Musk’s endless weirdness about anyone who isn’t a straight white man, or just the accumulated brain poisoning of a bunch of people who have been watching Fox News for decades, these people have other priorities.

Look at the above language again: they want to block grant workforce funds “instead of funneling taxpayer dollars to progressive nonprofits finding work for illegal immigrants or focusing on diversity, equity, and inclusion (DEI).”

Now think about the premises here: that serving “illegal immigrants”5 is somehow zero-sum for American workers and not additive to economic growth (as the evidence suggests it is), and that DEI impairs rather than accelerates growth (which research also rejects).

One can see this dynamic play out again and again in the administration’s policy priorities.6 Its immigration policy as a whole is depressing economic output—and even hurting native-born American workers. Its disinvestment in the sustainability economy has put the country far behind China and other competitors, while raising prices for American consumers. The administration’s war on higher education and related cuts in federal research spending is going to hamstring innovation and related economic growth in ways that we probably can’t even imagine yet.

I could go on. The dismantling of the federal bureaucracy without any plan or clear goal other than dominance and inflicting trauma, with the destruction of the federal Department of Education as the rancid cherry on top. The tariff regime. Trading strong relationships with vibrant market-based economies for closer alignment with authoritarian states that mostly make their money from extraction. Effectively ending prosecution of white-collar crime.

None of it contributes to growth. And note that I haven’t even mentioned the gutting of the safety net, which will have severe effects not only in terms of human suffering7 but impacting state and local budgets as communities that aren’t (as) indifferent to that suffering try to adjust, and bear the opportunity costs of doing so.

In a functioning information environment, I feel like this all would be a bigger deal than it is. Yes, it’s a cliché that a lot of Republican voters pull the lever against their own economic interests. But even among Trump voters, I can’t imagine that a majority would truly favor racial essentialism and cultural grievance over prosperity.

As a teaser, and after rereading this amazing analysis from the Century Foundation released last year, I’m not quite as WIOA delenda est1 as I’ve previously represented. I still maintain that it’s a failed policy—but I’d add an asterisk, which is that there’s a big conceptual gap between what it does well (labor market entry) and what we as an economy and a society need workforce policy to do (worker economic mobility, firm competitiveness). In terms of a forward approach, this means less a gutting of the current WIOA structure, and more the urgent need to add a bunch of thoughtfully designed and well-built floors on top of what’s in place.

Of course, this is only even possible if we actually fund the damn thing. After rereading that TCF piece, I got to thinking about money, starting with a point I’ve referenced before that Professor Harry Holzer made to Congress a few years ago: federal spending on workforce programs reached its all-time high in 1979, and has steadily fallen since.

Thus I went to the Dreaded Apparatus and asked it to make a graphic of all federal investments since 1979 in workforce development, K-12 education, housing, and transportation, with trend lines:

Yep, that’s us on the bottom, in thematically appropriate red. Workforce actually got more than K-12 education in 1979, or more than a third of what the federal government spent for transportation and a quarter of total spend for housing. Forty-five years later, workforce got about one-nineteenth of K-12, one-twelfth of housing (which itself was viciously cut in the ‘80s and roughly flat since), and about one twenty-eighth of transportation.

Then for funsies, I added all federal spending:

In this (post-Reagan) reality, total federal spending more than quadrupled over that period, while workforce obviously cratered. But… what if it had gone the other way? I asked It to show what workforce spending would look like if had risen at the same pace as overall federal spending.

Holy mackerel.2 If WIOA were funded at $97 billion per year, that presumably means a vastly higher cap on training vouchers, substantial funding to support sector consortia, way more dollars available to pay for enabling conditions of work like childcare and transportation… well, it would make the Springfield Elementary oil windfall look like finding spare change in the couch cushions.

Because I’m apparently some kind of masochist, I then asked to see this in table form, with percentage changes:

I should note here an important point from the Century Foundation piece, which is that they detail how, beginning in the 1970s and ‘80s, the federal government began making fairly significant workforce investments in legislation and through agencies other than the successive workforce laws administered by the Department of Labor.3 The total they cite, from the Progressive Policy Institute, is $28.2 billion across all programs. As the table above shows, nominal investment in 1979 would have been equivalent to $22 billion in 2023—so if you count all of that spending, in things like Supplemental Nutrition Assistance Program Employment & Training, or a whole raft of Biden-era investments that have probably since been rescinded… you’d say that we’ve just barely outpaced inflation.

But wait! Holzer’s other point big point was that the plunge in dedicated workforce spending came even as the the labor force had grown by 50 percent since 1979. So I asked for that to be added in:

Claude thoughtfully noted4 that the labor force has actually grown by 61 percent since ’79. In combination with federal workforce investments plunging by more than three-quarters, actual spending per labor force participant has fallen in constant dollars from around $210 per worker back then, to about $30 today. I guess if you used PPI’s number for total workforce funding—which again, outside of WIOA, is usually very population-specific and otherwise limited—the number would be $170 or so.

The last chart is workforce spending (WIOA) as a percentage of the total:

So we’re trapped in a vicious cycle of underinvestment, underperformance, and ever greater irrelevance. As I’ve set out in earlier posts, I think there are reasons this might not be the case in the future. But we need to start making arguments that resonate.

My last question to Claude was to summarize research around ROI on federal workforce spending. To summarize the summary: it’s complicated, and the evidence is mixed. But the investments that seem to yield the strongest return for both participants and taxpayers are intensive support services, and advanced training—the interventions that tend to cost more.

Not that spending alone is sufficient: program design must be on point as well. And Claude added a point that almost makes me suspicious he’s buttering me up:

Which of course brings us back to *how* we need to redesign this system—a question we’ll leave for another day.

Ok, fine: I just did a Google image search for the title of this post, this came up, and I dig it.

One of my go-to phrases early during my last job in city government was that we aimed to achieve “maximum progress with minimum disruption.” Implicit in this idea was that excessive disruption, which we’ll define as significantly altering what people outside one’s own team primarily do in their daily work, and/or how they do it, carries an opportunity cost that is unlikely to be justified by what’s gained as a result.

Eventually I moved off this view, at least somewhat, as I came to understand that the sort of system change we were trying to build required fundamental reconsideration of the objective itself. Where partners across the government might have previously understood their jobs as delivering excellence at a program level, we needed them to think about the cumulative impact that programs were having on the young people they served. The real question wasn’t whether a program could be fully enrolled, or retain participants over a certain percentage all the way through; it was how much value that experience was providing for their long-term prospects of career success and economic opportunity.

This didn’t require a full-on shift in worldview for every program manager across every agency with a role in youth development and career readiness, much less a total change in their day-to-day activities. But it was something we wanted everyone to have in the back of their heads, and for a few key people in each partnering agency, it had to be top of mind. How does a given experience fit into the larger picture, connect a young person to their next step, or enable them to fill the conceptual and experiential toolkit we believed they needed for the long term?

The importance of the change had to be strong enough to overcome the inertia that, as I’ve written before, is the most powerful force in public policy.1

I’m thinking about this in the context of a couple things I’ve read recently that I understand to be calling for a more or less comprehensive rethink of the American public education model. First was the finale, or at least I think the dramatic climax, of a series of blog posts from All4Ed they’ve branded as Normal Gets Us Nowhere.

In this series, two longtime career pathways practitioners2 more or less interrogate their own presumptions and prescriptions related to career-connected learning, and they reach some disquieting conclusions. “Pathways” turns out to be another failed attempt at a shortcut to career success, which remixes old patterns of attempting to use education to solve economic or political problems, classifying students as higher-level thinkers or lower-level functionaries—an exercise that, you guessed it, is and always has been heavily coded by race and class—and changing the rules, the incentives, and the resources at any sign of progress toward true equity.

They write:

[John] Dewey argued that you can’t separate vocational and liberal arts education without creating the conditions for what he called “social predestination.” Our current elite version of career-focused education avoids exactly the separation Dewey warned against. Law school, medical school, Ph.D. programs: all have an explicit vocational focus, all require bachelor’s degrees that include a grounding in the liberal arts, and all integrate broad intellectual formation—medical ethics, legal reasoning—alongside career preparation. Nobody stripped the liberal arts out of the pathway to becoming a doctor or lawyer. But when the pathways movement talks about career-connected learning, we’re almost always talking about the other kind of vocational education: the stripped-down version that prioritizes efficiency over exploration and uses language about ROI to insist that the young people furthest from opportunity need to finish their educations and get jobs as quickly as possible.

The debate isn’t about vocational education versus liberal arts (or the closely related false choice between credentials and bachelor’s degrees). It’s about who gets the version that includes both and who gets left out.

Emphasis mine. Obviously, and as Charlotte and Kyle eloquently articulate, we want a society in which every learner gets both. They offer an outline of ideas to build toward this system, including more explicit targeting of bachelor’s degrees within pathways, insistence on credentials that deliver measurable value (a focus of one of their earlier posts), building liberal arts competencies like critical thinking into pathways (e.g. an IT sequence that includes consideration of AI governance or social media’s impact), and centering career exploration—including, critically, room for young people to change their minds.

I am conceptually onboard with all of this. Hell, I’m driving the train. I have long believed that beyond foundational literacy and numeracy, probably the most important things for young people to learn are economics, so they can understand the world, and philosophy, so they can hopefully understand themselves.3 If that was the focus through, say, middle school, by the teen years students might be ready for a deeper dive into whatever subject or subjects most interested them, and then move to higher education or advanced training. It could all work if you have strong tracking of progress not just on academics but social-emotional development, career exploration and experiential learning, and wellness. You’d also want or need a rethink of the social contract to include employers operating under the expectation that they have a stake in youth development and career readiness.

The blog concludes by noting that a group of national pathways leaders has recently called for a shift toward greater emphasis on “purpose, belonging, and social capital.” Again, this is squarely in line with things I’ve thought and written about for years: the vital importance of facilitating “informed and intentional decision-making” by students and families.

It echoes another fascinating report I read a few weeks ago that calls for reimagining high school so that 50 percent of learning is experience-based. The report argues that “disengagement is not a marginal issue in American schools—it is the default experience.” The path to greater engagement and higher motivation, the authors argue, is to center experiential learning. I read it as a fun thought experiment, but—and we’re getting to the dead hand pretty soon—probably not plausible as a model.

The lead author of that report released another essay this week that arguably draws the line from the provocation of her “50 percent” paper, to All4Ed’s call for a systematic rethink. It directly names the overarching problem that I think the field has been edging toward for years now: K-12 education is preparing students for an economy, and a world, that is almost gone. Jobs that feature “routine, rules-based work” will disappear:

What remains looks more like the opposite. Work where the problem is not neatly defined in advance. Work where there is no clear script to follow. Work that requires judgment about what matters, initiative about what to do next, taste about what good looks like, and the ability to move forward when the instructions run out.

…

Our schools and universities are still far too oriented around producing students who can follow directions, absorb content, and give the right answer. We are much less good at helping them identify a problem, set a goal, marshal resources, and learn their way toward a solution.

To be sure, this isn’t a new thought. All the way back in 2005, Bill Gates said, “Training the workforce of tomorrow with the high schools of today is like trying to teach kids about today’s computers on a 50-year-old mainframe. It’s the wrong tool for the times.”4 Students then are now the (nearly middle-aged!) workforce of today, and while schools are far more technology-enabled now than was the case back when the Red Sox had only just broken the Curse of the Bambino,5 it’s fair to assert that what they’re teaching with that tech is more similar than not.

And this is where the dead hand rises. If we’ve seen this coming for more than 20 years, if tremendously influential philanthropists and industry leaders like Bill Gates (or, in New York City, Jamie Dimon) are banging the table for change and putting money on the table to make it so, if the federal Department of Education6 under both Margaret Spellings (Bush 43) and Arne Duncan and John King (Obama) was trying to drive major policy shifts, why hasn’t it happened?

In part, I think it’s a collective action problem. If you’re a student who’s stopped out of college, or who has finished your degree but are now facing underemployment, you might blame your high school advisor, or the college you attended, or even yourself; you’re unlikely to blame the system for, in effect, not teaching you the most important things. But that seems to be a real factor.

The other, I’m guessing bigger problem is our old friend inertia. Most public school K-12 teachers I’ve known are at once profoundly committed to the success of their students, and permanently exhausted at an almost spiritual level. They are over-scheduled, overburdened, over-regulated. Standard curricula are often less a platform than a straitjacket. And their unions are (understandably) focused on easing teachers’ day to day experience. Both in their negotiating pose vis-a-vis the school districts with whom they collectively bargain, and their political activities within Democratic campaigns, they are almost definitionally skeptical of reform.

So we have a status quo that is drifting ever further off target, if the target is “prepare all students for success in the labor market”—even as that status quo is succeeding on its own terms of high school graduation. What can flip the dynamic?

Here in New York, we’ve seen a recent development that I think should inspire hope and dread in roughly equal measure. Last year, our state Education Department moved to phase out the Regents exams as a graduation requirement, finally getting us off the ever shrinking list of states that require standardized tests to earn a diploma. That’s good.

They replaced it with something called the New York State Portrait of a Graduate, which “serves as a forward-looking framework, where academic excellence intersects with the demands of a dynamic, rapidly changing world… [and] reflects our commitment to preparing students not only for college and careers, but for life as engaged, thoughtful citizens.”

Yeah. I’m not sure how to operationalize that either.

I just know that my single smartest colleague on all these issues, who was involved in the commission that voted for the change, is terrified about it—even as she notes that some schools which have been moving in this direction for years will thrive under the new flexibility. Removing Regents exams should open up a ton of time in the school day, which could be filled with essential skills development, career exploration, critical thinking and other instructional priorities that should better position learners to thrive in the world beyond high school.

But it’s going to require energy, leadership, focus, and determined optimism. Whether we have enough of those virtues to overcome the inertia within the system very much remains to be seen.

A heroic job creator of note.

Not infrequently, shifts in who holds power drive schizophrenic shifts in policy. Since 2002, New York City has been led by Michael Bloomberg, himself the avatar of the business community, whose administration often tilted progressive right up to the point where a policy choice might interfere with someone making unlimited profit; then Bill de Blasio, whose rhetoric was more effective at making enemies than making change;1 and then Eric Adams, who (as mentioned in my last post) was never happier than when meeting with very wealthy captains of industry; and now Zohran Mamdani, who said on the campaign trail that he doesn’t believe billionaires should exist.

To be sure, there’s a lot of daylight between “the business community” and the ultra-wealthy. It’s one thing to doubt the societal value of billionaires in a time of both extreme inequality and sharply rising political spending by the absurdly rich, and another to enact policies with indifferent or hostile intent to how they will affect employers of all sizes.

Still, given that Mamdani is a Democratic Socialist, and that he was sworn into office by leftist demigod Bernie Sanders, it’s not surprising and maybe even understandable that business interests got spooked at the news that Mamdani had named a Deputy Mayor for “Economic Justice,” rather than one for “Economic Development.”

The premise underlying that concern is that economic justice—which I think we can fairly define as the struggle to elevate the agency, dignity, and security of working people—is a zero-sum struggle. If you buy that, you’re saying that for workers to win, employers must lose.

It’s a neat narrative that offers the comfort of having one’s priors confirmed. It also runs counter to a ton of research indicating that employers tend to prosper when their workers thrive. Firms that take steps to raise their workers’ pay or otherwise help with economic or life challenges benefit from lower turnover and related costs, and can see profit margins rise as a more experienced workforce with higher morale delivers better service for customers.2 Particularly for small businesses and startups, keeping workers happy—and in place for longer—might mean the difference between staying afloat and going under.

Of course, some inputs to worker satisfaction are closer to zero sum than others. Higher pay and more generous benefits have direct costs for the employer, and (simply from the employer’s perspective) only make sense if modeling suggests they’ll yield a strong return. Other worker-friendly practices, like moving toward reliable and family-friendly scheduling, might be disruptive to some extent, requiring shifts in operations.

But a third category of changes, in which employers help workers better contend with common challenges outside the workplace, are close to pure win-win solutions. This bucket includes reliable and affordable high-quality childcare, transportation access, resources to help families find healthcare (including mental health), housing assistance, financial planning, and anything else that helps facilitate access to a job.

If pieces of this sound like Mayor Mamdani’s affordability agenda, that’s not a coincidence. They’re his priorities from the campaign into his first months in office, and rightly so. And while my feelings about Governor Kathy Hochul are a bit more nuanced, I give her credit for getting on board, especially on childcare, and apparently trying to figure out how to say yes to the mayor and the city, rather than no.

What I don’t know, but wish I did, is the extent to which this will inform the work of the new Economic Justice portfolio. The coverage I’ve seen has focused on the unclear future of the Economic Development Corporation under this new mission.3 Deputy Mayor Julie Su has written and spoken extensively about job security and workers’ rights in previous roles, including as Acting Secretary of the U.S. Department of Labor during the Biden Administration, but to my knowledge hasn’t yet revealed a detailed strategy in her new job.

Were I advising them,4 I’d suggest that the strategy must include, one, an emphasis on wraparound services—those enabling conditions of work, like childcare and transportation access—and two, opening the workforce development toolbox to help build vehicles supporting career advancement and economic security. This path, too, offers the possibility to advance economic justice without punishing anyone on the employer side.5 Quite the opposite, in fact.

Everyone has a stake in boosting the skills of the current workforce. Last year, the NYC Employment and Training Coalition estimated that one million New Yorkers of prime working age are either unemployed, working part-time when they would prefer a full-time role, or working full-time for poverty or near-poverty wages. Most local job creation since the pandemic, meanwhile, has been in low-wage work—jobs that don’t pay enough to afford a decent life in the city.

At the same time, the squeeze on immigration and reduced domestic migration threaten to decrease the steady influx of talent into NYC from around the country and around the world that by now is probably taken for granted. Many of the jobs they would have filled are still likely to emerge, presenting the possibility for new pipelines to move hometown talent into jobs and career paths—and real economic mobility.

Workforce development and job training can help make this happen. But it’s only possible if government welcomes employers to the table, and if employers then commit to staying engaged until solutions emerge.

The role of the workforce field here is no less important. To simplify only a bit, it’s to stop being so basic. We need less—or at least not only—“what are your current openings?” and “will you take this candidate?”, and more what their ongoing pain points are, past adverse experiences partnering with the sector, their own internal best practices for career advancement, which incentives matter and which don’t, how much insight they have into their own turnover numbers, including costs and driving factors, et al.6

I don’t think economic justice can be ripped out of the capitalists’ greedy claws. I think maybe it can be built together, as a shared project of the private and public sectors, through a focus on imbuing every job with as much dignity, security, and potential mobility as feasible. Between the alternatives of a politically impossible all-out class warfare approach, and keeping to the miserable status quo of a workforce system that at best keeps most of the adults it serves in a state of economic precarity, this is the path the new administration should follow.

Mayor Eric Adams, in happier times

The mayor’s State of the City address that followed a few months later, in early 2023, included the promise of “30 by 30”: 30,000 apprentices in New York City by 2030. The Future of Workers Task Force report released later in 2023 identified apprenticeship as one of five key levers to move the system forward.2 This was good news, too: apprenticeship is a proven and powerful vehicle to lift workers into career-track employment at living wages, and by the nature of the commitments required of sponsoring employers, apprenticeable roles are generally good and stable ones. Add that federal policy favored apprenticeships, at least at the level of rhetoric, and has continued to do so even through the Trump-dominated chaos of the last decade.3

Had Mayor Adams won re-election, his second term would have ended at midnight on January 1, 2030. In another reality, perhaps, we might encounter an older, gentler Eric Adams in his lame duck period of November-December 2029. Making the rounds for a series of legacy-focused interviews, he is both humbled and exalted after eight years of selfless service to the public, getting just a little misty when he shares that, among all of his many accomplishments over two terms in office, he’s most proud of building strong pathways to meaningful work and economic security for tens of thousands of New Yorkers.

But in this reality… well, that won’t be happening. With Adams gone and his works in disrepute, I’m doubtful how much of a priority it will be for NYC to reach that goal of 30,000 apprentices over the next four years.

I would suggest that this, too, is a very good thing.

One problem with working for government in a policy role is that, since you know more and care much more about your issue of focus than the people at the top of the hierarchy, getting their buy-in often will require promising some result that resonates with whatever they do care about. Chasing that result is very likely to have a distorting effect on the actual work, even as it hopefully adds resources—funding and political heft—to your efforts.

This is all especially true if the targeted outcome is a Big Round Number (and the targeted outcome is almost always a Big Round Number). When this is so, all the incentives align toward reaching the number, and the thing you’ve agreed to—or, more likely, proposed and won approval for—in order to advance the work, threatens to distort or even undermine the work.

What bothered me about “30 by 30” was that to get one apprentice at, say, age 18 to 20, you probably need a dozen young people who are potentially on track for that possibility at around age 14, and maybe three or four who are actively considering it as an option, or straight-up preparing for it, at age 16 to 18. That means career awareness and career exploration in the middle grades, and more specialized and challenging work-based learning experiences in high school, and solid advising to help put these experiences into context and translate them into a plan. Ideally you’d also have strong systems to track them along the way and figure out who remains in your candidate pool.

Apprenticeship is the tip of the pyramid, the icing on the cake. You don’t start a cake with the icing, and it was far from clear to me that the necessary preceding steps were in place—certainly not at the scale needed to reach the Big Round Number.

Without the supporting layers, the structure won’t stand. What tends to happen then is that the target gets defined downward: some elements of apprenticeship might fall to the side, or small print is added to make the goal “apprenticeship-plus.” Pretty soon, the number is the work, and what that number is supposed to represent—a well-lit and well-supported pathway to a better life for a hardworking person who needs it—has fallen by the wayside.

My hope is that without the burden of the Big Round Number, the work can advance in a more organic way. I’m not so naive as to believe that we can go without BRNs, in workforce or any other issue (hello, housing; good afternoon, job creation). But the responsible way to do it is to understand current capacity, determine what’s additionally needed in terms of policies and resources to ensure that you don’t just hit the number once, but can sustain it indefinitely, and at a level of quality where you’re actually delivering the value implied by the promise.

I’ll talk about three things here. First, what I think AI might mean for entry-level work, which is probably my natural starting point as someone who’s spent most of the last 15 years thinking about how youth and young adults prepare for, enter, and hopefully thrive in the labor market. Second, the potential reach of mid-career dislocation, which probably matters most with respect to politics. Finally, what the best-case scenario—in which we think about AI like the enlightened adults we want ourselves to be, and make laws and allocations accordingly—might look like.

Last thing before we jump in: I’m focusing here on how AI is likely to impact professional, white-collar jobs that involve more varied tasks—as distinct from automation, which is using machines to perform routine and repetitive tasks. The latter—think self-checkout at the supermarket or pharmacy, or ordering on a kiosk at McDonald’s—is deeply concerning for a whole bunch of other reasons, starting with how it will intensify competition for a shrinking pool of lousy jobs. But our political class has yet to show much interest in this.

I. IT’S A TRAP!

Admiral Ackbar in his moment of epiphany, looking surprisingly calm. Maybe it’s the technical limitations of early 1980s filmmaking. Or maybe he has no student loan debt.

I was on a webinar yesterday for Lightcast’s new report, “Fault Lines.” This is a semi-sequel to “The Rising Storm,” which I wrote about previously. It takes the essential point of that report, that a decrease in labor supply driven by demographics and other factors will lead to chronic shortages in key sectors, and adds in the additional challenges of rising geopolitical tensions and the emergence of AI.

Lightcast asserts that AI is adding capacity in professional and white-collar roles where we already have enough people to do the work. Of course, these tend to be the jobs that we’ve held up as the most desirable: cool tech roles and finance and professional services positions that confer both economic security and social esteem. The sort of thing you’d say to impress a date, or a home loan officer. Hence the trap: the roles we’ve told young people to work towards, and implicitly held out as a reward for being good at school and not raising too much hell along the way, might now go the way of the toll booth attendant.

On the call, one of the presenters noted that we currently have a record number of computer science majors in college while hiring for the field has dropped to a longtime low. This seemed like one of those just-so facts that might be technically true but misleading or incomplete, so I checked it out. And sure enough…

While CS graduates still command the highest starting salaries at $80,000, they rank 7th highest in unemployment among all college majors according to Federal Reserve Bank of New York data. Computer engineering majors fare even worse at 7.5% unemployment.

Compare this to fields tech leaders spent years mocking: philosophy majors have 3.2% unemployment, art history graduates just 3%, and even journalism majors do better at 4.4%. The “learn to code” promise has become a cruel joke. […]

On one side, you have the number of computer science graduates climbing every single year, hitting record highs in 2025. Universities kept expanding CS programs because, for a long time, software engineering felt like the safest career imaginable. […]

Even if tech hiring improves slightly in the future, the pipeline of new CS grads is already locked in. Universities can’t turn this faucet off overnight. So you end up with thousands of capable graduates competing for a shrinking job pool.

Emphases in the original.

Is this happening already? Last fall, the Stanford Digital Economy Lab released the so-called “Canaries Paper,” which found that “Early-career workers (ages 22-25) in AI-exposed occupations experienced 16% relative employment declines, controlling for firm-level shocks, while employment for experienced workers remained stable.”

I won’t steal nerd-valor or insult your intelligence by pretending to have read the whole study, much less the various counters to its findings. But note that:

1. College graduates in 2025 submitted more job applications than their predecessors a year earlier, but received fewer offers.1

2. Only 30 percent of 2025 grads reported finding a full-time job related to their field of study, down from 41 percent for the Class of 2024.2

3. Significantly fewer employers reported “difficulties in hiring entry-level talent” in 2025 (35 percent) than was the case in 2022 (65 percent).3

On the same Lightcast webinar, one of the authors noted that when he graduated from college in 1990, the labor market was unusually forbidding for entry-level workers. I graduated five years later, and was fortunate to catch the economic updraft. But my story actually helps illustrate a larger point: while these jobs might not go away, those with privilege and access will be even more likely to get them.

My first job out of college was with NBC Sports. I went to their offices in NYC one day in the late spring of 1995 to meet a cousin of my mom’s, for whom she’d babysat when they were kids but hadn’t seen in decades, and who was now a senior vice president there. He said his division wasn’t hiring, but recommended I look into applying for the NBC Page Program. (This would have been a hilariously awful fit.)

He suddenly had to run off to take a call, leaving me in the middle of the hallway. I happened to look into the office nearest where I was standing and saw a man sitting at a desk under a poster of Sen. Robert F. Kennedy, with whom I’d been somewhat obsessed as a college student. He looked up and asked if he could help me. I said I liked his poster, and it turned out that he was from suburban Philadelphia as well as being an RFK admirer and, naturally, a big sports fan.

My mom’s cousin came back from his call and said, “Kevin, I see you’ve met David. He’s a writer.” Kevin said, “I need to hire a writer.” I worked there for the next two and a half years.

I’ve told this story a lot over my career in workforce, as an example of the importance of networks as well as how you can’t always anticipate opportunity, but you can always be ready to meet it. I hope it hasn’t come off as too self-congratulatory: my talking to a stranger at that age was sufficiently uncharacteristic that it’s easy for me to ascribe it to fate or what have you. But the truth is that privilege—a family connection (and a “weak tie” at that)—put me outside Kevin’s office that day.

That said, fast forward 31 years and it’s a fair question whether he’d need to hire a writer at all. What I actually wound up doing in that job was produce online content for the very early days of the internet: Web 1.0 stories and features related to NBC Sports’ on-air properties and events. I would be shocked if AI couldn’t handle whatever the modern analogue might be to my Assistant Producer (later, Associate Producer!) job, even if it wouldn’t get as much of a kick as I did out of going to two Super Bowls, two World Series, two NBA Finals, an Olympics (not as much fun) and a Wimbledon.

A few years ago, I got to pay this forward. In 2022, during my last year working for the City, I hired a new team member straight out of college. She was rough and raw in all the ways that 22 year olds are, with personal drama and at least one conflict with a partner. But she was also tireless and creative, and it was a blast to work with her. She still uses me as a reference. Her role was “project manager,” and maybe they’d hire for that even now—but half of what she did for me was research, and I suspect that now I’d be told to use AI for that part and find another solve for project management.

Which, from an enterprise perspective, is probably the right answer. As a taxpayer, I don’t want government to hire needlessly. But I’m also not sure how we quantify what’s lost when a young adult with a degree is deprived of an opportunity to build skills, experience, and a professional network. I’m pretty confident that it’s meaningful, though, and that at the scale suggested by the “Canaries Paper,” it rises to the level of a long-term societal challenge.

II. WAITING FOR WHITE-COLLAR SPRINGSTEEN

I use AI for little pieces of work. Last week, I fed Claude notes from five focus groups that were sent to me and asked for a summary of crosscutting themes. What came back, in well under a minute, would have taken me perhaps 45, and was easily as good as what I would have produced. It was the first real moment that I felt a spike of dread about losing my livelihood.4

Now, I’m not (yet?) super worried about this, because I don’t see how code can synthesize the other big thing I offer in addition to communications, which is operational content knowledge both active and latent. It’s one thing to summarize, or even synthesize and analyze, (much) faster than I can. It’s another to know who to talk to for the raw materials or the analysis, or to make non-linear connections that might broaden the scope or offer a new perspective. So long as certain information exists only in people’s heads—and our heads fill with new information every day—it seems reasonable to assume there will be some premium for those who can effectively and peaceably extract it. For now, the AI works for me. I think.

But even if I’m safe (ish), a lot of people who do broadly similar things might not be. There were about 70 million professional workers in America as of 2023. For the sake of argument, let’s take semi-seriously the predictions of people like Microsoft AI CEO Mustafa Suleyman and techno-political pundit Andrew Yang,5 who are predicting the almost instantaneous wipeout of white-collar jobs at the hand of AI. In doing so, we might say that 10 percent of those jobs will disappear, by 2028.

This would be, to make an understatement, cataclysmically disruptive. Seven million professionals unemployed, probably at the hands of big companies like Amazon, which—in any human consideration—does not really need additional profit.6 This is where, as The Atlantic recently put it, “America Isn’t Ready for What AI Will Do To Jobs.”7 Or, if you prefer, there’s Nate Silver:

So here’s a take I consider relatively straightforward, but I don’t think has really sunk into the conventional wisdom. If AI has even a fraction of the impact that many people in Silicon Valley now expect on the fabric of work and daily life, it’s going to have profound and unpredictable political impacts.

Last June, Sam Altman, the CEO of OpenAI, published a blog post entitled “The Gentle Singularity”. If you’re not familiar with the jargon, the Singularity (sometimes capitalized, sometimes not) is a hypothesized extremely rapid takeoff in technological progress — so technologies that would once have taken years or decades to come to fruition might be realized in months, days, hours, minutes, microseconds. I’m not sure that I want to weigh in right now on my “priors” about the Singularity. […]

What I’m more confident in asserting is that the notion of a gentle singularity is bullshit.

Emphasis in the original.

Now, the point that the Atlantic article makes is that the whole question here is the pace at which the technology will eat the jobs. If it’s slow enough, then we’re talking about something comparable to past advances: as the Atlantic notes, if jobs change over at 3 percent a year, you probably won’t notice and the economy likely can absorb the change. But if it’s much more than that… well, here’s Silver again:

However cynical one is about the failings of the “expert” class, these are people who tend to shape public opinion and devote a lot of time and energy to politics. If a consensus develops among this cohort that their livelihoods are threatened, or that their children’s livelihoods are, I expect there will be enough political blowback that anti-elite pushback won’t be enough to overcome it.

And here’s something I just did on Claude:

III. RIGHT!… WE WERE TALKING ABOUT WORKFORCE DEVELOPMENT.

At the end of the last “Spotlight” post, I made a reference to the Butlerian Jihad, a fictional event in the universe of the science fiction classic Dune. The idea of the Butlerian Jihad is that the mass of humanity rebels against machines that act in the place of humans: when it ultimately wins, the guiding credo is “Thou shalt not make a machine in the likeness of a human mind.” By calling it a “jihad,” author Frank Herbert nods at the essentially religious nature of the movement.8

I do think that revulsion towards AI could turbocharge the general (and increasingly cross-partisan) sentiment of populism now coursing through the electorate. There are connections to be made between Epstein-class impunity, self-enrichment at the public expense, and AI implementation that kills jobs while further redistributing income upward. It’s a matter of harnessing the bad vibes that are now pervasive yet ambient: a great leader might channel this into something useful, while a demagogue will simply ride the tiger.

The useful outcome is where workforce policies might come to bear. I’m always wary of indulging my own priors when it feels like wish fulfillment, and this scenario is that in spades. I used to describe myself as a technocratic moralist, someone who puts policy expertise in the service of core values like equity. The core value here would be something like “AI should augment human capacity and enable human thriving.”9

As you’ll see if you’ve checked all the links in this lengthy post, new stuff is coming out on AI literally every single day. So it was that a day or two ago, I saw this piece from the Brookings Institution on “pro-worker AI.”10 It offers a helpful framework, but at least as interesting to me—albeit kind of obvious—is its explanation of why investments and policies in this space are unfolding as they are:

Pro-worker tools are arguably pro-firm tools as well because they enable workers to produce more value. Yet anecdotal evidence suggests that a great deal of the current AI focus is on task automation and the development of high-level capabilities in line with artificial general intelligence (AGI), with less energy and investment flowing toward the development of pro-worker AI. Why isn’t there more investment in developing pro-worker AI? One reason is incentives: Leading firms perceive greater economic return to building and deploying technologies that automate expertise than those that create new tasks or workers and increase the value of skills and expertise.

Emphasis mine. The logical conclusion of this “incentives” approach can be seen in a report Bernie Sanders issued last fall, which basically calls to box AI in by passing broadly pro-worker policies: making it easier for workers to organize and strike, implementing a work-based safety net (32 hour standard workweek, guaranteed paid leave, higher minimum wage), and rewriting the tax code to stop rewarding investments in automation and AI over human capital.

I’d love to see all of this come to pass. But if you believe, as I do, that the oligarchs essentially wield a veto by controlling maybe 35-40 percent of the electorate through their bought media, it’s tough to see how Bernie’s pro-worker agenda prevails before or without getting big money out of politics, which itself is very difficult to imagine right now.

So for better and for worse, I think we’re stuck with AI, which presents the opportunity for workforce development. From traditional workforce functions like resume development, interview prep and job search, to education and training, to the broader emphasis on job quality that I think is the future of the field, AI can be an enormous force multiplier for well trained workforce professionals. At a practice level, my friends at NYC’s Workforce Professionals Training Institute are thinking hard about this, creating facts on the ground that deserve policymakers’ attention.

But this will only happen if we get better at telling our stories and selling our value. I used to believe that employer engagement was the single greatest challenge for the field: our bad and seemingly unbreakable habit of approaching businesses through a purely transactional lens, starting with a demand to help with our immediate need rather than a question of how we can support them over the long haul. That’s still real, and related. Yet it too is bound up in our deficits of narrative.

Where AI meets workforce, or should, is that both hold the potential to help build a more just and equitable society—or to accelerate our drive toward techno-feudalism. This emerging technology offers us a new opportunity to make the case for the value of what we offer as a field. We need to use it.

I’d like to think I’m better positioned than most to understand how one’s imagination can falter in the face of actual American fascism. Last year I completed a novel that I’d been working on since 2022, which started as a way to process my experiences working in New York City government and wound up as an outlet for my anxieties about where I saw the country going. The story is mostly set in 2037 and 2038, but two key segments flash back to 2028-29, as a contested presidential election leads to a period of chaos and eventually the army imposing martial law. On the other side of that, America more or less divides itself into two models of governance, sharing a common currency and not much else.

(Most of the story takes place in a blue city within a red state, in which gambling is the only viable industry left. The main plot concerns researchers and political figures battling over a referendum on whether or not to allow “gaming” on the life and day to day outcomes of young kids. There’s also a possible young adult suicide epidemic, a Free Range Fighting League, a new model of indentured servitude, and eventually something like full-on feudalism. It is, of course, a comedy.)

I had intended to self-publish this story once finished, but I held off last summer because there are bits that harshly mock social service philanthropy and philanthropists, and at the time I had a job that involved trying to raise money for an intermediary. But meanwhile the world moved in ways that actually overtook some of the themes and dynamics that I’d been speculating might happen maybe, someday, back I was writing this in 2022 and 2023 and 2024. AI got ridiculously powerful, to the point that the way my researcher characters use it in the story would seem simple to a high school student. Gambling continued to spread, to the point where it got abstract and collateralized, and it even began to show up in sports in ways that would have horrified an American of the 1990s (which is basically what I am) but that I’m guessing most contemporary people will shrug. And of course the Trumpian abuses have been more blatant and vicious, if perhaps not yet as pervasive, as those my characters are reacting to in 2028. Currently I’m trying to go through the draft to figure out if it can be revised into something I can self-publish, but it honestly might be like trying to play the cello in a marching band.

I did make one prediction in that novel that will take a little while to bear out, but in which I remain overwhelmingly confident: Donald Trump will run for president again in 2028.

That so many people in public life seem to dismiss this possibility so readily—this piece on Friday, which I swear made me dumber as I read it, was what prompted this whole outburst—feels to me like another failure of imagination, a failure to face the sheer velocity at which things are getting worse. Paid political opinion artists are perfectly able and usually willing to call out unprecedented things, this abuse or that corrupt act or those gleeful transgressions of norms, yet seem unable or unwilling to follow this logic to its obvious conclusion: Trump will not be constrained by any ordinary forces.

It comes down to two very simple questions. First, what in Trump’s history or character, what from all we know of this person, suggests that he will honor the Constitution and not run again? Why might he—not “one,” not “a normal politician,” but he, Trump—chose to depart a role that offers greater access than anything else in the world to his two addictions: attention, and grift?

And second, what signal have Republicans really given that they will act, ever, as individuals or as an organized party, to say no to this person?

So yes: of course he will want to run again. My fictional prediction is that in late winter 2028, the Supreme Court rules 7-2 that the 22nd Amendment does not apply because there were legitimate questions about the validity of Trump’s previous elections, so it’s not clear whether or not he was ever elected, let alone twice; as such, it would violate his due process rights to bar him from running again. Or maybe they never decide it, but issue a stay to make Trump presumed eligible.

The point is that they will find a way to “legalize” his campaign. Authoritarian regimes always have a paper trail of their encroachments.

The Republicans will embrace this, whether out of terror—because, yes, he has something pretty close to a personal army now, not to mention countless additional heavily armed aggrieved randos—or true belief, or just the realization that the most certain way they could win is if the administration steals it. (See also this story, which—talk about being overtaken by events—was published between my first draft of this post and when I’m hitting “send” on it.)

The only scenarios I see in which Trump wouldn’t run for president again involve either his dying in office, or some enormous conflict coming even sooner that leads either to the administration canceling future elections, or to his forcible removal. None of that would be much fun to live through.

In my novel, the outcome of the 2028 election is that the administration applies enough pressure—arrests of Democratic organizers, poll restrictions, intimidation—to win most of the close states, but then loses a bunch of western states and Alaska, which tips the Electoral College to the Democrat (who never really appears, but I have in a notes file as something like “imagine Gretchen Whitmer, but the fictional version played by Melanie Lynskey at her most badass”) by a final margin of 271-268. He refuses to accept the results and authority collapses in a handful of states and unrest spreads for months until the military declares martial law.

What will actually happen ? No idea. But if he’s alive and even plausibly functional, Trump will run for president in 2028. I think the sooner we treat this as a high likelihood, the better prepared we will be to figure out how to respond.

I was lukewarm on both the organization and the subject until after 9/11, when two things happened. First, the job-loss consequences of the attacks brought home to me how important this area of policy really was. Second, I got my hands on a compelling story about how decisions made by Mayor Rudy Giuliani and his appointees made that awful situation much worse: their ideological refusal to spend federal workforce funds meant that when the attack on lower Manhattan drove tens of thousands out of work, the City had no infrastructure in place to help them.

After that, I was hooked, and as workforce went from a total afterthought under Giuliani—who saw “job training” as a refuge for those too lazy to work—to drawing some attention and investment under Mayor Michael Bloomberg, I both got better as an analyst and found some audience for my output. In 2010, I moved into city government, and my focus began to shift toward youth issues. By 2015 I was the founding executive director of the NYC Center for Youth Employment, a role I held for more than seven years. I left in 2022 to consult, which I’ve done since, other than that dubious foray into nonprofit management which I referenced in my last post.

Over all that time, as a researcher or a government administrator or a consultant, I was trying to tell workforce development stories that emphasized how this area of policy provided a set of tools for both individual economic mobility and broader prosperity at a firm, sector, or community level. At the local level, it arguably “worked”: while my own contribution to this is unknowable, NYC unquestionably got far more sophisticated with respect to workforce, and significantly boosted local investment, from the early 2000s to the early 2020s.1 Nationally, though, funding continued to decline, and the direction of federal workforce policy remained frustratingly irrelevant to the major levers of the economy. (Again, see my last post.)

Then in October 2024 I read something that simultaneously rewired my previous understanding of this work I’d done for so long, and offered an explanation for why workforce development—this field I’d made my life’s work—is never at the center of the agenda despite its relevance the near-universal experience of work.

This thrilling and destabilizing report came from Lightcast, and is called The Rising Storm. I described it on LinkedIn as the policy equivalent of an early Stephen King novel: “impossible to put down and scary as hell.”

The premise of The Rising Storm is that the American labor market is in imminent danger of facing an open-ended talent crunch in key sectors of the economy like healthcare and construction. The consequences of these shortages will include higher prices, longer waits for lousier services, and generally reduced quality of life. The biggest driver is a birth rate in long-term decline, but other contributing factors include earlier retirements, lower labor force participation, and an aging population with a growing demand for care. (Sure enough, healthcare job growth in January 2026 accounted for the large bulk of overall employment gains, and basically has accounted for all net job growth over the last two years.)

The report—which, again, was written in 2024—includes a section called “Immigration is Keeping the Labor Market Afloat.” It notes that a surge of newly arrived workers was a big reason the U.S. pulled out of its post-COVID inflationary spiral, and made the case for the importance of foreign-born talent in key industries:

Healthcare in particular clearly illustrates the US labor force’s reliance on immigration. Nearly 2.8 million—or 18%—of the healthcare workers in the US have come from outside the country. Foreign-born workers account for one in every four doctors, one in five registered nurses, and one in four health aides. (In some states, these numbers approach one third, or even one half, of workers in these healthcare occupations). The US currently employs 143,000 nurses from the Philippines alone (about 5% of all nurses) and more than 350,000 Filipinos in healthcare overall.

So if the question is whether immigrants can supply the workforce that the US-born population isn’t providing, the answer seems to be yes—for now. Foreign-born workers make up such a large share of the US workforce that it could not function without them.

“For now” is now “back then.” The Brookings Institution found that in 2025, the U.S. had net negative immigration for the first time in at least half a century. Brookings also notes that “[r]educed migration will dampen growth in the labor force, consumer spending, and gross domestic product (GDP). We estimate the sustainable pace of monthly job growth to be between 20,000 and 50,000 in late 2025 and believe it could be negative in 2026.”2

The situation here in New York sharply illustrates the danger. Here’s the state’s recent and projected population change, from a newly released report by the Public Policy Institute of NY:

The first two columns reflect the relatively high level of immigration through the Biden administration, which more than accounted for our state’s population growth earlier in this century and mitigated the more recent decline. I’m not sure how pessimistic the assumptions are for the projection in the rightmost column. But either way, we’re going to get older, and immigration level is the only real variable unless you project, in the face of all logic and evidence, a second Baby Boom.

Economics is an enormously important subject that I’ve read good books about and taken courses on, and yet still live in terror of being exposed as a dimwit whenever it comes up. But I’ll embrace the risk here, to say that my understanding is that when you have less of something, you tend to value it more.

Between the long-term decline in the fertility rate, and the newer policy choice to repel would-be migrants and deport countless workers already here, that’s where we’re headed with respect to the labor supply. The logic of market capitalism calls for excess labor supply in a context of growth, to keep wages down and forgive inefficiencies in talent development.3 The logic of fascism and white supremacy seems agnostic on quantity of available talent, so long as the good jobs are reserved for the in-group.

For now, though, we’re a market economy led by a crayon-eating would-be dictator. I read a great piece the other day making the argument that although, yes, Trump is a fascist, the country is not—at least, not yet. And as The Rising Storm and other analyses from groups like the Georgetown Center on Education and the Workforce4 point out, demand for talent isn’t likely to decline.

If I’m reading all of this right, it’s almost impossible to see how we avoid some of the damage predicted by The Rising Storm. But one way to mitigate it would be an approach we might call Every Worker Matters: the idea that every individual in the workforce should be supported to be as economically productive as possible.5

What does this mean in practice? I think two things. One, it’s leaning all the way into a new mission for education that incorporates career readiness and economic agency—not to make kids widgets or exploitation-ready automatons, but to help them figure out what best prepares them for lives of purpose and security. And two, it’s helping adults maximize their talent, along with their earning power. That might be investing in those enabling conditions of work like childcare and transportation access, so they can get and keep jobs. It might mean helping them identify a certification or other credential that would qualify them for a promotion or a wage gain. For millions, it could mean helping them return to college to complete degrees toward which they earned credit before stopping out.

Workforce development is the toolbox for these solutions, which we’re going to need even if by some miracle the fascists are stopped and Lady Liberty’s welcoming torch is relit. Unfortunately, we’ve mostly just used the rapid attachment wrench and the low-cost job training screwdriver. But there are stronger instruments in the box, just waiting for someone to dust them off. I think that moment is coming soon, and I hope my field will be ready.

That’s Part 1 of why workforce development is about to matter much more. Part 2 is how AI is already messing with the labor market, and the gathering pushback. As with so much else in this timeline, I mostly hate that it’s happening, but I’m admittedly intrigued to see “Butlerian Jihad (America’s Version)”.