By the third quarter of 2024, the United States had committed more than $1.1 trillion in announced industrial investment, the largest peacetime capital reallocation toward physical industry in fifty years, channeled through subsidies, tax credits, and direct authorizations distributed by a Washington that has rediscovered industrial policy after a forty year hiatus.
The map of where the money is going reads like a political document: Arizona, Ohio, Georgia, North Carolina. Counties chosen, in many cases, for what they pay back at the ballot box. What the map does not say is who is going to operate any of it.
This piece is about that question, but not the way most commentary has framed it. The dominant framing, pro subsidy or anti subsidy, free trade or fair trade, has obscured a more fundamental problem: the United States cannot now staff the industrial economy it has spent the past four years deciding to subsidize, and the reason it cannot is the same set of policy instincts that produced the subsidies in the first place.
That is the thesis. The rest is the argument.
The Familiar Mistake
Americans tend to discuss industrial policy as if it were a recent invention. It is older than the country, and its record is consistent enough that the current enthusiasm should give a thoughtful observer pause.
Hamilton's Report on Manufactures (1791) proposed bounties, tariffs, and subsidies for nascent American industry against the open warning of agrarian republicans. It is fashionable to credit the resulting policy regime, completed by Clay's American System and McKinley's tariff schedule, with American industrial dominance. The more honest accounting credits land, immigration, capital formation, and the steam engine. The tariffs were a drag on growth, not its cause. The country grew despite them, not because of them.
The wave that followed 1945, Sematech, NASA, the GI Bill, was different in kind. It was conducted under the discipline of an existential threat and produced, alongside genuine industrial capability, the largest peacetime expansion of federal payroll and credentialing in American history. Whether the gains exceeded the costs is a question of accounting choice. What is not contested is that the infrastructure of credentialed labor produced in that period, the IBEW apprenticeship halls, the UA training trusts, the community college trades programs, was the workforce backbone the current reindustrialization assumes will still be available. It is not. It was eaten alive, between 1980 and 2010, by another set of federal priorities.
The third reindustrialization, announced by the CHIPS Act in 2022, is being defended on grounds that should be familiar to any student of the prior two. National security, supply chain resilience, geographic redistribution. Each of these is real. None of them are arguments for industrial policy in the form being administered. They are arguments for modernizing lawful immigration, deregulating energy permitting, and reforming the occupational licensing regime that has made the American labor market more sclerotic than at any time since the Wagner Act. The current approach has chosen the path of subsidy and tariff instead. The choice is the central planning instinct, dressed in 2024 vocabulary.
The fabs will be built. Whether they will be staffed is the question we keep failing to ask out loud, because the honest answer requires admitting which policies created the shortage.
What the Subsidies Actually Buy
The case for the third reindustrialization, on its own terms, deserves direct engagement. The strongest version is roughly: the pandemic exposed acute supply chain fragility, the country needed a coordinating signal to overcome market under investment in domestic capacity, and a small targeted subsidy was an acceptable price for strategic insurance. By the end of 2024, more than $400 billion of new semiconductor investment had been committed in the United States. Maricopa County, Arizona added 21,000 manufacturing jobs in a single year. These are real outcomes.
The weakness of the case is that the same outcome was achievable, on a longer timeline and at lower fiscal cost, through unilateral liberalization. A serious tariff reduction on imported manufacturing capital goods, a permitting reform that brought American industrial siting time in line with Texas, and a clear eyed program for vetted, high skill engineering talent would have produced the same investment flows without the trillion dollar fiscal hangover, the politicized recipient list, or the structural over commitment to incumbent firms that the subsidy regime locks in for a generation.
Crony capitalism is not a slogan; it is a description of what happens when the federal government picks winners. The current recipient list, Intel, TSMC, GlobalFoundries, Micron, the major battery integrators, is also the list of firms that maintain the largest Washington government affairs operations. This is not coincidence, and it is not corruption in the narrow legal sense. It is the predictable equilibrium of a policy regime that hands out $52 billion through discretionary grant programs.
The capital is here. It is here on terms that distort price signals across the entire industrial economy and that lock public commitments into private balance sheets which will, in due course, become political assets of their own. Reversal will be hard. That is by design.
The Labor Problem the Planners Made
What the subsidies could not buy, and what the country cannot manufacture in the timeframe the subsidies contemplate, is the workforce. This is the part of the conversation we have been least honest about.
The Bureau of Labor Statistics' replacement rate analysis puts the skilled trades shortfall at roughly 1.9 million roles by 2033. Industry estimates run higher. Pipefitters, electricians, controls technicians, instrumentation specialists, journeyman welders rated to ASME standards, every one of these requires four to six years of structured apprenticeship to certify, and we did not start. The pipeline has been hollowed out across three decades of policy choices that nobody connected, at the time, to the industrial capacity question they were eventually going to ask.
The collapse of American vocational education is the relevant case study. Vocational and technical high schools in the United States were cut by approximately 60% between 1980 and 2010. The proximate cause was a federal education establishment that, through Title IV financial aid and a generation of Department of Education policy, oriented every dollar of public secondary investment toward four year college preparation. The student debt that the same federal policy underwrote, $1.6 trillion outstanding, bought the country a labor force credentialed for office work, exactly when the industrial policy the same government had just authorized required credentialed welders. The two policies are administered by different cabinet departments, and neither has the institutional incentive to acknowledge the other's role in producing the outcome.
Occupational licensing is the second case study. Roughly 22% of the American workforce now requires a government issued license to perform work that did not require one fifty years ago. Cosmetology, interior design, real estate, dental hygiene, electrical work, plumbing, each governed by a state licensing board, typically captured by incumbents, with the explicit effect of restricting labor supply. The interstate compact reform that would let a journeyman electrician licensed in Tennessee perform commissioning work on the new Hyundai plant in Georgia is, in 2026, still not fully ratified. The cost is paid in months added to every project schedule.
The third case is lawful immigration, and the distinction matters: the argument is about legal, vetted, high skill entry, not about border enforcement, which is a separate question with its own legitimate answer. The short version is that the country which built its industrial dominance on the ability to recruit the world’s best engineers spent the past several years making lawful recruitment harder, and made it harder at precisely the moment the subsidy regime ramped up. The right hand and the left hand are administered by different cabinet departments. Neither has the institutional incentive to acknowledge that the other one is working against it.
Why the Fix Isn't More Planning
The instinct of the policy class, on encountering the workforce shortfall, has been to propose more programs. More apprenticeship subsidies. More workforce development grant rounds. More targeted training funds. There is a Department of Labor announcement to that effect approximately every six weeks.
This is the precise category error that produced the shortage. Federal workforce programs do not produce welders; they produce administrative overhead and regional disparities in program quality that correlate with which state has the more aggressive grant writers. The Job Corps, which has existed since 1964 and spends roughly $1.7 billion annually, graduates approximately 30,000 individuals. The per graduate cost runs above $55,000, which is approximately the price of buying every graduate a community college trade program from the private market. The federal program persists not because it works but because its constituency is well organized.
The reforms that would actually move the needle are unglamorous and politically inconvenient. Three deserve specific mention.
- End the federal preference for four year college over vocational alternatives. Equalize Pell Grant treatment between accredited trade programs and bachelor's degree programs. Disallow Title IV refunds where the underlying credential does not produce a labor market return. The federal government has spent a generation oriented around producing administrative track graduates; it now needs people who can fabricate to ANSI weld specifications, and the policy infrastructure cannot pivot.
- Reform state occupational licensing for the relevant industrial trades. Interstate recognition compacts, sponsored at the state level and modeled on reforms Texas and Arizona have already advanced, would let a tradesman licensed in one state work in another without recertifying. The state licensing boards and the trade guilds will object loudly. They are also the constituency most directly responsible for the labor shortage they will deny producing.
- Modernize lawful, high skill entry. The statutory caps for skilled and seasonal categories were set in 1990, when the question of whether the country needed industrial mechanics at scale was not on anyone’s agenda. The categories were written around seasonal labor, not industrial millwrights. A lawful, vetted, employer sponsored pathway for skilled tradespeople, with credential recognition against the major allied trade certifications, is a high leverage intervention entirely consistent with secure borders and the rule of law. It rewards exactly the kind of immigrant the country has always wanted: skilled, employed, and contributing on day one.
Each of these is a deregulation or a return of discretion to the states, rather than a new federal program. Each requires government to do less, or to do its existing job more sensibly, in pursuit of an outcome it claims to value. The coalition that built the subsidy regime is, on the merits, the wrong coalition to enact the reforms that would make the subsidy regime succeed. This is the contradiction we have not named.
What We Tell Clients
The question we get more than any other, in mandates ranging from Plant Manager to Independent Director, is some version of: how do we hire for this, when this does not yet exist?
Three answers.
First: the workforce shortage is not a temporary distortion that will mean revert. It is the structural property of a labor market that has been progressively constrained for forty years, intersected with an industrial policy that assumes a market response. Hires made today against a thin candidate pool will not get easier. The companies that pretend otherwise will pay for the pretense.
Second: the firms that will navigate this best are the ones that have stopped waiting for the federal apprenticeship program to deliver and have begun building internal training capacity. The model is not "hire to fill the role." It is "hire the operator who will train the next three." A small number of clients, generally private equity backed industrial platforms with capital allocation discretion, have begun structuring their executive compensation around demonstrated team development outcomes, not just operating performance. This is unusual. It is also the only response available to a labor market the public sector has constrained.
Third: lawful skilled immigration is now one of the largest variables in American industrial competitiveness, and it is being treated as if it were the same question as border security, which it is not. Firms that have built in house immigration counsel, structured international rotations through Canadian and British subsidiaries, and treated lawful visa sponsorship as a balance sheet asset will compound advantage against those that have not. This is unusual practice in 2026. It will be standard practice by 2030.
Capital has come home. The question we have not asked clearly enough is whether the country has the institutional confidence to admit that the labor it now needs is on the wrong side of rules it freely wrote, and whether it has the practical maturity to rewrite them without pretending that doing so means abandoning either its borders or its principles.