If You Can't Get Into an Ivy, Join the Army

The last institution that cannot get rid of you is the last one that will make you into someone.

Here is the pitch, and it is a good one.

You are eighteen. You are bright, you are not rich, and you did not get into the school whose name would have done half the work for you. You can borrow six figures against a degree in a field that is being automated while you study it, or you can sign a piece of paper.

Sign the paper and the following happens. Someone feeds you, houses you, and pays you from the first week, before you are worth anything to anyone. Someone trains you in a technical specialty, at their expense, on their equipment. Someone puts you through a background investigation and, if you pass, grants you a clearance that no ordinary employer can hand a twenty-two-year-old. Someone gives you responsibility at twenty that a graduate trainee will not touch at twenty-six. You come out at twenty-two with four years of paid operational work, references from people who watched you do the job, and no debt. If you still want the degree, they will buy it for you: four years of active duty clears the threshold for the full Post-9/11 GI Bill, which pays in-state public tuition outright and adds a housing allowance on top. [1]

The catch is real and it is not small. For four years they own your location, your schedule, your appearance, your conduct, and your right to leave. They can also send you somewhere you do not come back from, or do not come back from whole, and no bonus prices that. You cannot resign because your supervisor is a fool. You may not get the specialty you were promised, and if you fail the school they will simply put you somewhere else. It is not a four-year apprenticeship with a refund policy.

Weigh it honestly against the alternative and it holds up. For a large number of capable eighteen-year-olds, this is now the better deal.

And that is the part worth stopping on. Not that the advice is good. That it is true.

Nobody made the military better

The interesting thing about the pitch is that it has not changed. Almost none of it is new. The military was making that exact offer thirty years ago, and thirty years ago it was a fallback, the thing you did when the other doors were shut. The doors were the point. The military was what remained after them.

The offer did not improve. The alternatives collapsed around it.

They collapsed in a specific way, and it is the part that gets missed when this turns into a debate about whether college is worth it. The real question is what a young person is supposed to walk through in order to become someone who can be trusted with things, and who is still building that passage.

What formation actually requires

Think about what it takes to turn an eighteen-year-old into a competent twenty-two-year-old. Not to certify them. To make them.

It takes someone willing to carry a person who is, for a long stretch, a net cost. Someone who will pay them before they produce, correct them when they fail, and stay attached to the outcome long enough that the correction matters. Formation is expensive and it is slow, and no institution does it out of goodwill. Institutions form people when they cannot get out of the consequences of not forming them.

Foreclosure alone will not do it. A prison cannot get rid of you either, and it forms nobody, because it needs nothing from you at the end. The institution has to be stuck with you and it has to need what you would become. Then it has no choice but to build it.

That is the whole mechanism. Formation survives where exit is foreclosed.

Hirschman named the escape route sixty years ago: when leaving is cheap, an institution stops having to listen and stops having to repair. [2] He was writing about the member who walks out. The version that matters now is the one where the institution holds the exit and the person cannot.

Look at what happened to the places that used to form people, and you find the same thing every time. Not a loss of virtue. An acquisition of an exit.

The ordinary university found the simplest one. Its revenue arrives at enrollment, before it has formed anything, and it bears no cost whatsoever if the person it graduates turns out useless. Its reputation is set by who it admits, not by who it produces. It selects, and it vouches, and it has quietly stopped doing the thing in the middle, because nothing in its position requires it to. A university that graduated nobody employable for four straight years would notice a decline in applications. A university that graduated nobody employable for one year would notice nothing at all.

You can watch the consequence in the numbers. In the first quarter of 2026, about forty-one percent of recent American graduates were working in jobs that do not require the degree they bought, and the unemployment gap between young degree-holders and young people the same age without a degree had closed to a point and a half. [3] None of that reaches the university as a signal it has to answer. The money was collected at the start.

The employer found a different exit and found it recently. Formation was never the employer’s goal, only a byproduct of needing the work done. The junior rung existed because the boring, careful, half-skilled tasks had to be performed by somebody, and the somebody got better while performing them. Being a side effect of production is exactly what made it reliable and exactly what made it disposable. The moment the tasks could be done without a person, the rung went and the training went with it. It was scaffolding that happened to hold people up, and it came down the moment it stopped holding anything else. This is the ladder that is gone, described from the demand side. Somebody has to want the junior work done in order for the junior to become senior. Nobody does anymore.

And then there is the elite university, which looks like the exception and is in fact the purest case in the set.

It seems to form people. The seminars are small, the standards are enforced, the graduates are formidable. But run the mechanism and ask where its exit is, and the answer is that its exit is the admissions office. An institution that selects hard enough does not have to form anybody. It takes in the already-formed: the products of decades of enforced friction that somebody else paid for, in households and schools and summers designed to produce exactly the person the admissions committee is looking for. It receives them finished. It confers the name. It sends them onward, and the outcomes are excellent, and almost none of the earnings difference was made on campus.

That last claim is checkable, and it has been checked. Compare students by the schools that accepted them rather than the schools they attended, and the earnings premium for going to the more selective one largely disappears: the ones who turned it down and went elsewhere ended up earning about the same. [4] There is one exception, and it is the finding underneath the finding. Students from poor families do gain. They are the ones who arrive without the decades of friction already behind them. The elite university adds something to the students it admits fewest of, and almost nothing to the ones who fill its class.

The ordinary university escaped formation by not caring about outcomes. The elite university escaped it by controlling the input so tightly that outcomes take care of themselves. It charges the most and it forms the least, and nobody notices, because its graduates really are exceptional and were before they arrived. So the rule does not have an exception. It has a loophole, and the rich have it.

The last complete institution

Which leaves the one that cannot get out.

Yes, the military can discharge you. It can wash you out of a school, reclassify you, deny you advancement, separate you. But look at what discharging you accomplishes, which is nothing. It does not produce the twenty-two-year-old the service needs, and there is no lateral market for a competent young sergeant. Nobody else is making them. The service cannot buy one, cannot poach one, cannot contract one in.

And underneath that, the reason no such market can exist: the person may be ordered to kill, and may be ordered to die. A contractor can be paid to accept the risk. No contractor can be placed under the duty to obey, to remain, and to bear it without limit, and that duty is the thing being bought. It has to be someone the institution made, and made from the start, because what is being built is not a skill set. It is a person who stays in the seat when leaving is the only rational move. Everything decent about the military and everything frightening about it comes out of that one fact.

So it pays you before you are useful, and corrects you when you fail, because it will be living with the result either way.

It is also the only place left that still does all four things under one roof. It selects you, with a test and a physical and a background check. It forms you, over years, at its own cost. It vouches for you, in a way that is legible to employers and to the state, and the clearance is the purest instrument of that vouching. The government’s own adjudication standard calls it the whole-person concept: an examination over a sufficient period, weighing the variables of a person’s life, all available information, favorable and unfavorable, ending in an affirmative determination. [5] Compare that to a transcript. A degree is a standardized record that you completed a curriculum. A clearance is an individual judgment that you can be trusted. And then the service places you, into work that exists, at a rank that means something.

Select, form, vouch, place. That used to be one continuous passage, and it ran through half a dozen institutions. Now the four steps have come apart. The university selects and vouches and skips the middle. The employer places and no longer forms. The credential market vouches without knowing anything.

Other institutions still form people, and the mechanism tells you which. The medical residency forms, and it cannot buy a lateral resident either. The union apprenticeship forms, and it binds both parties by design: the shop is stuck with you and needs the journeyman it is making. The religious order forms, and it takes you in and keeps you. But the certificate program forms nobody, and neither does the bootcamp. They charge tuition, they vouch on completion, they carry no one through the years when a person costs more than they return. They are the ordinary university on a shorter clock. The line does not run between college and the trades. It runs between institutions that are stuck with what you become and institutions that are paid before anyone finds out.

And every survivor on that list is a door you must already be qualified to reach. The residency wants a medical degree. The apprenticeship wants a place in a queue and somebody to vouch for you. The order wants a vocation.

The military screens too, and harder than the pitch admits. On the Pentagon’s own count, only about twenty-three percent of Americans aged seventeen to twenty-four can enlist without a waiver: weight, drugs, a record, a diagnosis, a test score take out the rest. [6] The services move that line when they need bodies, but most of the cohort does not clear it.

Look at what it does not screen for, though. Not what your parents did. Not what school you went to. Not who will vouch for you, and not what you can pay. This is the last complete institution in the country whose entrance exam is not, in some laundered form, the class you were born into. That is a narrower claim than the pitch makes, and it survives contact: not open to everyone, but the only one not closed by inheritance.

What it costs to be kept

The institution that cannot leave you is the institution that owns you: one fact, seen from two sides. It invests in you because it is stuck with you, and it controls you for the same reason. The care and the coercion come out of the same root. That is why every institution that still forms people looks, from the outside, faintly monstrous: the residency that works you past the point of safety, the order that governs your hours and your speech, the apprenticeship with its hazing and its closed shop. The unbearable parts are not a flaw in the arrangement. They are the arrangement.

And it forms you for itself, which is the concession the pitch never makes. The institution builds the person it needs, and you overlap with that person only as far as its needs happen to run. It will teach you to be reliable under pressure and it will not much care whether you learn to refuse. It can produce obedience and call it judgment. Some of what it makes of you does not travel, and some of what it makes of you should not. The formation is real. It was never designed with you in mind.

Which brings up the study that damages this argument.

The Vietnam draft lottery handed economists something rare: service assigned at random, no selection to correct for. A decade after the war, white veterans were earning roughly fifteen percent less than comparable men who never served. [7] Not a wash. A penalty.

Read one line further and the pattern turns familiar. Nonwhite veterans showed no lasting earnings loss at all. Set that beside the college finding and the two studies point the same way. The elite university adds almost nothing to the affluent and something real to the poor. Service cost the men who had civilian ladders waiting and cost nothing to the men who did not. Neither institution is a machine for raising your wage. Both seem to matter most to the people who arrive without an alternative already prepared.

Take the damage seriously, then notice what it damages. Formation is not an earnings premium and never was. The institution that forms you is not optimizing your income; it is producing what it needs. Four years is four years, and the man who spent them elsewhere was compounding something. What you get back is that you were made into somebody, by someone with a reason to bother, in a world that has largely stopped bothering. Whether that trade is worth it is the question, and anyone who tells you the arithmetic is obvious is selling something.

You cannot have the one without the other. Anyone offering you the formation without the possession is offering you a university.

A selection or a term

So there are two doors left into a formed adulthood, and only one of them actually forms you.

The first is bought, and what you are buying is the selection. The name, the sorting, the confirmation that you were already the sort of person who ends up fine. It is narrow by design, because the narrowness is the product, and it is mostly opened before you are old enough to know it exists.

The second is served. You pay in years and in autonomy, the institution takes possession of you for the duration, and at the end you are handed back with the skills and the clearance and the bearing of someone who has been through something.

Buy a selection, or serve a term.

What is gone is the thing that used to sit between them. The ordinary state university, the graduate scheme at the regional firm, the apprenticeship anybody could walk into, the long middle where an unexceptional young person of decent capability went in one end and came out the other as somebody. That road is now a toll booth without a road behind it, charging the price of a passage it no longer performs.

This is why the joke works, which is also why it is not really a joke. Telling a bright, broke eighteen-year-old to enlist sounds like a provocation, or like something a recruiter says. It is a description of what is left.

The part that should worry you

Here is the objection, and it is probably right.

The military’s junior rungs are made of exactly the material that is being eaten. Signals processing. Intelligence support. Network administration. Log review, first-line triage, the patient careful entry-level work that a young analyst has always done badly at first and well by the third year. That is not adjacent to what the models are taking. That is the center of it.

Some of it survives longer than the rest. The clearance is a judgment about a person and not a task, and no model can be adjudicated. The leadership pipeline is a genuine bottleneck, because someone still has to command. But the mass of junior technical work underneath both is being hollowed out on the same schedule as everybody else’s, and the service that finds it can run its intelligence shop with a third of the analysts will run it with a third of the analysts. It will just take longer to notice, because it does not answer to a quarterly number.

And there is a trap inside the part that looks most protective. The state will keep a human in the seat longer than a company would, because a model cannot hold a clearance, cannot be court-martialed, cannot be the name on the finding when something goes wrong. But that requirement preserves the position, not the passage. A person retained so that failure has a face is not a formed adult, and accountability can be satisfied by a signature. The billet survives. The years of doing the work badly and then less badly, which were the only reason the billet ever formed anybody, do not. That is the quiet way this door closes. The posting remains, and the thing that used to happen inside it is gone.

So the door narrows before it closes, and more slowly than the others, because the military’s need for the twenty-two-year-old is structural rather than economic, and structural needs are stubborn. But it is the same force, and the honest version of the advice has a clock on it.

Which is not a reason to tell the eighteen-year-old no.

It is open now, one of two that are, and the other mostly required being born through it. If you are eighteen and capable and unrich and the letter did not come, the thing in front of you is an institution that will take you as you are, pay you while you are worthless, teach you something real, put its name behind you, and hand you back at twenty-two with a start the university would have sold you at a loss.

Take it while it is there. Go in with the specialty in writing and the years counted and no illusions about who owns your calendar for the duration, because they will own it, and that is not the price of the formation. It is the mechanism of it.

The last institution that cannot get rid of you is the last one that will make you into someone.

Notes

[1] U.S. Department of Veterans Affairs, “Post-9/11 GI Bill (Chapter 33) rates.” Eligibility for 100 percent of the full benefit requires at least 1,095 days (36 months) of active duty; a standard four-year enlistment clears it. The full benefit covers net tuition and mandatory fees at a public in-state institution, with a monthly housing allowance and a books-and-supplies stipend on top. https://www.va.gov/education/benefit-rates/post-9-11-gi-bill-rates/

[2] Albert O. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States (Harvard University Press, 1970). Hirschman’s subject is the member or customer who can leave, and what their departure does to an organization that would otherwise have had to hear them. The application here inverts the parties: the institution holds the exit, and the person cannot.

[3] Federal Reserve Bank of New York, “The Labor Market for Recent College Graduates,” data through Q1 2026. Underemployment among recent graduates (those working in jobs that do not typically require a bachelor’s degree) stood at 41.5 percent; unemployment for graduates aged 22 to 27 was about 5.7 percent, against 7.2 percent for those in the same age group without a four-year degree, down from 7.7 percent the previous quarter. https://www.newyorkfed.org/research/college-labor-market

[4] Stacy Berg Dale and Alan B. Krueger, “Estimating the Payoff to Attending a More Selective College: An Application of Selection on Observables and Unobservables,” Quarterly Journal of Economics 117, no. 4 (November 2002): 1491-1527. Matching students by the selectivity of the schools that admitted them, rather than the school they attended, removes the effect of the unobserved characteristics on which elite colleges select. The earnings premium for attending the more selective school largely disappears under that correction. The authors find one significant exception: students from low-income families did earn more for having attended the selective school. The follow-up study, using Social Security administrative earnings data for the 1989 entering cohort, reaches the same result (Dale and Krueger, “Estimating the Return to College Selectivity over the Career Using Administrative Earnings Data,” Journal of Human Resources 49, no. 2 (2014): 323-358). https://www.nber.org/papers/w7322

[5] Security Executive Agent Directive 4, “National Security Adjudicative Guidelines,” signed by the Director of National Intelligence on December 10, 2016, effective June 8, 2017. SEAD 4 establishes the single common adjudicative criteria across the executive branch and sets out the whole-person concept, under which an adjudicator weighs all available information about a person, favorable and unfavorable, past and present, to reach an affirmative determination of eligibility. https://www.dni.gov/files/NCSC/documents/Regulations/SEAD-4-Adjudicative-Guidelines-U.pdf

[6] U.S. Department of Defense, 2020 Qualified Military Available Study (Office of the Under Secretary of Defense for Personnel and Readiness). Approximately 77 percent of Americans aged 17 to 24 would not qualify for military service without a waiver, up from 71 percent in the 2017 study. The most common single disqualifiers are being overweight (11 percent), drug and alcohol abuse (8 percent), and medical or physical health (7 percent); 44 percent of disqualified youth are disqualified for more than one reason. Waivers are widely used and the standards move: roughly 17 percent of recruits received a medical waiver in 2022, up from 12 percent in 2013.

[7] Joshua D. Angrist, “Lifetime Earnings and the Vietnam Era Draft Lottery: Evidence from Social Security Administrative Records,” American Economic Review 80, no. 3 (June 1990): 313-336. The randomly assigned induction risk removes the selection bias that contaminates ordinary comparisons of veterans and nonveterans. In the early 1980s, white veterans earned approximately 15 percent less than comparable nonveterans; Angrist finds no evidence of a lasting earnings loss for nonwhite veterans. The estimates cover conscripts in a wartime, draft-era force and do not transfer directly to a modern volunteer enlistment in a technical specialty. They are cited here as the strongest available causal evidence against the argument, not as a description of it. https://economics.mit.edu/sites/default/files/publications/Angrist%201990%20-%20Lifetime%20Earnings%20and%20the%20Vietname%20.pdf