Macro Featured

Where Did the Entry-Level Job Go?

Recent-grad unemployment is 5.7% and underemployment just hit its worst level since 2020. Everyone blames AI. The Fed's own data says the real culprit is something else — and which story is right decides whether rate cuts can fix it.

Alpha One

20 Aug 2026 • 6 min read

The class of 2026 did everything right. They picked practical majors, stacked internships, learned the AI tools everyone told them to learn. Then they graduated into the worst entry-level job market in a generation — and walked straight into an argument about whose fault it is.

The numbers first. Unemployment among recent college graduates aged 22–27 hit 5.7% as of June — well above the national rate of 4.1%. That inversion matters: for most of modern history, a bachelor's degree bought you a lower unemployment rate than the rest of the workforce, not a higher one. Underemployment — degree-holders working jobs that don't require one — reached 42% this spring, the highest since the pandemic year of 2020. On the platforms where entry-level hiring actually happens, listings on Handshake fell more than 16% year over year while applications per posting jumped 26%.

Meanwhile the broader labor market has stopped generating jobs at all. July payrolls printed at negative 23,000 against expectations of +85,000. And yet layoffs remain historically low. Nobody is being fired. Nobody is being hired. The entire adjustment is landing on the people trying to get in the door.

The popular explanation is two letters long: AI. The truth is messier, more interesting, and — for anyone allocating capital around this — more useful.

The Case for the Machines

The strongest evidence that AI is eating the entry-level job comes from Stanford's Digital Economy Lab, where Erik Brynjolfsson and colleagues have been running what they call the "Canaries in the Coal Mine" project — a partnership with ADP Research that tracks payroll data covering 4.6 million workers across more than 730 occupations.

Their findings are hard to dismiss. Workers aged 22–25 in the most AI-exposed occupations have seen relative employment declines of roughly 16% — while employment for experienced workers in the same occupations at the same firms stayed stable. Young software developers are the starkest case: employment for 22-to-25-year-old developers has fallen around 20% since late 2022, almost exactly when generative AI arrived. As of the lab's February 2026 dashboard update, employment for young workers in the most exposed jobs is shrinking at 3.8% a year — and accelerating.

The mechanism isn't mass firings. It's quieter than that: the openings that used to exist for juniors simply stopped being posted. Companies aren't replacing young analysts with chatbots — they're declining to hire the next class of them.

"AI is not the whole story, but it's part of the story," Brynjolfsson says, "and the evidence is building."

Briefings like this land in members' inboxes before the market prices them in. Join free →

The Case Against the Machines

Here's the problem with the AI story: the timeline doesn't fit.

In June, researchers at the Federal Reserve Bank of New York published a study that should have gotten far more attention than it did. Their finding: roughly 64% of the rise in unemployment among college-educated workers under 29 — versus pre-pandemic levels — is attributable not to AI, but to the expansion of remote work.

The logic is uncomfortable but intuitive. In occupations that can be done remotely, unemployment for young graduates rose about a full percentage point between 2017–2019 and 2022–2024 — while for workers 29 and older in those same occupations, it actually declined. Employers, the researchers argue, are reluctant to hire someone with zero experience into a job where nobody will sit next to them, correct their mistakes, or mentor them between meetings. One Fortune 500 firm told researchers it simply hired fewer inexperienced workers during the remote-work era because training them from afar didn't work.

Crucially, the deterioration in young-graduate hiring began before generative AI was publicly available. Whatever started this, it wasn't ChatGPT.

Layer on the macro picture and the AI narrative weakens further. The hiring rate across the whole economy has collapsed to levels normally seen in recessions. Jobless claims ran near multi-decade lows this summer even as payrolls went negative — the signature of a frozen market, not a technological one. When no one is hiring, the people hurt most are always the ones who need a first job. That was true in 1982 and 2009, long before anyone fine-tuned a transformer.

Even the New York Fed's own researchers concluded that AI didn't explain the rise in unemployment among younger workers. An OECD report this summer found young university graduates at elevated risk of unemployment across the developed world — including countries with far less AI adoption than the United States.

Why the Answer Is Worth Money

This isn't an academic dispute. The two explanations imply opposite futures — and opposite trades.

If the freeze story is right, the graduate job crisis is cyclical. It's a product of high rates, post-pandemic over-hiring hangovers, and a remote-work equilibrium that firms are already partially unwinding with return-to-office mandates. In that world, the entry-level market thaws when the Fed eases and corporate confidence returns. The cohort still carries scars — the economic literature on recession graduates finds earnings penalties that persist a decade or more — but the structure of the labor market survives.

If the AI story is right, there is no thaw coming. The bottom rung of the white-collar ladder is being permanently removed, and the 5.7% becomes a floor, not a peak. The implications compound: if firms don't hire juniors, they don't create seniors, and five years from now the experience premium explodes. Mid-career professionals who already have the judgment AI can't replicate become the scarcest asset in the economy. The ROI on a generic four-year degree — already under attack politically — gets repriced in earnest, with consequences for the $1.6 trillion student-loan complex, tuition-dependent universities, and every lender underwriting young-adult credit.

The honest reading of the evidence: this is mostly a freeze with an AI accelerant in specific occupations. The Stanford data is real, but it's concentrated — software, customer support, clerical work. The NY Fed data explains the breadth. Both can be true, and both are.

What's not in dispute is who pays. A 22-year-old who can't get a first job doesn't form a household, doesn't buy the starter home (existing-home sales just hit a three-month low with the median price falling), doesn't build the credit file, and delays every downstream purchase the consumer economy is built on. Roughly four million Americans graduate into this market every year. The compounding cost of a scarred cohort is one of the least-priced macro risks of the decade — and in February, when the AFL-CIO put young workers on the streets of Washington in the first "Young Worker March," it became a political fact too. In an election year, expect policy aimed squarely at this cohort: hiring credits, loan relief, apprenticeship subsidies. Someone will monetize that.

What to Watch

The hires rate in the monthly JOLTS report — the single cleanest freeze/thaw indicator. A thaw shows up there before it shows up anywhere else.
Stanford's Canaries Dashboard — if the 3.8% annual decline in AI-exposed young employment keeps accelerating after the Fed eases, the structural story wins.
Fall campus recruiting — Handshake posting volumes for the class of 2027 will reveal whether employers are re-opening the pipeline or closing it for good.
Return-to-office follow-through — if the NY Fed is right, RTO mandates should quietly improve junior hiring in 2027. Watch for it.

The entry-level job isn't gone. But it's being rationed — by interest rates, by office attendance, and at the margin by software that does what juniors used to do. The market is treating this as a soft-landing footnote. It's closer to a slow structural repricing of how careers begin.

If this analysis was useful, this is what AlphaBriefing does every day — geopolitics and markets, connected to what it means for your money. Free members get the daily brief in their inbox; paid members get the investment frameworks, scenario pricing, and catalyst calendars behind the paywall.

Get this level of intelligence every day. Subscribe to AlphaBriefing — free, member, and paid tiers available.

Sources & Further Reading

Disclaimer

AlphaBriefing is an independent intelligence publication. The content in this article is produced for informational and educational purposes only. Nothing published by AlphaBriefing constitutes financial, investment, legal, tax, or regulatory advice, nor should it be construed as a solicitation or recommendation to buy, sell, or hold any security, asset, or financial instrument.

All views expressed are those of the author at the time of writing and are subject to change without notice. Markets are volatile and unpredictable; past performance is not indicative of future results. Any investment involves risk, including the possible loss of principal.

AlphaBriefing and its principals, employees, or contributors may hold positions in securities or assets mentioned in this article. This should be considered a potential conflict of interest. No material relationship with any company referenced exists unless explicitly disclosed. Readers should conduct their own due diligence and consult qualified financial, legal, and tax advisors before making any investment decisions.

Information in this article is drawn from public sources believed to be reliable at the time of publication. AlphaBriefing makes no warranty, express or implied, as to the accuracy, completeness, or timeliness of any information herein. AlphaBriefing accepts no liability for any loss or damage arising from reliance on this content.