Entry-level jobs are disappearing, AI is accelerating the shift, and inflation is making the labor market feel worse than the headline suggests.
KEY TAKEAWAYS
The labor market headline looks healthy, but the underlying structure is weakening. The unemployment rate may be 4.3%, yet entry-level hiring pipelines are clearly deteriorating.
Consumption remains the backbone of the economy, which is why employment matters more than almost anything else. High inflation and energy prices become much more dangerous when workers begin losing confidence in their ability to earn.
Entry-level white-collar hiring has contracted sharply over the past two years. Industries that once absorbed college graduates are now shedding jobs instead of adding them.
The economic value proposition of a four-year degree is changing rapidly. Skilled trades are now producing better employment outcomes than many traditional white-collar career paths.
Artificial intelligence is accelerating labor market restructuring at the junior level. Companies are increasingly redirecting spending away from entry-level labor and toward AI infrastructure and efficiency initiatives.
MY HOT TAKES
The unemployment rate has become an increasingly misleading economic indicator. It masks where labor weakness is actually occurring and creates a false sense of stability.
The first real casualties of AI disruption are not senior executives or highly paid specialists. They are young workers trying to get their foot in the door.
America sold an entire generation on a college-to-career pipeline that no longer exists in the same form. The market is repricing the value of white-collar credentials in real time.
Gig work and freelance work are being normalized as substitutes for stable employment. That may help companies manage costs, but it does not build long-term consumer confidence or financial security.
The labor market problem is not simply about unemployment. It is about underemployment, stalled career formation, delayed wealth creation, and the slow erosion of future consumption power.
You can quote me: “The return on a four-year degree has been degrading for a decade, but this is the first time the math has actually inverted.”
Sub-surface. I am obsessed! Obsessed with consumption–my long-time followers know that and are likely rolling their eyes as they read these words. Why the obsession? Because we–consumers–make or break the economy, contributing some 70% to GDP. Is the price of crude oil important? Indeed it is, if it affects consumption patterns. $4.50 gasoline means folks are spending $20 - $50 more per fill up than they were a few months ago. That is $20 - $50 less than they have to spend on food. If you don’t need a car to get to work, great. Walk–a little exercise is good for you. Do you use public transportation? Well that is going up too, so you are paying up no matter what. In fact–as I have detailed ad nauseum–high energy costs will have inflationary effects across most of the things that we consume in one way or another–at one point or another. So, indeed, energy cost is important.
But there is something more important than energy costs, or even inflation itself. And that is employment–the labor market. I like to say that you can buy gas at any price if you don’t have a job. There it is, employment. My other famous quote: confident consumers consume–employed consumers are confident. So, because I am obsessed with consumption, I am naturally very much focused on the labor market. Should I be concerned about the labor market? Well, according to the headlines–er, no. According to last Friday’s BLS labor situation, the US added 115,000 new jobs in April and the unemployment rate is 4.3%, which is healthy according to the Fed’s central narrative. If you agree, you can just skip the rest of this blogpost / newsletter. If something doesn’t feel right about that–keep reading.
Let me tell you about a young woman who went viral a few weeks ago. A 2026 college graduate –freshly minted degree in hand, student loans ticking in the background–posted a video explaining that she had submitted over a hundred job applications without landing a single interview. Not a callback. Not a rejection email. Silence. The internet did what the internet does: half the comments told her she wasn't trying hard enough, and the other half said she'd perfectly articulated something they hadn't been able to put into words. Well, my friends, I fear that she, indeed, had. And the data backs her up completely.
Here is what the sub-surface of the labor market actually looks like. Entry-level job listings on Indeed (a hiring website) fell 7% in 2025 compared to the year prior. LinkedIn data shows entry-level hiring dropped another 6% between December of last year and February of this one. Those aren't rounding errors. That is a structural contraction in the pipeline that new workers depend on to enter the economy. And this is happening while the headline unemployment rate sits at 4.3% and Washington congratulates itself on a healthy jobs market.
The jobs that are being created are not the jobs that college graduates are chasing. April's 115,000 Nonfarm Payroll additions came largely from health care, transportation and warehousing, and retail trade–honorable work, every last bit of it, but not what four years and $60,000 in tuition debt was supposed to buy. Meanwhile, the sectors that historically served as the on-ramp for white-collar graduates–finance, information services, professional services– have been hemorrhaging jobs at an average of 9,000 per month since 2023. Before the pandemic, those same industries were adding 44,000 jobs per month. That is not a slowdown. That is a structural reversal. And it doesn't show up in the headline unemployment rate because the headline unemployment rate doesn't care where the jobs are–only that they exist.
There is a data point that I keep coming back to, and it is one of the most remarkable things I have seen in nearly forty years of watching this market. For the first time since the federal government began tracking these figures, college graduates have lost their employment advantage over workers without degrees. In the middle of last year, workers with a skilled-trades associate's degree–plumbers, electricians, pipe fitters–posted better employment outcomes than college graduates. Not roughly equivalent. Better. The return on a four-year degree has been degrading for a decade, but this is the first time the math has actually inverted. That is a generational shift in the labor market's underlying structure, and it is buried on page fourteen of reports that lead with the headline number.
Now layer in the accelerant. Artificial intelligence is not the cause of this problem–the entry-level contraction was already underway before every corporate earnings call started featuring the phrase "AI-driven efficiency." But AI has compressed the timeline dramatically and is doing so first at the bottom of the white-collar pyramid, which is precisely where new graduates are trying to enter. Anthropic’s CEO has said publicly that within five years, half of entry-level white-collar jobs may disappear. Companies are not waiting five years. Block cut 40% of its headcount this year, explicitly citing AI efficiency, and told other firms to follow suit. Meta followed with a 10% workforce reduction in the middle of heavy AI capital spending. Oracle. Microsoft. The pattern is identical in every case: replace labor at the bottom, redirect the savings to infrastructure at the top. New hiring budgets do not survive that reallocation.
The graduates are adapting–I will give them that. Nearly 38% of the class of 2026 are now considering starting their own business. Another third are looking at gig work. A quarter are exploring freelance arrangements. This is not the economy these young people were sold when they enrolled, and the resourcefulness they are showing deserves respect from people who graduated into a very different labor market. But gig work does not come with health insurance. Freelance income does not automatically fund a 401(k). And none of it resolves the inflation problem sitting on top of all of this.
Which brings me back to my obsession. Inflation hit 3.3% year-over-year through March, driven in large part by energy–gasoline alone accounted for nearly three-quarters of that month's entire price increase. The April read is out this morning and I expect it to come in hotter. With the Strait of Hormuz still disrupted and crude prices elevated, the energy pass-through into everything else is not finished working its way through the system. Futures traders, for their part, have essentially abandoned any expectation of a Fed rate cut in 2026. JPMorgan, even in its most optimistic scenario, does not see inflation returning sustainably to the Fed's 2% target before early next year. So the cost of carrying student debt stays elevated. The cost of credit card balances stays elevated. The cost of…um, existing stays elevated—for the cohort least equipped to absorb it.
That is the compound stress scenario that doesn't make it into the headline: a young worker who cannot find a full-time job is simultaneously failing to build savings, failing to pay down debt, and watching whatever cash cushion they have erode in real terms every month that inflation runs above their income growth. A confident consumer consumes. An unemployed consumer–or an underemployed one, stuck in part-time work they didn't choose–does not. And that 4.9 million number, the Americans working part-time in April because they couldn't find full-time work, is up nearly half a million from last month. That is not a healthy jobs market. That is a healthy-looking headline with a very different story underneath it.
So what do you actually do with this? You stop using the headline unemployment rate as your primary labor market indicator and start watching the sub-surface numbers: entry-level job postings, sector-specific hiring trends, real wage growth for workers under thirty, and the part-time-for-economic-reasons figure that BLS publishes every month and almost no one reads. If you are a young person navigating this market, the skilled trades are no longer a fallback, they are, right now, statistically better employment bets than a white-collar job search in finance or information services. If you are an employer, the talent pipeline you are not building today is the management problem you will have in five years. And if you are a policymaker, a headline unemployment rate that tells you everything is fine while a generation enters the workforce into structural contraction is not a policy tool–it is a public relations document.
The viral graduate was not complaining. She was reporting. And the data says she got it exactly right.