The low-hire economy traps entry-level graduates
The class of 2025 is walking into a paradox. They have degrees, often with honors, sometimes with certificates stacked on top. They have internships. They have projects. And they have an unemployment rate that has climbed 2.2 percentage points faster than any other group in the labor force. The Federal Reserve Bank of St. Louis published this number in July 2025, and it tells a story that is not about a recession in the classic sense. The economy is not shedding jobs. It is simply not creating them for people who have never held one before.
This is the low-hire phase. Businesses are holding onto the workers they already have. They are not firing, but they are also not hiring. For a recent graduate, this is worse than a downturn. In a downturn, companies eventually need fresh talent when the recovery comes. In a low-hire phase, the door is simply not opening. The Fed researchers William M. Rodgers III and Alice Kassens traced the effect across demographic groups. For workers aged 25 to 64, the slowing market added 1.1 percentage points to unemployment. For young workers with only a high school diploma, it added 1.23 points. For recent college graduates, the number was 2.2 points. The gap is not small. It is a structural shift in who gets to start.
The Experience Trap That Nobody Designed
Riley
Reed graduated from Webster University in May 2025 with a bachelor’s degree in advertising and marketing, plus two specialty certificates in the same field She applied to about 15 places. She struggled to get a single interview. The jobs that paid entry-level wages required several years of industry experience. Nobody seemed willing to train a newbie. This is not a story about a bad resume or a weak job market in one city. It is a pattern that the St. Louis Fed identified at the national level and confirmed in the Eighth Federal Reserve District, which covers parts of eight states.
The experience trap works like this. During the tight labor market of 2022 and 2023, companies hired aggressively. They took chances on candidates who needed training. They created pipelines for junior talent. Then inflation hit, economic uncertainty set in, and the hiring spigot turned off. But the experienced workers who were hired during the boom are still there. Companies are keeping them. They are not creating new roles. And when a position does open up, they can demand experience because the pool of available workers includes people who were laid off from other companies or who are looking to move. The new graduate cannot compete with someone who has three years of real work under their belt.
The Fed researchers tested several competing explanations for what is happening to young workers. Maybe it was the drop in the foreign-born population during the Trump era, which reduced consumer demand in sectors like retail where young workers traditionally start. The timing did not match. Maybe it was contraction in manufacturing or the shrinking of federal employment. Neither explained the data. What did explain it was a simple drop in job openings. Overall, openings fell by about a third over the past few years. The main culprit was not AI, not trade policy, not demographic change. It was the simple fact that businesses stopped hiring new people.
The AI Skill Requirement That Blocks the Door
The St. Louis Fed found that AI-related job requirements have added a separate headwind for recent college graduates, one that is smaller than the broader slowdown but still meaningful. The researchers calculated that AI-related labor market changes have increased the unemployment rate for recent college graduates by 1.68 percentage points. That is more than five times the impact for workers aged 25 to 64 and eight times higher than for workers aged 18 to 24 with no more than a high school diploma.
This is not about AI replacing jobs. It is about AI raising the bar for entry. A survey by the National Association of Colleges and Employers found that more than a third of employers now require entry-level workers to have AI skills. That sentiment has grown three times compared to the fall of 2025. The jobs that college graduates typically enter are more exposed to AI than trade or hands-on occupations. An accountant, a marketer, a financial analyst, a writer — these roles are being reshaped by AI tools. Employers want candidates who already know how to use them.
The problem is that most college students say their institutions are not effectively incorporating AI into degree programs. The skills gap upon graduation is real. A graduate who spent four years learning traditional methods arrives at a job market that expects them to already be fluent in tools that did not exist when they started college. The result is a mismatch. The jobs are there, in theory. But the requirements have shifted faster than the curriculum.
The Historical Precedent That Is Being Ignored
The low-hire, low-fire economy has happened before. After the dot-com crash of 2001, companies stopped hiring new graduates for years. The class of 2002 faced unemployment rates that stayed elevated well into the recovery. The same pattern repeated after the financial crisis of 2008. Graduates who entered the labor market during those years experienced lower lifetime earnings, slower career progression, and higher rates of job dissatisfaction that persisted for a decade or more.
The reason is simple. The first job matters. It sets the trajectory. A graduate who starts in a low-hire economy is more likely to take a job below their qualifications, which depresses their starting salary. That lower salary compounds over time because raises are typically percentage-based. They are also less likely to receive training, because companies that are not hiring are also not investing in development. And they are more likely to change jobs frequently in the early years, which signals instability to future employers.
The current moment has one difference that makes it worse. The AI skill requirement is not cyclical. It is structural. Even when the broader labor market recovers, the bar for entry will not return to where it was. The jobs that used to be available to a generalist with a degree now require specific technical competence. The graduate who missed that window will not get a second chance to learn on the job.
The Question of the Standard
Who defines what is good enough for an entry-level job? The answer used to be the employer, but the standard was implicit. A bachelor’s degree signaled that the candidate could learn, follow instructions, and complete tasks. The employer would fill in the specific skills through training. That model is breaking down.
The St. Louis Fed researchers found that the increase in job postings requiring AI skills is cutting out more young workers. The employers are not being unreasonable. They are responding to a real change in how work gets done. A marketing associate who cannot use AI tools for content generation or data analysis is less productive than one who can. A financial analyst who cannot automate routine tasks is slower. The problem is that the standard has shifted faster than the supply of workers who meet it.
The result is a market failure. Employers want experienced workers with AI skills. New graduates have neither. The employers wait for the experienced workers to become available. The graduates wait for the employers to lower their standards. Neither side moves. The unemployment rate for recent college graduates stays elevated.
The Insight That Is Technically Solved but Socially Unresolved
The technical solution to this problem is straightforward. Universities need to integrate AI skills into their curricula faster. Companies need to create training pipelines for new graduates. The government could fund apprenticeship programs or subsidize entry-level hiring. All of these are possible. None of them are happening at the scale required.
The social problem is harder. The low-hire economy is a symptom of a deeper shift in how businesses think about labor. During the pandemic, companies discovered that they could do more with fewer people. Automation and AI accelerated that trend. The recovery from the pandemic was supposed to bring back hiring, but it did not bring back the willingness to train. The risk of hiring a new graduate and investing in their development now seems too high compared to the certainty of hiring someone who already knows the job.
The Fed researchers put it plainly. Since April 2023, hiring has slowed, and young workers, especially new entrants, have borne the brunt of that softening. The AI adds an additional headwind at the point of labor market entry, but its effects remain smaller than those of the broader decline in job openings. The main problem is not AI. It is the simple fact that businesses have stopped creating the roles that used to be the first step on the career ladder.
Riley Reed eventually got an interview. She does not know if she will get the job. She is one of thousands of graduates in the same position. The low-hire economy is not a temporary blip. It is the new normal, at least for now. And the question that nobody has answered is what happens to a generation that cannot get its foot in the door. The answer may determine not just their careers, but the shape of the labor market for decades to come.