“Frontline Workers in High Demand, College Graduates Struggle to Find Jobs”: The ‘College Premium’ Erodes as Foundational Skills Decline and AI Takes Over Entry-Level Work

Aging frontline workforce and rising graduate supply reverse labor-market dynamics
AI displaces entry-level office work, narrowing hiring prospects for new graduates
University credentials lose signaling power as degree-based hiring recedes 

Employment prospects for young Americans are diverging sharply between college graduates and those without degrees. Hiring has increased for frontline and in-person roles in construction, manufacturing and care services amid labor shortages, while opportunities have contracted in the office and professional occupations dominated by college graduates as the supply of qualified workers expands and artificial intelligence (AI) proliferates. AI’s displacement of entry-level tasks traditionally assigned to recent graduates has contributed to the downturn, but confidence in college degrees had already weakened amid declining foundational skills and grade inflation. As universities lowered academic and graduation standards to retain students, they produced inadequately trained workers, prompting employers to prioritize demonstrated skills and practical experience over academic pedigree. With AI also accelerating the turnover of knowledge and skills, the retreat from degree-based hiring is expected to gain further momentum unless universities restore the rigor of their admissions and educational standards.

Labor-Market Indicators for College Graduates Aged 22–34 at Their Worst Since 2003

According to the labor-market research organization Burning Glass Institute on the 20th, the unemployment rate among Americans aged 22–34 without college degrees was close to its lowest level in two decades. Employment conditions for college graduates in the same age group, by contrast, have deteriorated. The only periods over the past 20 years when their employment prospects were worse were during the COVID-19 pandemic and the sluggish recoveries that followed the 2007–2009 global financial crisis. Shifts in labor supply and demand lie behind this trend. Skilled workers in construction and manufacturing are retiring rapidly as the workforce ages, but too few new workers are entering these fields to replace them.

The number of college graduates, meanwhile, continues to rise. As the population of Americans with bachelor’s degrees reaches a record high, companies are increasingly using AI to perform the entry-level tasks traditionally assigned to newly hired graduates. According to research published by the Federal Reserve Bank of Dallas on the 1st, job postings in occupations with high exposure to AI in Texas—the center of the AI industry—declined by approximately 8% more than those in occupations with low exposure. In addition to software development and web design, white-collar occupations spanning management, administration and editing also showed high levels of AI exposure. The researchers suggested that recent college graduates may be bearing the initial impact of AI adoption as companies reduce new hiring rather than dismiss existing employees.

By contrast, unemployment rates were lower than in the past for occupations involving a high proportion of manual or in-person work, including blue-collar construction and manufacturing jobs, security guards, nursing and care work, and restaurant and retail employment. August employment data reflected the same pattern. Food services and drinking establishments added 59,000 jobs in a single month, accounting for more than one-third of the 162,000 jobs created across the United States during the same period. The Burning Glass analysis similarly found that employment conditions had improved for in-person and frontline workers such as construction and manufacturing production employees, security guards, nurses, restaurant workers and retail staff, while science, technology, engineering and mathematics (STEM) occupations and managerial and professional positions remained comparatively weak.

Declining Foundational Skills Erode the “Degree Effect”

Declining foundational skills have also contributed to the employment difficulties facing young college graduates. Although college enrollment and graduation have become commonplace, a degree no longer provides an adequate guarantee of practical workplace competence. In the 2024 National Assessment of Educational Progress (NAEP) administered by the National Center for Education Statistics (NCES) under the U.S. Department of Education, 45% of 12th-grade students were rated “below basic” in mathematics and 32% in reading. Only 33% of students were deemed prepared for college-level mathematics coursework, down from 37% in 2019.

Mathematics scores fell to their lowest level since the current assessment framework was introduced in 2005, while reading scores reached their lowest point since testing began in 1992. Learning losses during the COVID-19 pandemic deepened the decline, but the deterioration in academic achievement had already been under way since the mid-2010s. In the Organisation for Economic Co-operation and Development’s (OECD) 2025 Programme for International Student Assessment (PISA), U.S. students recorded a mathematics score of 463, trailing the 612 points achieved by students in Beijing, Shanghai, Jiangsu and Zhejiang by 149 points. The share of top-performing mathematics students was also 8% in the United States, compared with 54% across the four Chinese regions. Even allowing for the limitation that China’s results were calculated using selected regions with strong educational conditions, the figures demonstrate a substantial gap at the foundational education stage responsible for cultivating talent for advanced industries.

Table 1. Disconnect Between Foundational Skills and Grades Among U.S. College Students

Collapse in Confidence in Degrees and Grades

A disconnect between school grades and actual academic proficiency was also evident during the transition to higher education. According to a report released this year by the University of California, San Diego (UCSD), the number of incoming students who lacked high-school-level mathematics skills increased nearly 30-fold between 2020 and 2025, reaching one-eighth of all freshmen. One in 12 incoming students failed to demonstrate even middle-school-level proficiency. Among these students, 94% had taken more courses than the minimum high-school mathematics requirement for college admission, while 42% had completed calculus or precalculus. Their coursework and transcripts suggested that they were prepared for college education, yet placement tests determined that they required remedial instruction covering elementary- or middle-school material.

Even after students entered college, rising grades diverged from actual learning outcomes. UC Berkeley researchers analyzed approximately 500,000 grades awarded at large U.S. research universities between 2018 and 2025 and found that, following the launch of ChatGPT in 2023, the share of A grades in courses readily amenable to generative AI use, such as writing and coding, rose by 13 percentage points. This represented an increase of approximately 30% from 2022. Citing the larger gains in courses where assignments accounted for a substantial share of final grades, the researchers attributed the increase more to AI completing coursework on students’ behalf than to broad improvements in learning outcomes.

As degrees and grades became less reliable indicators of applicants’ capabilities, companies began to change their recruitment practices. In this year’s survey by the National Association of Colleges and Employers (NACE), approximately 70% of companies said they used “skills-based hiring,” which evaluates candidates according to demonstrated competencies rather than academic qualifications, while nearly every respondent identified internship experience in the United States as an important criterion. Data compiled by the Federal Reserve Bank of New York in the second quarter of the same year showed that 42% of college graduates aged 22–27 were employed in jobs that did not require a college degree. In practice, skills-based hiring was used primarily during interviews (87%) and candidate screening (65%), while the share of companies using grade point averages as an evaluation criterion plunged from 73% in 2019 to 42% this year.

Soaring Costs, Diminishing Returns

Belief in a university’s name as a guarantee of employment and income had already begun to weaken before the advent of generative AI. A Pew Research Center survey of 3,445 U.S. adults last year found that 70% believed American higher education was heading in the wrong direction, 14 percentage points higher than the 56% recorded in 2020. Notably, negative assessments were more prevalent among college graduates, at 74%, than among those without degrees, at 69%. Seventy-nine percent said colleges were failing to keep tuition affordable, while 55% said they were not preparing students for well-paying jobs. Another 49% said colleges were failing to develop students’ critical-thinking and problem-solving abilities adequately.

Skepticism over the return on educational spending was equally pronounced. In last year’s Pew Research Center survey, only 25% of respondents said a four-year college degree was extremely or very important for securing a well-paying job, while 49% said its importance had diminished compared with 20 years earlier. Only 22% believed college was worth attending even if doing so required student loans. By contrast, 47% said it was worthwhile only when loans were unnecessary, while 29% said it was not worth the cost. According to the U.S. Federal Reserve System (Fed), 48% of bachelor’s degree holders and 54% of graduate degree holders used student loans to finance their education. As high tuition costs collide with uncertain employment outcomes, the longstanding practice of treating a university’s name as a guarantee of career prospects is also losing ground.

AI Dismantles Degree Barriers and Raises the Value of Skills

Skepticism about the value of degrees is also evident in changes to corporate hiring criteria. Following the 2007–2009 global financial crisis, companies imposed college-degree requirements even for jobs where such qualifications had little practical relevance, using them as a convenient means of screening applicants. As labor shortages intensified, however, employers began moving in the opposite direction by lowering degree barriers. An analysis of job postings by Harvard Business School and Burning Glass Institute found a marked decline in degree requirements between 2017 and 2019 across 46% of middle-skill occupations and 31% of high-skill occupations. Researchers at the time estimated that, if the trend continued, an additional 1.4 million jobs could become accessible to workers without college degrees over five years.

The proliferation of AI has accelerated this trend. Global consulting firm PricewaterhouseCoopers (PwC) analyzed approximately 1 billion job postings across six continents and found that, between 2019 and 2024, the skills demanded by employers changed 66% faster in occupations with high AI exposure than in those with low exposure. The proportion of jobs requiring degrees in occupations with significant automation potential also fell by 9 percentage points, from 53% to 44%, over the same period, outpacing the 5-percentage-point decline in occupations with low AI exposure. Jobs requiring AI skills, by contrast, paid an average of 56% more than other positions in the same occupations. PwC concluded that as AI lowers the barriers to acquiring specialized knowledge and shortens the replacement cycle of existing expertise, the ability to apply current skills and exercise critical judgment is carrying greater weight in recruitment than past academic credentials.