A degree used to be a safety net. The data says it isn't anymore.

A college degree used to be the most reliable economic safety net in America. The data no longer supports that assumption.

According to the Federal Reserve Bank of New York, recent graduate unemployment hit 5.7% in Q4 2025, nearly 1.5 percentage points above the overall unemployment rate of 4.2%. That gap is the widest it's been in over a decade.

Meanwhile, entry-level white-collar job postings, the jobs college graduates have historically walked into, have collapsed. According to data from Indeed and Burning Glass Institute:

• Entry-level software development postings: down 67% since 2018
• Entry-level financial analysis postings: down 56%
• Entry-level marketing & communications postings: down 46%

The lines have been diverging since mid-2023, with recent graduate unemployment climbing while overall rates held steady.

Is AI the cause?

It's part of the picture, but not the whole picture.

The entry-level knowledge-work collapse began before generative AI went mainstream. Automation of routine analytical tasks, offshoring of early-career functions and the flattening of corporate hierarchies were already reducing the number of "starter" white-collar jobs. AI has accelerated the trend but didn't create it.

What AI has done is raise the bar. The entry-level jobs that remain increasingly require applied technical skills like data analysis, prompt engineering, workflow automation and systems thinking that most undergraduate programs don't explicitly teach. The gap isn't between "having a degree" and "not having a degree." It's between having applicable skills and not having them.

Software development entry-level postings took the hardest hit, losing two-thirds of available positions in six years.

Three things about how Texas is responding

1. Texas is scaling apprenticeships faster than any other state

The Texas Workforce Commission reports 40,000+ active registered apprentices statewide, a number that has grown over 300% since 2014. Texas now has more registered apprenticeship programs than any state except California and it's growing faster.

This isn't just construction and trades. Texas apprenticeships now cover IT, healthcare, advanced manufacturing, cybersecurity and financial services. The TWC's apprenticeship expansion aligns with HB 3260, which allocates state funding for earn-and-learn models that produce industry-recognized credentials.

A 300% increase in a decade. Texas is now second only to California in total registered apprenticeship programs.

2. HB 3260 is reshaping how Texas funds workforce education

House Bill 3260, passed in the 88th Texas Legislature, does two things:

1. It ties community college funding to student outcomes (credential completion, employment, wage gains) rather than enrollment alone.
2. It creates dedicated funding streams for earn-and-learn programs like apprenticeships, clinical rotations, co-ops and other work-based learning models.

This is a fundamental shift. Institutions that can demonstrate a direct connection between their programs and employment outcomes will receive more state funding. Those that can't will receive less. The incentive structure now rewards exactly what the underemployment data demands: alignment between education and employment.

3. The geographic hiring shift is real

Corporate relocations to Texas continue at pace. Since 2020, 200+ companies have announced relocations or major expansions in the DFW metroplex alone. These companies are hiring, but they're hiring for specific skills, not generic degrees.

This creates a paradox: Texas has more open jobs than almost any state, but its universities are producing graduates who don't match the demand profile. The gap between where the jobs are and what graduates can do is the central workforce challenge of the next decade.

What forward-thinking leaders are doing

• University provosts are running OBBBA simulations on every program and building "Program Health Dashboards" that combine enrollment trends, completion rates, placement data and earnings outcomes.
• EDC directors are packaging workforce pipeline data into site selection proposals, showing relocating companies not just available land and tax incentives but a trained, ready talent base.
• Workforce board executives are mapping apprenticeship completions to industry sectors, building the data case that earn-and-learn models deliver better outcomes per dollar than traditional classroom-only approaches.

From the Field

"We stopped asking 'how many students enrolled' and started asking 'how many graduates are working in their field at a living wage within six months.' That single metric changed every conversation at the leadership table."

Workforce development director, Texas public university system