Pay for Performance

As the feds roll out $145M in performance-based funding for apprenticeship, California and Maryland offer examples of how to do it well. Also, a vibe shift on AI as potential job displacement looms larger, and essays by DeRionne Pollard on why community colleges need better storytelling and from Pew on its research about how students use and pay for short-term credentials. (Subscribe here.)

Taking Politics Out of Apprenticeship Funding

Performance funding is having a moment in the apprenticeship space. The federal government recently announced that $145M in spending on registered apprenticeship would be distributed under a pay-for-performance model, following examples set by California, Iowa, and Maryland.

The U.S. Department of Labor has yet to unveil specifics about its approach. But the feds can draw lessons from California, by far the largest U.S. pay-for-performance experiment to date, and one that’s largely been a success.

Colleen Connolly reports for Work Shift about the program California launched four years ago. Intermediary organizations or other sponsor groups get $3,500 a year from the state for each active apprentice, with an additional $1K when the apprentice completes.

The goal is to create a more stable stream of money for apprenticeship intermediaries, which typically must chase one-off government grants.

“We’re looking to create that sustainability and growth pathway for program sponsors that are meeting the needs of employers, supporting apprentices through their journey, and ultimately really scaling this apprenticeship system,” Adele Burnes, chief of the California Division of Apprenticeship Standards and the primary architect of the new model, told Connolly.

The program has boosted the number of apprentices in targeted industries by 76% over the last three years, supporting almost 30K apprentices so far. The early childhood education sector has been one of the largest recipients of the state funding, along with energy and natural resources, transportation, healthcare, advanced manufacturing, and IT.

Early Care & Education Pathways to Success is a California-based sponsor of registered apprenticeships in the childcare and early learning industry. The nonprofit has benefited from the discretionary funding, which features less paperwork and eliminates the sort of competition that can be gamed through politics.

“If you have 500 apprentices, you’re getting 500 times $3,500, and if you have three apprentices you’re getting three times $3,500,” says Randi Wolfe, executive director of ECEPTS. “Power and influence and the old boys’ club will not have any bearing. It takes out the politics.”

While pay-for-performance has led to more predictable and sustainable funding for apprenticeship in California, the program faces an uncertain future in a tight budget year. Connolly reports that it may see a big budget decrease, or even be fully eliminated.

Likewise, the $145M in funding from the federal government is currently slated to last for just four years. The money also isn’t new and will be drawn from the primary federal workforce system.

If the funding stream were to pay out $3,500 per apprentice, it would cover about 40K apprentices—a drop in the bucket as the Trump administration seeks to create 300K+ more apprenticeship slots each year to hit its goal for 1M active apprentices nationwide. And even if the U.S. reached that mark, it would still lag behind many countries where apprenticeship is far more established.

Click over to Work Shift to read Connolly’s in-depth reporting on California’s approach.

Going Beyond Recruitment: The Labor Department also can learn from Maryland’s new pay-for-performance funding stream, Lancy Downs, a senior policy analyst at New America, writes in a blog post for the nonprofit group.

The Maryland Apprenticeship Incentive Program rewards sponsors and employers for the retention of apprentices, not just the recruitment, she writes. It also prioritizes youth apprentices by more than doubling the funding for each high school–level apprentice—$7,500 for those participants compared to $3K for each new adult apprentice in a federally registered program.

Funding this year under the program would support a maximum of 1,600 apprentices, about a third of Maryland’s target for 5K new apprentices.

While it will be a while before results emerge from the pay-for-performance strategy, Downs says the state’s design is thoughtful and intentional.

Vibe Shift on AI

Experts who watch artificial intelligence closely say the technology has just leaped forward, with an impact that could rival OpenAI’s introduction of ChatGPT in late 2022.

Key drivers were the release last week of new models from Anthropic and OpenAI. Both make it easy to create autonomous coding agents and manage complex workflows.

“You can just build things and do things,” writes John Bailey, a nonresident senior fellow at the American Enterprise Institute and a keen AI observer. “If you tried AI in 2023 or even 2025 and walked away unimpressed, that experience is no longer relevant. The gap between public perception and current capability is widening much more quickly than people realize.”

Anthropic’s release of industry-specific plugins for its Claude Cowork—a workplace assistant tool and essentially a nontechnical version of Claude Code—helped trigger a major selloff of software stocks last week, as investors increasingly worry that AI could disrupt finance, law, software development, and other industries. The market anxiety has been accompanied by noteworthy predictions that substantial AI-driven job displacement is looming.

Anton Korinek thinks the U.S. will see major job losses as soon as this year. An economics professor who leads the University of Virginia’s Economics of Transformative AI Initiative, Korinek told The Atlantic’s Josh Tyrangiel that he’s worried about AI’s speed. If the tech moves as fast as he expects, the damage will arrive for many workers before institutions can adapt.

Most corporate leaders also have gotten quiet about AI and job loss, Tyrangiel writes in his in-depth take on what comes next. “Anyone who has seen a shark fin break the water and then disappear knows this is not reassuring.”

Paying for Fixes: Billionaires have an obligation to help solve AI-driven job displacement, writes Anthropic CEO Dario Amodei in a recent opus that’s worth reading. So do governments, he says. 

“The natural policy response to an enormous economic pie coupled with high inequality (due to a lack of jobs, or poorly paid jobs, for many) is progressive taxation,” Amodei writes. “The tax could be general or could be targeted against AI companies in particular.”

Government won’t help the AI job transition, Phil Gramm, an economist and former U.S. senator, and Mike Solon, a senior fellow at the Hudson Institute, argue in The Wall Street Journal.

“Previous efforts to cushion the transition from jobs of the past to jobs of the future have done little to benefit those making the transition—and have raised the cost for society as a whole,” they write, citing government studies that have found lackluster results for the federal Trade Adjustment Assistance Program. (Newer research has shown that the program is more effective than once believed.)

Gramm and Solon instead call for a national, AI-powered employment system that would match unemployed Americans with jobs and provide personalized training.

Politics and Policy: The national debate over what to do about AI and jobs might not shake out across a predictable partisan divide. Figures as far apart as Pete Buttigieg and Steve Bannon say the technology will fundamentally reshape politics, requiring new thinking and new coalitions.

For now, however, the vast majority of policy action on AI is happening in state capitals. 

A newly released atlas tracks enacted policies across 50 U.S. states on AI, machine learning, and related technologies and infrastructure. The nonpartisan project from an interdisciplinary team of faculty members at the University of Illinois System and the Institute of Government and Public Affairs can be sorted by legislation across education, healthcare, privacy, business operations, and other categories.

AI and CTE: A new report from CTE Futures takes Wharton professor and AI researcher Ethan Mollick’s concept of co-intelligence—essentially that AI is best incorporated into work as a collaborator—and adapts it for career technical education.

“Mollick’s framework is primarily applicable for knowledge workers whose tasks are abstract, cognitive, and often desk-based,” Cameron Sublett and his co-authors write. “For CTE learners, however, the nature of work can be much different.”

The authors propose a framework for applied co-intelligence that accounts for the more specific technical skills required in CTE fields, emphasizes the importance of durable skills in applied contexts, and doesn’t assume any baseline familiarity with AI.

Open Tabs

Workforce Pell
The U.S. Department of Labor plans to soon announce a grant competition for community colleges, including $65M of funding with the potential of an additional $65M, Henry Mack, the assistant secretary of labor, told community college leaders this week. The Association of Community College Trustees reports that Mack said the grant program would focus on helping the sector grow short-term programs that would be eligible for federal Workforce Pell grants.

CTE and Employment
A decade after graduating high school, 71% of North Carolina students who concentrated in career and technical education were employed in 2024, with median annual earnings of $40K, according to data released by the state’s Department of Commerce. Students who focused on manufacturing had the highest wages (roughly $50K), followed closely by concentrators in transportation, distribution, and logistics and in STEM fields.

Momentum for LERs
The learning and employment record ecosystem is stronger and more visible than when the SkillsFWD project began in 2023, according to a report describing the foundation-backed effort, which featured six grants of $1.4M each for LER-related experiments. The report cites momentum and recent federal interest in talent infrastructure. Yet mass adoption remains elusive, and many learners have yet to benefit from mobile, interoperable skills records.

Skills-First Blueprint
More state and local governments are tapping a blueprint for implementing skills-based hiring, with Opportunity@Work tripling the size of its accelerator project. The nonprofit’s effort now includes 15 participating governments around the country. With support from the Volcker Alliance, which focuses on the public sector workforce, the project offers tailored assistance and a focus on measurable outcomes in both hiring and advancement.

Job Moves
Ben Pring has been hired as vice president of Jobs for the Future’s Center on Artificial Intelligence & the Future of Work. Pring previously held senior roles at Gartner, a business intelligence firm, and at Cognizant, a major tech company.

Thanks for reading. If this newsletter was forwarded to you, subscribe here. —PF