Predictions 2026
Drumroll please
Is February too late for some predictions? We say no!
Michael Horn is one of the most forward-thinking minds in education. He’s a prolific author, a long-time thought partner, and now our first-ever repeat guest on Purposeful Paths. (I hope that makes it to the top of his resume quickly.)
My only complaint about Michael is he’s so prolific that I can’t keep up with all of his great work.
This is my first time writing and sharing publicly my predictions, which is a moment to pause. What’s the point?
Hopefully, the point of predictions isn’t to be right. It’s to force clearer thinking about where we’re headed, and what we should be doing about it. I’m sharing to sharpen my own thinking. I hope the conversation sharpens yours, too.
If you want to watch the full conversation, you can do so here. And I pulled some short clips from the conversation that I think are worth sharing. More on those below.
TL;DR
The entry-level labor market will get worse, not better
The Skills-based hiring movement loses momentum
Cosmetology schools fail the gainful employment measures
Social media bans for youth expand to 10 countries
Trump accounts will revolutionize philanthropy
Mid-size private colleges continue to get squeezed
Last Year’s Scorecard
Last year, Michael made four predictions. Here’s how they held up.
1. Growing momentum for apprenticeships, particularly in non-traditional sectors. Grade: C+
Michael’s take? “More talk than action.”
We were expecting a lot from the incoming Trump administration on apprenticeships. That momentum didn’t materialize at first. In fact, the opposite happened. But there’s been a correction since, including a $145 million Department of Labor funding announcement for pay-for-performance apprenticeships. California announced $30 million in apprenticeship funding focused on healthcare and education. Indiana and Colorado continued to make progress.
2. Increased focus on work-based and experiential learning at all education levels. Grade: B+
This one held up. Colorado’s executive order to reorganize the state’s talent development system. California’s additional $64 million for CTE programs. Indiana’s new diploma is creating more concrete experiential learning requirements. Networks like CAPS and Big Picture Learning are growing. And AI’s impact on entry-level hiring has accelerated the urgency of this conversation.
Burning Glass and Strada’s “Talent Disrupted” report found that college students who had an internship were almost 50% less likely to be underemployed after graduation.
3. Less linear education-to-career pathways becoming the norm. Grade: B
More of an ongoing trend than a single-year prediction. The narrative continues to shift, but we’re still early.
4. Continued pressure on traditional colleges as students seek alternatives. Grade: A-
Around 16 colleges closed in 2025. Multiple mergers. The demographic cliff is now here. Michael and his team have expanded their analysis from 44 New England colleges to 315 mid-size private colleges nationally. A third of them face financial solvency challenges in the next five years. Of the original 44 New England schools, 38 list growth as their primary strategy.
Our Predictions for the Year Ahead
Prediction 1: The entry-level labor market will get worse, not better.
There’s a lot of attention on AI’s impact on entry-level jobs, and it’s real. But we don’t talk enough about the broader context: economic uncertainty, trade policy disruption, and the lesson tech companies learned after overhiring during the pandemic that they can maintain growth with fewer people.
Michael agreed, and added an important frame. Companies now expect entry-level hires to perform at year two or three. That requires experience and cultural knowledge that favors those who are well-off and have strong social capital.
There is some room for optimism: entry-level roles are growing in low-AI-exposure fields like healthcare, skilled trades, and specialized services.
At Willow, we believe AI fluency & literacy is career readiness. We’re integrating AI fluency into our curriculum because, frankly, I can't imagine hiring someone without expertise in AI tools, and a infectious drive to learn. We actually launched a dedicated AI fluency quest earlier that day, and a second partner called asking to pilot it before we’d even announced it.
Prediction 2: Skills-based hiring will face a more public reckoning.
Last year, Michael and I were murmuring quietly about this. This year, we’re more convicted.
Skills-based hiring is not working in practice. Only 1 out of 700 hires benefited from skills-based hiring practices, according to the Burning Glass Institute. I’m fairly smart, and even I find it too hard, too complicated, and the incentives don’t align for companies.
The aims are right. The goals are right. But this was a group of funders and policymakers who got together and said, wouldn’t it be great if this world existed? And it would be great. But it doesn’t.
That’s bad news for students without networks. Which is exactly why career-connected learning and work-based learning that build social capital matter so much.
Prediction 3: Cosmetology schools are the canary in the coal mine.
Michael predicts that under new gainful employment measures, almost no cosmetology schools will pass the earnings test of whether graduates earn at least what a high school graduate does. 91-92% of cosmetology programs have a negative ROI. That means states will be forced to radically rethink licensure for these roles, or accept that nobody new gets trained (or they self-pay).
The bigger implication: if this forces a conversation about unnecessary licensure barriers in cosmetology, maybe it ripples into healthcare, nursing, and other fields where training bottlenecks are creating artificial shortages.
Michael is net positive on the gainful employment exercise, and so am I. He’d prefer regionality over state-level metrics, and an ROI focus rather than just a minimum earnings floor. That’s actually exactly how we do our personalized ROI projections at Willow: we project net cost and take a 10-year view of whether a student will be better off. Then we let families decide.
Our belief is informed choice. Be honest with young people. They don’t have the data when they’re making decisions. They deserve it.
Prediction 4: 10 countries will follow Australia’s lead and ban social media for students under 16.
I think three of those countries will also ban AI companions for young people. Here’s a stat I find mind-blowing: 70% of young people have tried AI companions, and half are regular users.
I’m here for it. I think this is good for the future of our society.
Countries I’m watching: France, Spain, Denmark, UK, Germany, New Zealand, Malaysia.
Michael surprised me by leaning closer to my position than I expected. He takes Tyler Cowen’s free speech concerns seriously. But he’s a Wait Until 8th household for his own kids.
Where we disagreed: I support school-level bans on cell phones. Michael wants to empower individual educators to make that call, because he wants room for productive uses of phones in schools. It’s a fair point, and I think it’s too hard for individual educators/schools to own that battle.
Prediction 5: Trump accounts (baby bonds) will revolutionize philanthropy and economic mobility.
This one surprised Michael. (”I did not expect you to go there.”)
The concept is simple: the government establishes an investment account for every child at birth, seeded with public funds.
Here’s my thinking. Baby bonds create infrastructure for direct investment in students’ futures. Philanthropists will see this and realize they no longer need to accept 30% off the top for overhead, with an uncertain programmatic impact. The recent Dell commitment more than doubled the foundation’s total historical giving in a single commitment.
Michael agreed on the structure, noting it builds an ownership society. I think we’re going to see a lot more philanthropy flow through this channel.
6: The higher ed squeeze
Michael’s ongoing prediction that will continue to be true: mid-size private colleges are in trouble with the demographic cliff. His team analyzed 315 institutions enrolling between 1,000 and 8,000 students. A third face financial solvency challenges within five years.
No higher ed leader currently working has ever managed during a period of sustained shrinkage. Growth has been the norm since at least the GI Bill. We’ve had two-year blips, but never a 15-year downward trend.
Standing up for Community College
One highlight of our conversation wasn’t a prediction. It was one of our guests, Sophia.
She unmuted and shared her story: community college to Harvard, from a low-income community. And she pushed back on the idea that community colleges can’t serve multiple missions.
“I’m an exception right now. But I shouldn’t be.”
I told her the world needs more of her self-described “delusional optimism and tenacity.” I’ve been in touch with Sophia since our conversation, and excited to see the big impact she makes in our space.
So What?
Anyone who knows me knows I’m an optimist and don’t like to dwell in the problem-space too long. Here’s what I’m still thinking about.
The entry-level labor market is getting harder. Skills-based hiring isn’t the silver bullet. Social capital matters more than ever. And the postsecondary landscape
which has never delivered for all students, is going to be asked to do more… quickly.
For those of us in education, the job to be done is clear. Give every student career-connected experiences. Be honest about program quality and ROI. Build AI fluency as a core competency. And support ALL students, not just the ones going to four-year colleges.
That’s what we’re building with our partners at Willow Education.
Again, if you want to watch the full convo, you can find it here.
What am I missing? What do you disagree with? I’d love to hear from you.
Social Capital
Our next Purposeful Paths conversation is scheduled for March 12, 1pm EST with the leading thinker on Social Capital and Economic Mobility, Julia Freeland Fisher. I hope you’ll join us! You can register here.
Onward,
James
p.s. we launched our new website last month, would love to hear thoughts and feedback!




Tremendous. James - for the social media bans, what is your prediction on actual change to total screentime hours? I.e., could be that consumption declines, could be that it's unchanged.