College enrollments are up, we are told. College enrollments are declining, so we are also told. Both may be true. The situation isn’t simple. Let’s look at the numbers, consider the causes, ponder how much artificial intelligence (AI) has to do with this, and reckon with the fate of small liberal arts colleges.

By the Numbers

Enrollment statistics make for dry reading, but it is good to have some facts at hand before we venture into explanations. In this case, the official facts are like estimates of the number of earthworms in a garden. The answer varies: lots; not enough; before or after we go fishing?

The numbers vary by wide margins depending on the sources. Overall enrollment in “post-secondary institutions” has increased. By some counts, it now stands at 19.4 million, of whom 11.5 million are full-time. College enrollments peaked in 2010 and have fallen about 20 percent since then. About 40 percent of undergraduate students are enrolled in community colleges (i.e., 7.76 million). More than half of those community college students are part-time. 

But every one of these numbers floats in a sea of competing statistics. Some analysts claim there were 8.6 million community college students in 2022-2023; other analysts put the fall 2024 number at 12.4 million, of whom 72.2 percent are part-time. The U.S. Department of Education (ED) puts the figure at 4.7 million—full-time and part-time combined—for fall 2003.

Those are discrepancies beyond the reach of any plausible reconciliation. My guess is that the ED numbers come closest to reality, so I will stick with those for the other institutions. 

Everyone agrees that community college enrollments are rising. Some say by about three percent per year; some by 5.8 percent, roughly three percent year over year. But how can anyone really tell? 

Post-secondary education, of course, includes elite colleges and universities—the Ivy League institutions, a handful of private universities, and the top-drawer private colleges. Fewer than one percent of undergraduate college students are enrolled in the most elite colleges. The category, stretched to its broadest definition, enrolls about five percent of undergraduate college students. And the total enrollment at these colleges and universities barely changes year to year. They may have to dip deeper into their application pools, but they are in no danger of having empty seats or dorms.

Four-year public universities enrolled 9.3 million students in Fall 2023. Four-year private universities enrolled 5.2 million students in Fall 2023. Four-year for-profit institutions enrolled 0.8 million students in Fall 2023. This is where the trends get interesting, and the lump sum statistics get seriously misleading.

Large public universities appear to be reaping a significant increase in enrollments. The estimates range from 1.4 to 1.9 percent as of fall 2025. Some prestigious private universities are seeing comparable numbers. But smaller public colleges and especially private colleges are facing declining enrollments. Private colleges saw a 1.6 percent decrease in enrollment. 

Those who like to extrapolate a trend from a few data points can view these numbers as a forecast of disaster, and those who have a wait-and-see disposition can shrug them off. In what follows, I take these numbers not as omens but as fairly reliable indicators of where we are and where we are going. 

Here’s Why

These numbers are snapshots of a train that is roaring past. Everyone in higher education knows full well that the “demographic cliff” has arrived. The number of young people graduating from high school and ready to attend college has fallen sharply and will continue to decline for at least the next two decades. Perhaps, given the reluctance of Generation Z to get married and have children, the demographic sweepstakes will make college enrollment in 2044 and beyond an even worse proposition. 

Added to the demographic slump are several other factors that “everyone”—or if not everyone, all of the people who are paying attention—knows. What follows is mostly a list of those factors, each of which would require an essay in its own right to make full sense of the situation. But here I aim to fly over the forest and merely note the important trees. The value of doing that is that it offers a perspective on how many different developments are contributing to higher education’s lopsided decline.

The factors are:

• AI is changing the calculation of whether going to college is a good investment if the goal is to land a well-paying job in a secure field.
• AI is also raising new questions about what kind of post-secondary education is most adaptable to a highly uncertain marketplace for college graduates. 
• A majority of adult Americans now look upon college with a high degree of skepticism. The skepticism is most pronounced among conservatives—74 percent doubt the value of college— but is abundant among liberals as well—51 percent. Overall, 63 percent of Americans polled by NBC News in Fall 2025 said, “a four-year college degree is not worth the cost.”
• Decades of spiraling college costs and rapid accumulation of student debt have created the highly visible phenomenon of unemployed and underemployed college graduates. The current underemployment rate for college graduates is 42.5 percent.
• The 2025 One Big Beautiful Bill Act included a provision that makes federal funding for universities dependent on the median income of their graduates compared to individuals who finished high school but did not enroll in college. The provision draws public attention to “return on investment” (ROI). This means that those who doubt the value of a college degree by institution or by field of study will have facts at hand, not just impressions.
• Opportunities for young people who skip college to get started in well-paying careers have never been better. The stigma of not having a college degree has mostly evaporated.
• The required college degree has disappeared from many jobs that previously required candidates to have it.

Each of these points could be elaborated, but they are well enough known to let stand, except for their uneven effect on different types of colleges. Several other factors are less well known but worth citing:

• New obstacles in the path for foreign students to enroll in American colleges and universities. Many colleges have grown to rely on international students, a large portion of whom pay full tuition, unlike their domestic counterparts.
• An accreditation system that has locked in leftist ideology in many colleges and universities.
• Revisions to the rules governing federal student loans that make it more difficult for under-performing students to obtain loans or to stay enrolled.

And other factors complicate everything:

• The steep decline in the academic preparation of college-bound students.
• The need for colleges and universities to lower admission standards in light of the decline in preparation and ability among the pool of students who seek admission.
• The downward spiral as lower admission standards result in decreased quality of courses and graduates who are less reliably qualified for professional employment.
• The admission of a significant number of illegal immigrants, whose education is subsidized by the tuition of other students
• The psychological fragility of a large percentage of students.
• The settled disposition of many universities on one side of the culture war, and their hostility to traditional American values.

A really sophisticated form of research might be able to assign weights to these different factors, but it will suffice to say that while all of them play some part, concern about getting a good job and the start on a good career is far and away the most important factor when high school students are weighing whether to attend college and, if so, which college to attend.

Americans are perfectly willing to put up with high expenses and numerous annoyances if that is the price they must pay for an education that promises well-paid work in a secure occupation. Year after year, for decades on end, more than 80 percent of college freshmen assert that “getting a good job” is their primary reason to go to college.

AI threatens that. 

How AI Enters the Picture

AI is a threat, not a verdict. Maybe AI won’t get in the way of launching a successful career at all; maybe higher education will adapt. But people worry that it will destroy employment opportunities and that the long-assumed advantages of the college degree will evaporate.

The advent of AI means that certain fields of study now look like very poor bets. Who will go to college to “learn how to code,” when AI has already taken over that function? Some fields look more secure than others. Nursing, for example, is not likely to be automated into obsolescence. Accounting? Who knows?

A 17-year-old faced with this situation might well hedge his bets by enrolling in an institution that offers many possible options. This would allow him to adjust his studies along the way in light of shifting opportunities for graduates. That kind of prudence is a large factor in the growing popularity of large state universities that offer an abundance of options.

That said, those universities are also aware of the changing landscape and are busy paring down the options. Typically, the programs they are cutting are in the liberal arts, which may not reduce the flexibility the students desire. If French literature is off the table, c’est la vie. I can still major in engineering or physical therapy.

Uncertainty about the marketplace didn’t begin with AI, but AI has increased uncertainty dramatically. No one knows where this technology is going. What jobs will disappear? Which occupations will survive but be radically altered? What new kinds of work will emerge? 

Such transformations have happened before, but which of them is the right reference point? Is the advent of AI comparable to Native Americans getting horses from the Spanish? Those horses opened up whole new ways of hunting and made previously uninhabitable portions of North America into valued hunting grounds. They rearranged the tribal landscape. Or is the arrival of AI more like the arrival of the transistor radio that made it possible to bring recorded music to the beach? Is this a Gutenberg moment or the Nicolas-Jacques Conté—inventor of the graphite pencil in 1795—moment? Has OpenAI done the equivalent of domesticating wolves into dogs? Or the equivalent of discovering how to make opium out of poppies? 

Possibly, it is a technological blip of no real consequence at all, but few of today’s college-eligible students believe this. They have already incorporated ChatGPT and Gemini into their lives. They expect their choice of college to be at least as savvy as themselves.

Because we can’t answer such questions, we have to look for ways to mitigate the risks. One way is to skip college and concentrate on learning skills that will be useful no matter what AI does to the economy. This doesn’t necessarily mean frontier survivalist know-how, such as how to hunt and dress a deer, but it might mean knowing how to build a house, do plumbing, or teach children how to read.

Increasing numbers of young people are pursuing such options, but a majority are still willing to take AI in stride and go off to college. 

How We All Lose

This hedging comes at the expense of small colleges. The small colleges often do a few things extraordinarily well, but they aren’t known for their flexibility. As a result, they are fighting for their lives. 

Douglas Belkin in the Wall Street Journal (WSJ) recently reported that “442 private nonprofit colleges—out of 1,700 nationwide” are “at significant risk of closing or merging in the next decade.” Some analysts, such as Clayton Christensen, are far more pessimistic. Christensen, who died in 2020, forecast that as many as 50 percent of all colleges will close in the next decade.

I take that to be highly unlikely. There have been similar doomsday declarations in the past, but higher education as a whole has soldiered past the dangers of unaffordability, MOOCs, and public distrust. Why not this time too?

But colleges are indeed closing. Some 80 colleges have either closed or “merged” since March 2020—and “merged” means essentially the same thing as closed. More will close soon. Belkin’s WSJ article focuses on St. Michael’s College in Vermont, a 120-year-old institution that has not closed but is in the category of those at high risk. It has been running back-to-back deficits; laid off a third of its faculty; is rated by Moody’s as junk bond status; and its enrollment has declined from 2,100 in 2014 to about 1,120 in 2024. 

It is hard to see a way back for St. Michael’s, though I am among those who would dearly like to see it thrive again. That’s because small liberal arts colleges like St. Michael’s offer a form of education that is intrinsically valuable and almost impossible to replace. Being educated in a small face-to-face community where the students know their teachers and the teachers know their students is ideal. A small college can ensure that students build on a shared intellectual community, where they read many of the same books and experience the same intellectual traditions. A small college is also more of a moral community than any university can be. Universities can provide these things too, but only in the form of subgroups, such as a fraternity, an honors program, or a team. Nothing against those options, but the small college is something else, and much will be lost if America loses these educational oases.

Vermont was once home to a fair number of them. Green Mountain College, founded in 1834, gave up the ghost in 2019. A whiskey entrepreneur bought the campus at auction for $4.5 million and is offering it for free to anyone who will commit to spending $200 million to repair it and bring it back to life as a college. 

That seems like an outlandish proposition. But I hope someone tries.

Follow the National Association of Scholars on X.