At the beginning of this century, online education was a small but promising development. Though only about 6% of college students took any sort of online class at all, and only 1% of degrees were fully online, conditions seemed good for growth, and, if online education followed the pattern of other industries, good for controlling the seemingly unstoppable rise in tuition costs.
Two decades later, it’s clear that expansion was indeed significant, with yearly increases adding up to 8-fold growth in the number of students taking at least one online class, and 10-fold growth in the number of online degrees (though from a smaller base.) The number of students taking at least one online class in any given year crossed 50%
the year before COVID.
The hoped-for savings, though, have not accompanied this growth. Online education has lowered tuition, but only for some degrees and only at some schools. Why is that? The simplest answer is that it is possible to offer online education at significantly lower cost than in-person degrees, but it is not automatic, and it is not simple. Online instruction only reduces tuition if the college is willing to organize itself around reducing it.
Higher education is in a trilemma, a trio of tradeoffs like the famous “Good, fast, cheap. Pick Two.” Ours are between high quality, large enrollment, and affordability. Where there are savings to be gotten from digital platforms, they come from offering fully online degrees; just adding online classes to in-person degrees does not alter the core costs of educating a student
. Given this, schools wanting lower tuition can’t just digitize courses and call it a day.
Every college faces multiple claims on its resources – faculty salaries, student support, sports programs, research funds. Reducing tuition is just one of many options, and often not the one that seems most pressing. A fellow faculty member recently asked his students, in a class about college economics, what administrative headcount they would cut to lower tuition. Instead, they immediately argued for hiring more counseling staff.
The bulk of inexpensive online degrees are offered by state colleges and universities, which have the mission and resources needed to educate a broad swath of students online. At Penn State, an in-person course costs around $850
per credit hour for out-of-state students; online it is around $610
, a 30% discount. Arizona State (ASU) offers a 50% discount
for online study. At Purdue, it’s a 60% discount
. Maryland, 67%
With a few notable exceptions, state schools face little competition here. Community colleges typically lack the resources to launch online degrees, and many of their students lack the hardware or bandwidth needed to keep up with class. Colleges with fewer than a thousand students (nearly a third of the total) also typically lack the necessary income and IT staff. And private research universities and liberal arts colleges prefer the high tuition/high aid
model, subsidizing tuition for individual students, rather than lowering it overall.
Meanwhile, many for-profit colleges, previously the largest source of cheap online degrees, shrank or shut down after Obama-era regulations
led to implosions in enrollment (Phoenix, Strayer), restructuring (Grand Canyon, Kaplan), or outright closure (Argosy, Education Corporation of America.) Phoenix alone shrank by more than 400,000 students over the course of the last decade. While the Trump administration paused enforcement, and COVID provided the background for modest growth among the remaining for-profits, the Biden administration’s skepticism about for-profits
suggests these schools will never return to their previous scale.
The one other large-scale source of inexpensive degrees is a pair of private non-profits, Southern New Hampshire (SNHU) and Western Governors University (WGU), which enroll over a quarter million students between them. By contrast, average enrollment for all bachelor’s granting institutions is a little under 6ooo, and the median is a little over 2000, a consequence of the proliferation of small schools.
At SNHU, which has dramatically expanded its online enrollment in the last decade to become the country’s largest university, the cost per credit hour is $320
, compared to an average of $1492
for all private non-profits. Western Governors’ University (WGU), online from its founding in 1997 and the second-largest university, offers competency-based education
and charges a flat fee of between $3475 and $3635
per semester. Expensive colleges like USC
or Harvey Mudd
charge nearly $2000 a credit hour, meaning a single 4-credit course at those schools can cost more than a year
The savings don’t come from the technology
Reducing tuition requires an increase in productivity, and the brute force approach is having more students per class. A decade ago, as tens of thousands signed up for Sebastian Thrun and Peter Norvig’s massive open online course in AI
, these MOOCs seemed set to break the quality/scale/cost trilemma in just this way. The prospect of a nearly free education for millions of students at high (or high enough) quality was thrilling. MOOCs might have achieved their goals if lectures alone were a good way to learn things. Lectures alone are mostly not a good way to learn things
. Though the literature on the inutility of pure lecture
is vast, the key weakness is the lack of human connection.
MOOCs showed us that there are a number of people who, given the right materials, can create a remarkably good education on their own. They also showed us that that number is around 3%
. Outside this very small pool of autodidacts, humans need interaction to stick with a course. Even with the cost at $0, and the fees for certification set very low, for the vast majority of students ‘go at your own pace’ meant ‘stopped’.
The excitement about MOOCs assumed the problem the students faced was a lack of instructional materials, not a lack of motivation. But education exists in a social context. As the most comprehensive review of the effects of college on students
…the extent to which peers and student groups are seen as friendly and supportive and to which faculty are seen as approachable, helpful, and encouraging have the most important positive implications for how much students report learning during college.
Peer pressure and engaged faculty turn out to be powerful enough to help overcome even considerable degrees of procrastination, distraction, or laziness among students, motivators that do not appear with self-paced video lectures.
The secret is just cost-cutting, unfortunately
Given these limits to the savings from scale, the more significant (and organizationally complex) adaptation is lowering the cost of instruction itself.
Hiring lower-cost instructors is already standard practice. In many introductory courses, a well-compensated faculty member lectures, but less expensive teaching assistants run sections and grade papers. Contract faculty are paid less than tenured faculty; adjunct faculty, who teach a course at a time, are paid less than contract, and so on.
Online tools do nothing to change these economics, but they make it easier to recruit instructors nationally; a teaching job paying $2500 a class will be more appealing to someone in Las Cruces than Los Angeles, and schools with low-cost options typically staff many classes, and sometimes all, with adjuncts. (Remote work makes it possible to imagine ‘faculty towns’ where online instructors cluster for cheap rent, good weather, and reliable bandwidth, with no actual college nearby.)
The big surprise from the spread of online education is how unsurprising it’s all been. For the last ten years, much of the attention has been on MOOCs and bootcamps, on the assumption that cheaper degrees would be brought to scale either by high-status schools like MIT and Stanford, or by startups like General Assembly or Flatiron School.
Instead, online degrees mainly conferred by state schools, same as in-person. Even the country’s largest university, SNHU, is not a Silicon Valley disruptor, but an ordinarily accredited university. (The mass adoption of the last twenty years also went under-reported because the vocabulary-word press obsesses over the Ivies and rich liberal arts schools, which, for online degrees, were very much where the action wasn’t.)
Online education is not a broad revolution, transforming the economics of every institution it touches. Instead, it has become another format, like residential programs, part-time programs, or executive education. The genius of the American system (“system”) of higher education has been the frequent invention of new formats – land grant colleges, community colleges, regional campuses. Inexpensive online education is the most recent addition to that list.
The COVID pivot
COVID stress-tested everything. Previous convictions that online teaching was too complex evaporated, as schools converted their entire curricula in a few days. If COVID had somehow ended in the summer of 2020, our collective pivot would have remained little more than emergency remote teaching
. But a later lesson from the pandemic is that people get good at what they do more of. Since most schools offered online classes for more than a year, both faculty and students passed from emergency adaptation through familiarity to arrive, in many cases, at something like fluency.
As we’ve collectively grown more competent, it’s become clear that online education is a complex alternative to in-person education, not a simple substitute. A lot of our new practices are still rough around the edges, but most of us got better at running our classes online after doing it month in and month out, as emergency remote instruction gradually turned into plain old instruction. (There are still some concerns about online quality control, concerns that come perilously close to revealing an awful family secret about how we do in-person quality control.)
The Delta variant was also surprisingly important, destroying the conviction that we could predict the pandemic’s progress. In the summer of 2021, without a foreseeable end to the waves of infection, online instruction shifted from being a crisis response to an ongoing capability.
So is this the moment of massive cost savings? Probably not. The internet’s flexibility has provided better control over tradeoffs between cost, quality, and size, but has not abolished them. And the COVID pivot greased away institutional frictions that came from unfamiliarity, but it did not introduce any dramatic new savings.
Post-COVID, the institutions likeliest to expand online fastest are the nation’s thousands of community colleges, which already practice stringent cost control, giving them less flexibility on cost. (CalBright
, a free competency-based online community college in California, is the largest current experiment, though there are others.) Lack of demand is also a factor: when students or their families say they want an inexpensive education, what they often mean is that they want an expensive education at a subsidized price
, something not offered by inexpensive online degrees.
At the same time, the capaciousness of the U.S. system means that new kinds of schools rarely drive old kinds out of business. Public schools co-exist with private; sectarian with secular; specialist with generalist; research with teaching. And inexpensive with expensive.
Online enrollment has increased every year of this century, and looks likely to continue growing for some time. We don’t know when that growth will level off, but ‘not soon’ is a safe bet. The boring truth is that the spread of online education has been a game of inches, played over decades, a pattern that COVID may have briefly accelerated, but has not fundamentally changed.
The (Continual) Transformation of Higher Education
is a newsletter by Clay Shirky
. Publishes weekly, more or less. Feel free to forward. Subscribe here