Once-unthinkable disintermeditiation: higher ed

Ever since MIT introduced OpenCourseware in 2001, higher ed faculty have been watching with curiosity to see how giving away free content would impact the school’s brand and corse business model. Last December, my alma mater announced MITx, where students could earn a “credential” from online courses, and then on May 16, it named the leader of MITx, Provost L. Rafael Reif as the 17th president of MIT.

MIT is far from the only university to embrace and promote this change. In 2007, Yale launched Open Yale Courses and Stanford helped launch iTunes U in 2005, and recently began offering courses with 10,000s of enrolled students. However, the biggest visibility for such efforts came on May 2, when Harvard joined MIT to announce edX, which will combined MITx and Harvardx to offer certificates but not course credit.

Many of us business profs have been watching with both curiosity and trepidation. On the one hand, there is no business model yet for the (well-endowed) colleges to give away content, even at low cost. On the other hand, we’ve seen this movie before — whether with CD Now and Wherehouse (or Sam Goody or Musicland), Amazon and Borders or Amazon and Circuit City.

In particular, the talk by these pioneers that these efforts are complementary to traditional education is just talk. As the outgoing MIT president said on May 2:

“Online education is not an enemy of residential education, but rather a profoundly liberating and inspiring ally,” MIT President Susan Hockfield said in announcing the project, according to a Harvard Gazette report. She explained that "you can choose to view this era as one of threatening change and unsettling volatility, or you can see it as a moment charged with the most exciting possibilities for education leaders in our lifetimes."

I recall how when I started my full-time computer career in 1980, the PC was a toy and any serious work was done on a mainframe. Now everyone carries more computing power in their pocket than the mainframes of 30 years ago, but most of our computing power and storage capacity is provided cheaply by a few on-demand clouds.

As we business profs are painfully aware, online higher education shows all the signs of becoming a classic Clayton Christensen “disruptive innovation.” It will start out lower quality, but eventually it will get good enough to replace the traditional product and destroy most of the market due to its larger scale and lower cost. (NB: When will this effect come to secondary education?)

An op-ed in this morning’s Wall Street Journal by John Chubb and Terry Moe — two Hoover Institution (i.e. Stanford) fellows and education reform authors — was a little more blunt:

Over the long term, online technology promises historic improvements in the quality of and access to higher education. The fact is, students do not need to be on campus at Harvard or MIT to experience some of the key benefits of an elite education. Moreover, colleges and universities, whatever their status, do not need to put a professor in every classroom. One Nobel laureate can literally teach a million students, and for a very reasonable tuition price. Online education will lead to the substitution of technology (which is cheap) for labor (which is expensive)--as has happened in every other industry--making schools much more productive.
…
For now, policy makers, educators and entrepreneurs alike need to recognize that this is a revolution, but also a complicated process that must unfold over time before its benefits are realized. The MITs and Harvards still don't really know what they are doing, but that is normal at this early stage of massive change. …

But like countless industries before it, higher education will be transformed by technology--and for the better. Elite players and upstarts, not-for-profits and for-profits, will compete for students, government funds and investment in pursuit of the future blend of service that works for their respective institutions and for the students each aims to serve.

Even this is a Panglossian (or, more likely, intentionally sugar-coated) view of the world. Retired USC professor Lloyd Armstrong wrote:

One could imagine that a few years from now, MITx would have the equivalent of entire degree programs on line. Students successfully completing the demanding sequence would not, of course, get an MIT degree, but rather a MITx "super-certificate" certifying their success in the entire degree program. They would not be equivalent to MIT grads, but because of the rigor of MITx courses, they would likely be better prepared than the grads of a large fraction of accredited schools. I believe that the MIT reputation is such that employers would see this MITx super-certificate as providing a meaningful description of the skill set of the recipient.

OK, maybe a BSEE from Berkeley (or even Cal Poly) is worth more than an online certificate from MITx, but how would the certificate compare to a BS from CSU Nowhere or ITT Tech?

At the Chronicle, the trade journal of higher ed, Kevin Carey is a little more blunt:

Harvardx won’t compete with Harvard University in the business of running admissions tournaments for aspiring members of the ruling class or assembling great minds in a single place to conduct world-class scholarship and reseach. Harvardx will be competing with everyone who isn’t Harvard University, or its general equivalent. Expensive, newly-arrived, brand-deficient for-profit online colleges probably have the most to fear, followed by over-priced private non-profits and then lower-quality non-selective public institutions.

This news comes at a time when California politicians continue to mismanage the state economy and state budget, spending billions on pet projects we can’t afford while defunding public higher education. The cuts are real, not hypothetical, and they are weakening all three levels of higher education — UC, CSU and JCs. Since college expenses are mostly people, as with the past decade of cuts, the latest cuts mean hiring freezes, layoffs and likely less students being served (or increased prices, or both). The generate incremental revenue, the UC graduate and professional programs — MBA, JDs, MDs, MS engineering — have already raised their prices to public school levels, and CSUs are heading in that direction.

It‘s clear where this trend is going. As with online retailing, scale economies reduce costs, increase scope and eliminate natural geographic boundaries of competition. All colleges will need to increase labor productivity (both efficiency and effectiveness) to survive.

Writing on TechCrunch, Gregory Ferenstein attacks the problem of effectiveness — the unspoken problem that rather than mastering material, students everywhere cram information into short-term memory just long enough to take a test:

[S]aying that EdX is "the biggest change in education since the invention of the printing press" ignores the fact that lectures are often the least educational aspect of college: after four years of instruction, research shows that many students haven't mastered basic reasoning or communication skills. Students forget most of what they hear in lecture and then only recall 40% of the tested material two years later. Lectures do little for students actually enrolled in the school, let alone the millions of online users who will study part-time, without a supportive community or frequent feedback from a professor.

The answer is to change the actual delivery approach in residential education, i.e. use online methods to eliminate the lecture and use classroom time for feedback and personalized instruction:

[l]ast week, two Stanford professors made a courageous proposal to ditch lectures in the medical school. "For most of the 20th century, lectures provided an efficient way to transfer knowledge, But in an era with a perfect video-delivery platform -- one that serves up billions of YouTube views and millions of TED Talks on such things as technology, entertainment, and design -- why would anyone waste precious class time on a lecture?," write Associate Medical School dean, Charles Prober and business professor, Chip Heath, in the New England Journal of Medicine. Instead, they call for an embrace of the "flipped" classroom, where students review Khan Academy's YouTube lectures at home and solve problems alongside professors in the classroom. …

Prober and Heath point to a recent one-week study that compared the outcomes of two classes, a control class that received a lecture from a Nobel Prize-winning physicist and an experimental section where students worked with graduate assistants to solve physics problems. Test scores for the experimental group (non-lecture) was nearly double that of the control section (41% to 74%).

"Students are being taught roughly the same way they were taught when the Wright brothers were tinkering at Kitty Hawk," they explain. After a revolution, an organization should bear little resemblance to its former self. Harvard and MIT have merely placed the 20th century education model online. Stanford, on the other hand, is completely doing away with the old model of the "sage on the stage" and embracing a learning environment that mirrors life forever connected to the world's information.

My new employer is pursuing exactly this approach of flipping the classroom to improve effectiveness. Being small, nimble and non-bureaucratic, I think we can get there quicker than most. But, like every other graduate professional program, we will have to demonstrate the value proposition and career benefits of our residential program.

Still, it’s clear to see where this is heading. Higher education has long resisted improvements in labor productivity, but online education efforts such as edX and OCW provide the path forward. Improved efficiency means less labor used for the same quantity of goods delivered. In a “flipped” classroom, much of the work today done by professors will be done by grad student TAs, people with master’s degrees, or otherwise less expensive or higher productivity workers.

The one missing piece was grading: the labor-intensive, non-automatable part of the education equation. Stanford’s CS faculty have been working to create an artificial intelligence approach to grade more arbitrary forms of homework (including someday essays). This would provide order of magnitude improvements scale, allowing one faculty member to teach not 500 but 50,000 or 100,000 students, without having (as in a traditional lecture-recitation approach) provide one section leader for every 30 students. (MIT says their electronics class of 120,000 students has “the number of TAs you would expect for a class with 100 or 200 people,” by relying on discussion boards as a substitute for class discussion.)

As a result, employment numbers for PhD-holding faculty will fall by at least a third (probably more like 50%) by the middle of this century: college professors will not quite go the way of cold type typesetters, keypunch operators or record store clerks, but our jobs are not as safe as manicurists or lifeguards. Like other elite research universities, MIT will continue as a contract research lab for industry, government and other sponsors, hiring PhDs and grad students to deliver that research (and less frequently, teach classes).