What Is College For?

Powerful forces threaten to re-order higher education. Predictions of massive destruction are overstated, but the coming storm will force colleges to figure out what they want to do, writes Dan Currell.

June 28, 2013

My college years were spent on a hill in a small town. I was in the company of 3,000 other people – students, faculty, staff – and we were set apart.  The only thing on the agenda was to continue being Gustavus Adolphus College, whatever that meant.  I didn’t know who first set that agenda, and I don’t recall a lot of active reflection on what it meant.  What did it mean to be a residential, liberal arts college in the Swedish Lutheran tradition? We discussed that a little bit, but mostly we just did it.

Now I am a trustee.  A lot has changed, but the basic character of the place hasn’t. Whatever it meant to be Gustavus in 1990 – well, it still means that in 2013.

On the horizon, I can see a lot more reflection about what exactly it means to be Gustavus. Everyone can sense the powerful forces affecting colleges; some would say they threaten to destroy the four-year residential model altogether.  Some expect this to happen fast.

Perhaps higher education has adapted slowly to the Internet, especially compared to the private sector. But the practical benefit of being last in line is that you can learn from everyone ahead of you. (There’s a reason you didn’t want to go first when it came to giving oral presentations in fourth grade.)  The parts of our society most affected by the Internet have learned a few important lessons, and I think these lessons can help to set a vision of the future for colleges.

I believe the main lesson is this: the Internet didn’t kill off whole sectors of the economy – as it was predicted to – but it did force management to take the threat of rapid extinction very seriously.  If the “old economy” conversation was about how to gain more customers for your service, the “new economy” conversation is about why any customers should want this sort of service to begin with.  It’s a scarier conversation, and it forces companies to think harder about what they are for. 

I hope colleges never discuss why their customers need their services. These are communities, not big box stores. But the college version of that conversation is this: What are we for?  What’s the goal?  Since there are now innumerable other (and cheaper) ways to be educated, why are we doing this?

The colleges with a compelling answer to that question – where all 3,000 people know the answer – are going to be fine. At Gustavus, we are in the midst of a campus conversation on this topic.

In the mid-1990s the business world began to recognize a series of forces that were thought to threaten nearly every industry.  By 1999, the forces had names:  digitization (products and services become electronic); disintermediation (intermediaries disappear); and unbundling (products are disaggregated into discounted components).  Doom was declared for incumbent companies, and the Pets.com sock puppet roamed freely at Super Bowl halftime.

Latching onto the emerging zeitgeist, GE's CEO Jack Welch declared that his goal was to destroy his business before someone else did.  He called the initiative destroy-your-business.com.

In the years that followed, much effort went into studying how innovation works -- and how it often leads large companies to fail.  In 2001, a study of the market for mechanical excavators and disk drives showed that markets are typically disrupted by cheap, inferior technologies aimed at customers who can’t afford the “full package.”  When the cheap products get better, it’s curtains for the incumbents – they have been disrupted.  Clayton Christensen, the author of the work, became the corporate world’s most famous professor.  He also made a compelling case that incumbent companies were nearly powerless to resist the undertow of this kind of disruption – they were stuck in what he called an “innovator’s dilemma.”

The forces that began threatening businesses 20 years ago took a long time to get to universities, but now they have arrived.  In the last year, hundreds of massive open online courses (MOOCs) have rushed onto the scene, digitizing the teaching of world-class scholars and making their classes available to anyone for free.  This followed Christensen’s model of a cheap (or in this case, free) and admittedly inferior substitute undermining a very expensive one.  The prospect of disruption has unleashed a flood of anxiety and not a few prophecies of doom for expensive universities. The problem for universities looks much the same as it did for “bricks and mortar” businesses in 1999 – digitization would lead to disintermediation, which would lead to unbundling. That sequence could destroy the incumbents.

Of course, colleges and businesses are different, and most people who take higher education seriously want to keep it that way.  But educational services are increasingly bought and sold on an open market.  If we don’t want colleges to become mere creatures of the market, the first step is to better understand the market.

Viewed this way, the incumbent company experience since 1999 should be largely encouraging.  It turned out that it’s harder for a bunch of web programmers to learn retail than it is for an established retailer to hire web programmers.  The sock puppet lost.  Petco now has over 1,000 locations, and PetSmart owns the defunct Pets.com domain name. (Pets.com lost a staggering $300 million for its investors, though the sock puppet got a second life pitching auto loans.)

That said, the last decade has forced every company to focus relentlessly on its raison d’etre, on knowing exactly what it is good for and adjusting its operations accordingly.  That soul-searching is probably the most lasting effect of the digital revolution: managers are constantly forced back to the question, “What are we good for now?”

This will be the main effect for colleges, too, and it doesn’t come a moment too soon.  Many of us agree that college is important; we have no broad agreement on why.  Now that an education can cost more than a house, it seems like a particularly important question.  Why is college important?  What exactly makes it worth so much time and money? 

Online education may force many universities to admit that they are not really in the transformation business.  Is a 200-student lecture hall with a graduate student at the front the path to transformation?  In many cases, it’s barely education.  Colleges are better-positioned than most universities in this regard – but online education will still bring real pressure to demonstrate the distinct value of what a college can deliver.

Questions like this have now taken on a sharp edge, because there will be a heavy price to pay for getting the answers wrong.  As with other industries, an analysis of the answer starts with digitization – and moves through disintermediation to unbundling.


Digitization turns a physical product into a data set. Most entertainment, financial, and information products have been digitized, which has dramatically reduced the cost to copy, store and transmit them.  Most importantly, digitization can allow a product to rapidly evolve into something different – and better.  This rapid evolution is the real threat to incumbents.  And tantalizingly, a digitized product obliterates physical and proximal limitations: employees can live and work in the Bahamas even if their services are consumed in Seattle.

The prime incumbent horror story for digitization is Encyclopedia Britannica.  The obvious effect of digitization was to make encyclopedias lighter, easier to search and far cheaper to distribute.  These were all improvements to Britannica’s product.  In an increasingly knowledge-based economy, Britannica circa 1999 seemed to be in an enviable position. 

But of course we now know that the main effect of digitizing encyclopedias was to transform the thing itself.  Wikipedia hasn’t quite killed Britannica – but it may just be a matter of time.  Wikipedia was inferior and free a decade ago, but it is rapidly increasing in quality, scope and depth.  It fits the disruption model perfectly.

The frightening thing about the Britannica story is how hard it is to think of a happier ending.  Could the executives at Britannica possibly have imagined that their greatest risk was that tens of thousands of amateurs would spontaneously create a free substitute for their product?

The lesson to be learned from our first experiences of digitization is that digitization alone changes very little – but when it allows a product to rapidly evolve into something else, the effects can be violent and unpredictable.  Digitization did not change the pet products business: dogs eat the same food they did in 1998.  Digitization, however, changed what an encyclopedia is, and that made all the difference.

That said, very few businesses were hurt as badly as Britannica.  For example, money has been digitized – so what happened to banks?  In 1999, there was no reason to think that banks should survive in their bricks-and-mortar form for more than about five years: depositors and borrowers could transact everything far more efficiently online and pocket the savings.  In 1999, Britannica seemed to be in a better position than Wells Fargo.

How wrong that was.  Retail banking products have been digitized but they are otherwise largely unchanged.  Because we haven’t yet figured out how to make virtual money do anything different than its predecessor, retail banking today would be entirely recognizable to George Bailey of It’s a Wonderful Life.  Money has been digitized, but it hasn’t been transformed.

And what about the virtual company – employees doing their work from notebook computers on the beach?  Digital products make remote work possible, but the financial sector – whose products now have no mass and no physical location - is still heavily co-located.  Bankers may vacation in the Bahamas, but they still work shoulder-to-shoulder in New York and London.  Perhaps the only sector more heavily co-located than banking, ironically, is technology: there are high places here and there, but the only true altar is in Silicon Valley, where code is written by software engineers sitting in adjacent cubicles.

In hindsight, it is no great surprise that universities have been affected so little by digitization.  Some elements of a university (lectures, books, lab activities) can be digitized, but as of yet their core product hasn’t changed.  If MOOCs are the first step toward transforming what a class is – then yes, we are at a crossroads in higher education.  If, however, they are simply a way of getting broader distribution for an existing form of teaching, they won’t destroy colleges.  In higher education as in other sectors, the litmus test is whether the product itself will change. 

This brings us back to some key questions: What is the product really, and why does anyone want it?  If cheap, simple substitutes are supplanting some components of a college education, what important parts, if any, are left?  And who is responsible for assembling the product and ensuring its quality?


Remember travel agents? They were the ultimate intermediaries. The Internet has nearly killed them by connecting consumers directly to airlines, hotels and cruise operators. That’s disintermediation, and it has been happening in every sector since 1999.  It is one of the forces that makes corporate strategists most nervous.

It was easy to foresee the death of the travel agent in 1999. But it was equally easy to foresee the death of real estate agents, publishing houses (readers buy direct from authors), music labels (same rationale) and a host of other intermediaries who are still alive and well in 2013. What happened? Why did so many intermediaries survive?

For the most part, industries still have intermediaries – just different ones. Consider the music business. Tower Records died pretty quickly – one intermediary down, score one for the consumer. The chain from production to consumption was getting shorter, and pretty soon consumers would buy their music direct from artists. 

Right? Well – in principle, yes.  In practice, rarely.

As it turns out, music labels (the intermediary everyone loves to hate) are still alive, and iTunes stepped in right where Tower Records used to be. There are just as many links in the music industry’s distribution chain now as before. The price of an album hasn’t dropped much, even though the cost of distribution is almost gone. As before, most artists are barely paid while the music industry generates billions in revenue. The most lucrative profits go to the intermediary with the best tollbooth. Right now, that’s Apple, and the tollbooth is iTunes.

But surely this reshuffling has had some effect on the product itself or the way we consume it?  Again, so little has changed. In “The Entertainer,” Billy Joel wrote “If you’re gonna make a hit, you gotta make it fit, so they cut it down to 3:05” – a reference to the power of radio stations who didn’t like to play long tracks. In the era of satellite radio, iTunes and Pandora, radio stations no longer rule the roost. So -- has the product been transformed?  According to thebillboardexperiment.com, the average length of a pop song has increased by all of fifteen seconds since the 1980s, and neither the sound nor the economics of the industry are all that different.

As with digitization, the question for disintermediation isn’t whether it will happen, but whether it changes the product. Travel agents weren’t driven out of the basic flights-and-hotels business just because they increased the price of a ticket. The problem was that Orbitz and Travelocity could do the job better at a lower cost. 

So it turned out that disintermediation was a misleading term, because in few cases were the intermediaries really eliminated. In most cases they gave way to new ones who did a better job in some way, but this didn’t make the transitions less painful. Most travel agents had likely never wondered exactly what value they created, or why their customers chose to use them, until it was too late.

Who are the intermediaries in higher education? Reed-Elsevier, Pearson and other content providers are obvious examples, and while they’ve altered and added to their strategies and offerings, for now their positions seem secure. But the college itself is an intermediary, too. Professors provide services that colleges and universities bundle together and sell. To be a little provocative, consider the music business again: the content creators are rarely paid much, even though there is a lot of money sloshing around the sector. The excess money goes to the intermediary. In this case, the money pays for counseling, career services, athletic facilities, housing and so on.

So, the economic form of “What is college for?” is this: is the college creating enough value to justify its position as an intermediary between professors and students?   

The question of whether colleges will be disintermediated likely turns on whether their service offerings can be unbundled. Unbundling was the third big force strategists were thinking about in 1999, and it has arguably been the most powerful one in the decade since.


Many products are made up of separate parts that consumers must buy together even though they won’t likely use them all – they are bundled.  A cable TV package, a Disney World pass, or a subscription to the Chicago Tribune – they are all bundles.  Bundled products are always more profitable than their a la carte siblings because they force a larger purchase and improve the economics of creating and selling the product.  They also always involve some amount of cross-subsidy from one purchaser to another: those who only use a few parts in the bundle subsidize those who use everything, since everyone pays the same price.

In 1999, we saw that the Internet might force the unbundling of products and services.  The most valuable parts of a bundle could be sold a la carte, which would be cheaper and more tailored for consumers who used only some pieces of the bundle.  Because those were the consumers who cross-subsidized the rest, this seemingly innocent development could undermine well-established business models.

Indeed, unbundling was lethal in many sectors.  Many newspapers sat on an economic foundation of classified ads: consumers made a small co-payment to get their comics and sports, and the real revenues came from commercial and citizen advertisers.  Commercial advertisers had other options – radio, magazines, billboards.  But private citizens had only one way to sell the lawn mower or get rid of an unexpected litter of puppies.  Classified ad revenues were very important, but classified ads were a piece of the bundle that not everyone used.

What newspapers didn’t know in 1999 was that classifieds aren’t very effective. Craigslist and eBay made this clear – they were a far superior way of getting the job done, and they could operate even better outside of the newspaper bundle.  Those two websites alone took the legs out from under the newspaper industry by stripping a key element out of its bundle. With classifieds stripped out of the bundle, the newspaper’s economic model doesn’t work. 

Unbundling may be the greatest threat to colleges, and a far greater threat to universities. At first, the college bundle just brought together several different professors to offer an integrated degree. More recently, we have added athletics, student services, career services, disability support, counseling, housing, and a host of other elements into the bundle.  Everyone pays for the whole thing, no matter how much of it they use. 

If something is cheap, it’s easiest to buy the whole bundle even if you pay for many elements that you never use. At higher price points that becomes less sensible. If the sticker price for a private college education is now about $200,000, that will buy a lot of private tutoring, lab time, and other a la carte educational services over the course of four years.  All that’s left is for the student to separately validate her achievements and the resulting competencies.

If this sounds far-fetched, consider that this is a major movement in high school education.  We know it as homeschooling, and it is no accident that the trend has exploded since the advent of the Internet.  As we now know, “homeschooling” is often a misnomer.  It is really the a la carte construction of a secondary education by families sharing resources and working in cooperative networks. It’s unbundled education.

We haven’t seen unbundled education at the college level yet.  This is surely in part because of the social status conferred by a traditional college degree.  But it is also because colleges have kept education and evaluation tightly bundled together.  The professor teaches and evaluates progress; the college offers courses and confers a degree.  As it now stands, the only way to get a degree is to actually attend a college of some sort.

It won’t necessarily stay this way.  There is no reason why education and evaluation will necessarily stay bundled together, and one can already see movement in the direction of the two splitting apart.  Right now it comes up as a question of compatibility: what transfer credits will a degree-granting institution accept?  A transfer student may have taken his courses on campus, or he may have taken them online.  If it was the latter, at least in some cases, it’s possible that he did little more than self-study followed by an exam or paper.  This comes very close to separating education from evaluation.  The next natural step is to accept credits from standalone online courses, at which point the degree becomes partly unbundled.

It might seem easy for universities to clamp down on this, but right now most colleges are in no position to be choosy when accepting transfer students, particularly if they can pay much or all of the sticker price.

Of course, a more concerning possibility is that diplomas won’t matter that much in ten years.  Some large and reputable employers -- particularly in the tech sector -- are now quite willing to hire people with demonstrable skills whether or not they have a college degree.  To take two prominent examples, neither Microsoft nor Facebook is in much of a position to insist that their employees have all finished, college since neither of their CEOs did.  What they need are demonstrable technical skills, many of which are now separately assessed through third-party testing anyhow.  For example, see www.brainbench.com.  In those cases, a college diploma is already a “nice to have.”

Granted, our culture is still very attached to the idea of a college diploma.  But the idea that it might not be necessary, even (or perhaps especially) at the elite end of the labor market, is catching on.  Apparently there’s something about a $200,000 price tag that will make you wonder how much you really want the product.  And there’s also the investment of time and energy.  Consider the Thiel Fellowship:  fellows are given $100,000 to skip college and pursue their entrepreneurial ambitions right away.  The presumption behind the Thiel Fellowship is that college offers little more than some learning that the brightest students will pick up along the way as they build the next big thing.  This reduces college to nothing more than the content-learning element; it ignores the maturing process, critical thinking, collective learning, ethical reflection and a host of other things more important than the principles of organic chemistry.

A less radical possibility than ditching college altogether is the unbundling of evaluation from education.  This would allow students to spend their college budgets as they see fit – online courses, live tutorials, study abroad and internship experiences, seminar classes or whatever – and test separately for the purpose of showing progress.

This kind of unbundling may seem a long way off, but at least one economic factor points in its direction.  Bundles will tend to come apart when the cross-subsidy between different buyers of the same package becomes too great.  In American higher education, we may soon reach that point.  At most private colleges and universities, the gap between “full pay” students, median-tuition students, and heavily aid-dependent students is enormous. 

Why would this cause the bundle to rupture?  It’s one thing to pay $5,000 for an airline ticket from New York to Los Angeles, as business travelers sometimes do. That kind of premium pricing cross-subsidizes those on the same flight who have paid $300, and it’s what makes airlines viable (at least in the years when they are).  But at some point on the pricing curve -- $15,000, say -- the New York-based executive hires a NetJet to fly her direct to Burbank.  The pricing isn’t very different and the service is more tailored.

What Next?

What became of GE’s destroy-your-business.com?  Jack Welch said in a 2004 interview that in hindsight it was unnecessary. GE's existing businesses didn’t need a separate initiative to prod them along – they focused relentlessly on creating real value for customers, and they incorporated new technologies as they emerged.  In fact, across the economy, nearly all incumbent businesses discovered that they offered far more value than what could be easily digitized.  It had seemed for a moment in 1999 that the internet placed them in mortal peril, but there was just too much connective tissue holding their businesses together -- creative work, advertising, accounting, quality control, audit, personnel management – for the Internet alone to kill them.  Incumbent companies that focused relentlessly on their raison d’etre came out of the digital revolution stronger than they went in.

If a college’s true product is a transformed student, then the main effect of the next decade should be to redouble every school’s commitment to that cause. The explicit goal of residential liberal arts colleges will again be to increase what a student knows and change who she is.  If this is true, then the conversations left to be had are about the transformative mission of the school. What exactly is it? Deciding on a clear and important set of goals will not be easy, but colleges cannot afford to kick that can down the road.  We each need to figure out what our college is for.


Dan Currell is executive director with the Legal, Risk and Compliance Practice at the Corporate Executive Board.


Be the first to know.
Get our free daily newsletter.


Back to Top