Thursday, June 19, 2014

Clay Squared

This post is written in the wake of Jill Lepore's The Disruption Machine and its critique of the concept disruptive innovation, attributable to Clayton Christensen.  He is one of the Clays in my title.  The other is Clay Shirky, who has also written about disruption and its impact, particularly about the decline of newspapers and the parallels with Higher Education.  My piece is mainly for those of us in Higher Ed.  How should we be thinking about innovation and disruption in light of Lepore's essay?  Lepore, more quietly than her critique of Christensen, makes an argument to affirm an incremental change approach.  Is she right about that?  Or are circumstances so dire that a more drastic approach is called for?

In this piece I'm going to be content with presenting my views on these questions.  I'm not going to argue that these views give the right answers.  They are simply my views, to which each reader can contrast her own. 

I need to give a disclaimer first.  While I've downloaded The Innovator's Dilemma for the Kindle, I've not read it.  What I know about it has been garnered from what friends and colleagues have written.  I don't recall reading anything else by Christensen either.  So I'm on thin ice in writing about his ideas.  Likewise, I haven't read Here Comes Everybody.  But in the case of Shirky I have read several of the essays on his blog.  Newspapers and Thinking the Unthinkable was particularly influential on me. 

There is something that needs to compensate for this weakness in background reading, especially if anyone else is to take my views seriously.   So here it is.  I used to do research in industrial organization and was a reasonably proficient applied theorist, writing several papers on entry deterrence and being a regular at the IO workshop in the Economics Department during the late 1980s and early 1990s.  It is with this background that I come at Christensen's ideas.  Most of my friends and colleagues in Higher Ed lack this perspective.  Yet I've also been a faculty member for many years and I was an administrator in the learning technology arena for upward of 14 years.  From this, I know that Higher Ed is not just another vertical and that we should be wary of looking at Higher Ed purely through a business lens.  It is this dual perspective where I believe the value is in my views.  Further, I believe the combination of these perspectives is superadditive (the whole is greater than the sum of the parts).  My title makes it seem that the effects are multiplicative.  I'm not claiming that.  Given superadditivity, the title is poetic license, nothing more.

Let's begin with the story behind the Innovator's Dilemma.  There is an industry that produces a valued product for its customers and in so doing delivers high profits for the incumbent firms.  A marginal fringe market emerges where lower cost and inferior substitutes are offered, inferior in the sense that the quality is shabby.  Over time producers in the fringe market move down the learning curve and their product becomes even lower cost while the quality rises.  Some of the customers in the main market try what the fringe market has to offer and are pleasantly surprised.  Price is much lower than in the main market and quality is not bad.  So they switch their suppliers.  This is the beginning of a chain reaction.  Before long the fringe market has become the new main market.  The old main market has become a dinosaur.

To this story let's add three additional factors that buttress it.  The first is Moore's Law, which for the sake of this story I will treat as an exogenous truth.  Both markets will see either quality improvement or cost reduction as a consequence of Moore's Law, but the benefits might not be equal across the them.  To the extent that Moore's Law favors the fringe market, obviously that supports the Innovator's Dilemma story. 

The second is that people my age (late 50s) or older or even those who are ten to fifteen years younger will have living memory that in the late 1970s and early 1980s IBM was the leader in computing.  Likewise, at that time we'd use the word Xerox when referring to photocopying, even if the copying machine was from a different vendor.  We also remember the purchase of our first personal computer, a glorified word processor and otherwise toy for adults, perhaps with a dot matrix printer to accompany it.  We know the story of both the PC revolution and the Internet revolution that followed it.  We know that with any prescience at all IBM would not have outsourced the writing of the operating system for its version of personal computers and that had it done so Microsoft would never had gotten off the ground.  Likewise, with Xerox the developer of the original mouse, it should have been them to come out with the early versions of the Macintosh computers, not Apple.

Third, none of us like to seem stupid.  For those of us who consider ourselves pretty smart, the vast majority of those who are in Higher Ed are in this category, we really don't want to seem stupid.  This makes us exceedingly wary about potential threats and far less confident with current success.  So it makes us receptive to the message in The Innovator's Dilemma.  You should be in the process of cannibalizing your current business line even if it is highly profitable now, because it won't be in the not too distant future.  And if not cannibalizing then at a minimum you should hedge your bets by investing fairly heavily in the fringe business as you continue to run the current business.  Ignoring the fringe business entirely you do at your own peril.

To this sense of insecurity economists can add a note of mirth, to wit:

“The only function of economic forecasting is to make astrology look respectable”
Found here.

In other words, we (economists, all human beings) are not very good at all in distinguishing a possible but highly unlikely threat from a probable one.  If there are many different potential threats as new fringe markets and we are unable to judge which are the most probable and which, if any, are actually likely, we may very well ignore them all as in most cases investing in them will prove unwise after the fact.  Christensen's approach suggests there is only one threat to focus on and the threat is real enough.  In his approach uncertainty is minimal.  Lepore criticizes him on this and rightly so.

But Christensen's story was never meant to tell all about the economics of innovation.  For example, it never explained the Microsoft approach, where they deliberately were not the first in firm as a new technology emerged, preferring to let others come in ahead of them, then enter after the new technology has proven its importance.  A combination of deep pockets, being able to reverse engineer the products of its rivals, and integrating the new offering with existing Microsoft favorites (taking advantage of network externalities) was an entirely winning strategy for Microsoft in the late 1980s and throughout the 1990s.  Examples include how Excel took over from Lotus 123, how Internet Explorer beat out Netscape, and how Outlook/Exchange beat out Notes/Domino, Groupwise, and other email/calendaring software.  Similarly, Apple was quite a late entrant into the smartphone market but totally overtook Palm and RIM (Blackberry) with the iPhone. Yet Microsoft's strategy, didn't work nearly as well after 2000 and Apple too may now be slowing down for similar reasons.

Christensen's approach also doesn't address fundamentally new markets where the big players are not present at all, so there is no business line that will go the way of the dinosaur as this new market emerges.  When email first came along, it was fundamentally new as a means of personal communication.  It co-existed with phone, snail mail, and FAX, but until eCommerce became a reality, itself enabled by broadband, none of these other modes of communication really competed with email.  Because most Internet access at the time was by dial up, it's hard to see any cannibalizaion due to the emergence of email.  Quite the contrary.  The various uses appeared complementary. 

Or to put it differently, the lag between when the fringe market product or service really competes with the main market offering is not determined in the Christensen story.  Absent a firmer prediction about such lags, there isn't much of a theory here.  Nobody uses an abacus now nor do they use a slide rule.  But we didn't need Christensen to tell us why not.

Let me turn to how Shirky's story of disruption differs from Christensen's.  First there can be multiple sources of disruption. For example consider the following set of factors that impacted how the Internet affected (paper based) newspapers:
  • It enabled entrance of a bunch of purely Internet based news periodicals - Slate, Huffington Post, Politico, etc.  These may have had lower overhead than traditional newspapers, could better tailor their stories to their audience, and in the process seem less stodgy than traditional newspapers.
  • Via the blogosphere and smart phones with cameras, essentially everyone could be a reporter, sans training as a professional journalist.  
  • People like me could become pundits.  Each of us might not have a big audience.  But collectively we might suit our readers needs better than those writing for a newspaper or for syndication via reproduction in other newspapers.
  • Reading patterns are/were different online than on paper.  Perhaps more importantly, people who had been willing to pay for home delivery of a paper newspaper nonetheless objected to paying for a subscription to the online alternative, perhaps citing the mantra, "information wants to be free."
  • Digital natives may have had still different reading patterns and didn't view reading the newspaper as an important part of their daily life.  Without intending this, the Daily Show and later Colbert became competitors to newspapers for this audience. 
  • To the extent that local or regional newspapers were reproducing news from a national or international source and didn't have sufficient content just about their own area served, they were getting beaten up by online alternatives that were more broadly available.
Second, because of the multiple factors that create disruption, there may be no coherence to what will come next and indeed the future may be worse than the past.  This is not your grandmother's view of innovation.

* * * * *

Let me now turn specifically to Higher Ed.  Eleven years ago I attended what was then called the Frye Leadership Institute - very intensive professional development for would be campus CIOs and Librarians.  We had a session that focused on the potential and threat posed by for-profit Higher Ed, with a particular interest in the University of Phoenix.  The arguments given at the time were implicitly in the spirit of The Innovator's Dilemma.  Similarly, the more recent fascination with MOOCs also seems very much in the spirit of the Innovator's Dilemma.  Nevertheless, below I want to take a Shirky-like approach in discussing multiple causes for disruption and only then consider possible ways to address the issues.

High Tuition and Excessive Student Debt

This seems to be the number one issue with regard to undergraduate education.  Here is something quite recent from Bernie Sanders' Web site on college costs.   And here is something from a few years ago about Rick Perry and the $10,000 college degree.  While you may not agree with the specific remedies in either case, when you see proposals like these from both the Left and the Right, we should be able to agree that high tuition and excessive student debt is the number one issue.

Yet even this issue I regard as derivative, a consequence of the soft labor market that has persisted since the housing bubble burst.  If the labor market was where it was in the late 1990s, with a good job essentially guaranteed by getting a degree, the issue of debt and tuition would not be nearly as important as it is now.  Both would seem to be parts of essential investments in human capital.

Whether the current labor market softness is the new normal or is simply the not yet tail end of an already long business cycle is anyone's guess.  Likewise, there seems no consensus on whether a smartly designed and implemented fiscal policy could lift the economy out of its doldrums and thereby make job prospects better for many.  There is only a general agreement that with the present gridlock in Washington no such fiscal policy will be forthcoming.  Yet the proper response from Higher Ed depends on which of these alternatives proves correct. 

Learning versus Credentialing

Recently the question has been raised whether young adults who know how to advance their own learning actually need college and what, if anything, college has to offer them.  Conversely, if young adults don't know to aggressively pursue their own learning before entering college, will college transform them in a profound way so that after graduation they do?

Two recent columns by Thomas Friedman, How to Get a Job at Google and How to Get a Job at Google Part 2 speak to these issues.  In a nutshell, the pieces indicated Google is more interested in learning to learn skills than already acquired specific knowledge, though the latter can matter too.  They also want their employees to be good citizens and care about fellow employees, with a genius at learning to learn but a ditz at being a good citizen not the type of person they want.

There is, of course, a question whether this characterization of a desirable hire typifies much of the rest of the labor market or not.  For the sake of argument let's say it is typical.  Then a different threat than that the labor market is soft is that employers have become more picky about the type of people whom they will hire and the degree itself, even a degree coupled with a high GPA, is no longer the credential it once was for this purpose. 

Increased Reliance on Adjunct Instructors

This issue has gotten a lot of press of late and there are many dimensions to it.  I want to frame the issue differently than it has been presented elsewhere.  The question I want to get at is whether instructors have ownership for the mission of the university and therefore give body and soul into the fulfillment of that mission or if, instead, instructors are merely employees who dispense their teaching but nothing more, and therefore are largely unconcerned with the two previous issues as their own instruction might speak to how they are addressed broadly.

One might argue there is no guarantee such ownership will be found in tenured faculty and indeed that at research universities in particular the tradition has been for many tenured faculty to ignore undergraduate education.  I don't want to disagree on that point.   But here we are talking about factors that will encourage innovation and ultimately change the way Higher Ed goes about its business.  If tuition from undergraduate education continues to be a much larger share of revenues than it has been historically, then it stands to reason that these changes will reorient institutions in favor of undergraduate education.  We'd like to know whether the instructors will buy into that and perhaps drive it or if they will be neutral about it or actively resist it.  

* * * * *

One might be able to expand this list of issues but it suffices for me now.  Let us turn to possible innovation on cost, on learning, and on staffing,  But first a note about measurement, particularly on the cost end.  Carol Twigg, the force behind the Pew Program on Course Redesign emphasized what I say next and I'm following her lead.  It would be good for others to embrace this idea as well.  There is a tendency to consider cost as expenditure per student and not control for how students perform.  One measure of performance, for example, is how many students get through the course.  If expenditure per student is $x, but only 50% get through the course, than the real cost per student is $2x, as the money spent on those students who don't get through is essentially wasted, but is spent nonetheless.  The same general notion applies beyond completion rates to learning, but there we don't have hard measures like we do for completion rates, so getting at the right cost adjustment is harder, both conceptually and empirically.

The current fascination with technology innovation and learning, particularly video lectures delivered online coupled with social network interaction among the students and perhaps automated interactions as well, is based on the idea that much of the instructional activity is fixed cost and that lower average cost can be attained with large online classes, ergo MOOCs.  Yet if one does the particular type of adjustment Carol Twigg says we should, and consider learning as much as course completion, then it simply isn't obvious that this is the innovation that should drive everything else.  You need a mindset a la The Innovator's Dilemma to come to the conclusion that MOOCs are the game to play.  With a Shirky-like mindset instead, you might conclude there is a need to try higher touch approaches, in addition or instead, in the hope of finding ways that both improve learning and don't break the bank.

In the rest of this piece I'm going to focus on public research universities with residential students, not all of Higher Ed.  Having spent my career at Illinois, I know more about such institutions than about other segments of Higher Ed.  I haven't seen anyone else make this next point, but it may very well be that appropriate innovation needs to be segment specific.  And, following Brown and Duguid, it may be that the innovation needs to be fundamentally social, rather than fundamentally technological, and yet consistent with current social structures.  Without that consistency institutional inertia would block the innovation entirely.

For the reader who is new to my posts, know the particular possible innovation I'm suggesting below has been a pet idea for quite some time.  Back in 2005 I wrote a series of posts on Inward Looking Service Learning that consider the innovation from the perspective of the student providing the peer-mentoring for the other students.  Below I'll consider the same idea from the perspective of students currently enrolled in the course.  Also, the reader should ask herself whether this represents a drastic innovation or is only the consequence of incremental change, as Lepore suggests.  In the late 1990s I used undergraduate TAs, who mainly interacted with students online, in a large section of intermediate microeconomics.  We don't seem any closer to Inward Looking Service Learning than we did back in 2005, so I'd vote that the suggested innovation is drastic.  Yet it might not seem nearly as threatening as MOOCs may seem to some, because it is a move to high touch and because it doesn't automate away the role of the instructor.

The key to the idea is to make the study group a formal organization that the campus recognizes and schedules.  In so doing, one might think of this idea as the flipped classroom on steroids.  Take what traditionally has been a 3 credit-hour course. Schedule it for 5 hours instead of 3.  Let one of those hours remain a place for ensemble teaching and learning in a regular classroom.  Let the remaining four hours be scheduled in two-hour blocks and be for the study group to meet.  Different study groups need not meet at the same time nor at the same location.

Let's say each study group has five members who are enrolled in the course and one peer-mentor who has taken the course previously and is concurrently mentored, along with the other peer-mentors, by the instructor for the course.

The primary goal for each study group is that members would give voice to their own formative thinking.  Having one student take notes for himself while the other students talked would be unacceptable.  Everyone needs to participate.  Above we discussed that instructors and administrators don't like to look stupid.  Students don't like to look stupid either.  The study group is aimed at creating a friendly enough environment that students become comfortable making mistakes in front of one another as part and parcel of their way toward understanding what is going on.  Having developed such comfort the hope is that students would care deeply for their fellow members of the study group, teach each other, and help one another when they struggle.  The mentor's job would then primarily be one of facilitation, also to monitor if their are group blockages and then try to work through those, and to have one-on-one sessions with any student who seems to be struggling much more than the others.  The particular activities of the group would otherwise depend on the subject matter of the course and how the instructor would like study groups to operate.  By scheduling the study group meetings for more than the time current courses are scheduled, the hope is that students would become used to putting in more time on task and to understand that real learning happens by working through blockages in understanding.

Where are we currently in regard to this specific innovation?  I'm aware of a few early adopter types among instructors who have tried something with peer mentoring, though obviously not yet with campus sanction.  The ones I know have reported very positively about their experiences.  Yet the idea hasn't been implemented outside the handful of early adopters and thus is not close to any tipping point.   The reader may think that is because the peer-mentoring is a cost add.  In my INSL essays, keenly aware of that concern, I worked through how it might be cost neutral or even cost reducing.  The key there is that mentoring is extremely valuable as an activity for the mentors, learning by teaching, so should be given credit accordingly.  That is the part which some might find more controversial, but in the current climate maybe not.

* * * * *

There is a different way to think about innovation, externally driven or internally generated.  Disruptive innovation implicitly conveys that it is externally driven.  Incremental change a la the continuous improvement model comes from within.  A moment's thought, however, suggests that these are not mutually exhaustive.  Why not drastic change that comes from within?

I'm afraid that most of us in Higher Ed are guilty of at least one of the following.
  • We tend to deny evidence that our students are not learning deeply.
  • We don't monitor instruction much at all and instead rely on course evaluations rather than direct observation.  So we don't have a good sense of issues like teaching to the test, whether courses offer appropriate challenge for students, and the relationship between these things and the type of evaluations that are generated.
  • We tend to want to jump on the train that appears to be leaving the station for fear of being left behind.

What if we could overcome our own limitations and then designed the change we thought best based on an empirical study of the actual conditions on the ground coupled with the big picture issues I've sketched above.  Having done this, what design would you come up with?  That's what each reader should be asking.

No comments: