Students anywhere can watch my old friend Ben Polak teach his famous Yale class.  They can’t get a Yale grade for the class but that possibility is coming ever closer:  A professor at Stanford is teaching a robotics class and everyone can sign up, do the assignments, take the exams and get a certificate of “accomplishment.  Prospective employers do not know whether your friend took the exam for you. This means the certificate has little value.  But surely it is only a matter of time before some verification mechanism is set up and this problem is dealt with.

The implications of this change are multifold but I just want to focus on one: the impact on the research university.  Universities produce research as well as teaching and this other dimension is often forgotten in all the discussion of virtual teaching. Here is one possible sequence of events:

1. Virtual teaching cannibalizes face-to-face teaching.  Tuition goes down and courses become quite cheap.

2. This destroys tuition-based universities which turn into vast teaching factories.  A few universities try an “elite” approach with tiny classes taught by excellent teachers.

3. Endowment based universities continue to survive.  Researchers become concentrated in these universities.  They compete for government funding and do mainly PhD teaching.

4. A “top heavy” university structure emerges with a handful of research universities and a number of vast teaching universities.

This analysis assumes there is weak complementarity between research and teaching.  If there is strong complementarity, the teachers have to be researchers to keep courses up to date, exciting etc. This will make step 2 above more difficult and leave a structure like today’s but with universities having virtual counterparts and huge scale.