The New York Times has a useful article today on MOOCs – Massive Open Online Courses. Reporter Alan Finder points out that online education has been around in one form or another for a decade. What’s different today is the rise of the MOOCs; the article walks through the three leading MOOC providers – EdX, Coursera, and Udacity – and describes how each works and the important differences between them. One day, I’m pretty sure, online education in various forms – through MOOCs, or in combination with traditional classrooms, or other ways yet to emerge – will be a genuine alternative for both educational content and higher education credential. For many reasons, however, that day is still a ways off. So I’m interested in asking what the value of online education is today – its value for an undergraduate currently in school, not waiting for institutional changes in higher education itself, at the level of the whole system.
If an undergraduate doesn’t plan on going on to further graduate or professional education – for which GPA matters – then one might be able to take classes for their practical, real world educational value, even if one’s GPA suffers (because, after all, a reason to take these courses in a terminal degree program is that one is relatively, somewhat less worried about the signal, but instead seeks the content, which means deliberately choosing courses where one has little background knowledge, at least by comparison to other students who, looking to protect GPA, only take classes for which they are already well-prepared). GPA matters in the real world more than one might think, especially in the early years of competing in the real world, especially for liberal arts majors looking for work in a tough environment: GPA matters because one is competing for those positions with people who, educated or not, got splendid GPAs, even if they were in fields like anthropology or religion or women’s studies – fields where one is graded rather less on sense than correct sensibility. There aren’t a lot of other signals available (well, not really true, in the day of on the Internet all is revealed).
However much GPA matters in the real world in one’s terminal degree, the reality is that if one is planning on competitive professional school – or you’re 18 and you just don’t know – GPA matters hugely. The credential trumps the education, pretty much. If one is applying to law school, for example, it’s just undergraduate GPA and LSAT score. It doesn’t matter what you studied as an undergraduate and, so far as I can tell, it doesn’t even matter within a pretty broad range where you studied; reporting for US News rankings does not favor one school over another, nor does it favor one undergraduate major over another. Certainly many law schools or business schools, especially at the top end, do take internal account of the undergraduate institution and some smaller number might look at what you studied – Brian Leiter mentions that University of Chicago Law School looks favorably on philosophy undergraduates, for example, and I’m sure that’s true in a variety of top end schools. But the reporting incentives to USNWR don’t reward you for it.
So successive levels of the educational system are reached as a matter of putting credential over education, while asserting that they are pretty much the same thing. It’s great if both those work out for you, but for many people, that not true. Which means that you need the credential in the short term, in order to go up the next rung of a system that compounds rewards more steeply at the top – AverageIsOver is really WinnerTakeAll. Yet, in the long run, you also need education in the sense of knowledge and skills. Unfortunately, in the long run we’re all dead, or at least downwardly mobile; credential beats education in the short run.
Moreover, I suspect Tyler Cowen is broadly right in his main thesis that, in the AverageIsOver economy (and in its implied class structure, which is scarily shifting against whole tiers of the New Class), the skills that matter are those that define one’s relationship to the computer, automation, robots, and machines. That does not mean, as he emphasizes, that everyone should be an engineer, or that if you can’t be an engineer, you are doomed. It does mean that students who aren’t going to be good, or highly successful, at studying STEM still need some level of education and competency in it. You may not be interested in machines, but machines are definitely interested in you.
The problem today is that non-technical liberal arts students often can’t risk what it takes to get that education, if the consequence is that their even modest relative lack of success in those courses (assuming ones at their level are even offered) undermines their undergraduate GPA. I’ve described why this is so in past Volokh posts; university grade inflation strongly favors risk aversion, because it’s not actually inflation – it’s compression against a fixed top that gives you many more possibilities to go down than up. Carrying that discussion over to this – the problem of how to get needed education – I’ve suggested that one solution would be for universities to invest in pass-final minors in technical subjects for non-technical students. Pass-fail so that there’s some incentive to attend and learn something, but no mixing in with GPA, which is to say, no mingling of educational functions with credential functions. Universities have many reasons they don’t do this, none of them persuasive to me as the parent of an undergraduate, but in any case it’s pretty clear to me that they won’t be helpful in these ways, to students seeking to navigate the AverageIsOver future.
This is where online education comes in. Much of the debate over online education has to do with trying to fit online education, particularly but not exclusively MOOCs run by enterprises outside of some particular university, into the existing credentialing structure of that university – get the online grades and credits recognized. At some point down the road, I imagine something will emerge – I assume it will have a lot to do with how to split income streams. But that debate is about the university’s credential of graduation, diploma, and GPA – something I’m not concerned with in this post.
Whereas at this moment, the value of online education is not being able to to get your university to accept it as credit. On the contrary, it’s the fact that it doesn’t enter your GPA that matters. These online courses give a student an opportunity to learn something that doesn’t show up on the transcript. If the problem is that you can’t risk credential for education, finding off-GPA-balance-sheet (if I can put it that way) ways to learn skills, and especially skills you might not be naturally good at doing, is a feature of online education, not a bug. Online courses are a good way to gain very basic introductions to subjects that you can’t risk messing up but which would stand you in good stead in a world defined by your relationship to smart machines. It won’t turn you into an engineer – but that’s not the point. The world of smart machines does not need, as I said, everyone to be a STEM technician; it needs people who can, as Cowen puts it, either (a) use machines to leverage their human skills and aptitudes, or else (b) use their human skills to leverage the machines. If, in today’s credential-mad world, you can’t risk GPA to acquire those skills – then off-GPA-balance-sheet online education might be the best answer.
Basics of computer programming for English majors, for example. Statistics for English majors. Accounting for English majors. Robotics for English majors. Very practical and frankly very minor skills, too, that surprisingly many students don’t acquire – Excel for English majors, etc. But there are two important issues that, if addressed, would make this model much more useful for today’s undergraduates. One is that the existing MOOCs are not really aimed at this kind of audience – rather, many of them are aimed at the world’s technical students, to put a Stanford engineering professor and technical class in their reach through the internet. That’s great, noble even – but it’s a different kind of educational purpose than the one I describe here.
We need courses designed specifically for the English majors that have genuine content, skills, and achievement, but suited for someone who might one day manage, or at least have to interface intelligently with, technical people. Or increasingly, if Cowen is right, interface with the machine itself. Reading and understanding Frankenstein as an English major is good – no one should mistake my endorsement of that, both for its own sake and for the utility to other things that it can convey. It’s not an accident that robots are in the first place an invention of literature; neither should anyone be surprised that so many technical STEM people are in deep thrall to a vision of robots in fiction and literature dating to many decades ago. But reading and writing about Frankenstein in the world of imagination is one thing – interacting with Frankenstein in the workplace, even one programmed in ethics by Asimov’s Three Laws, and in order that it produce added value to you, or else you produce added value to it, is quite another. Facilitating that interaction with the machine can be served by online courses, if (well-)designed for that purpose and that audience. (And I think there’s a significant business opportunity here, by the way.)
The other issue is that, while it’s in the student’s interest to keep such courses off-GPA-balance-sheet, it is nonetheless in the student’s interest to be able to have some form of credential indicating successful completion, as an indicator of an acquired skill or body of knowledge. The online achievement “badges” that have been widely discussed is a useful concept here – provided they accurately do what GPA does not, and capture the actual level of the material taught and learned. Badges are useful because they disambiguate skills – exactly the opposite of the GPA’s homogenizing effect – but they are only useful if the badge clearly distinguishes introduction to computer science for English majors from computer science for Stanford STEM graduate students. And if the badges are credentials that actually are signals of skills or knowledge acquired.
So if I were an undergraduate in the humanities or liberal arts today, I’d probably be doing my classes and, even in this WinnerTakeAll world defined by technological innovation, I’d still study philosophy. But apart from my classes, I’d also be finding ways to rack up online courses. As cheaply as possible, of course, and if free, even better. But free or not, they have to provide educational value, value defined as teaching me useful skills and knowledge for the real world, in fields in which I can’t compete on-GPA-balance-sheet. This seems to me one of the greatest utilities of online education as it is actually available to undergraduates today.
(As with all my posts, comments are closed. I’ll clean up some grammar and clarify some phrasing tomorrow.)