Bryan Caplan  

The Best Kinds of Free Education

PRINT
"Some Men Just Want to Watch t... Designing Men...
EconLog reader Alan Shields sent me some interesting comments on my observation that the best education in the world is already free.  Reprinted with Alan's permission:

I've been thinking about your thought experiment of "who would choose to take a course for no credit". Let's turn the problem around a bit: if one were offered two years of free schooling on the restriction that they would receive no credit for courses, no endorsement of skills, and no recommendation from the teacher, what courses would one take to best increase one's earnings? Programming courses came to mind (I am a programmer), but even there I found more good out of books and experience than courses. Materials already available are sufficient. My second thought was sculpture, craftwork such as sewing and design, auto repair (I know auto repair was extremely popular at the community college I attended when young). While these courses may not have the bang of a STEM degree, they still provide usable skill without certification.

So this is where it's interesting: all of those crafts already have an enormous amount of material online. My wife is a crochet fan and she learned from YouTube videos which only came out in the past two years - and she's repeatedly heard from older crocheters that the YouTube videos are vastly superior to all of the teaching books and supplies available before. I am unsure as to the quality of teaching in auto repair and plumbing, but I've heard of some home repair how-to projects. Just recently I learned how to use a strap wrench from an online video.

My theory is that we can already see where the most valuable uncredited courses are because the coursework is already being aggressively covered by online courses provided for free. I think it's no coincidence that these courses are in fields that do not require credentials (plumbing may be regulated, but I don't think being an apartment handyman or manager is regulated).

Thank you for your time,

Alan Shields


Is Alan right?  Other suggestions?


Comments and Sharing





COMMENTS (11 to date)
Scott Wentland writes:

Public libraries have been free for a very long time. One could effectively earn multiple PhDs worth of knowledge over a lifetime by checking out books there, without spending a penny (but I don't think anyone lists their library record on their résumé).

Ken P writes:

Udacity's Python course has enrolled 200,000 students who build a search engine in the process (experience) and discuss in the forum. Python, Machine Learning, NLP could probably put a person in STEM salary range as an employee.

I agree with Alan that many trade skills can be gained online. This has been taking place for some time without the need for an employer. Art, crafts, photography, web design, graphic design... If you can't find an employer, hire yourself.

Daublin writes:

I found my college education valuable, as a total package.

However, I confess feeling the same as Alan about the prospect of taking college classes to learn something. It's really hard to think of cases where college classes would do better than Wikipedia, YouTube, and Amazon.com.

I mean, take all the advantages you can think of for college classrooms. Now imagine you could skip over lectures that are not what you want to learn -- that's an hour of saved time each time you do it. Now imagine you can ask a different teacher whenever you are stuck on something. On top of all that, imagine you waste zero time stressing through tests and trying to figure out what the teacher wants you to answer.

I realize the dichotomy here, and I don't have a great explanation. My best try is that while a person is in college, they typically don't have a full-time job.

patrick writes:

I'm basically building a house with YouTube videos. I kid you not.

zgatt writes:

We all learn on our own, but do you expect you'd get a better foundation from reading about something or from interacting (even in a classroom) with an expert who is devoting time to explaining it?
There's this notion going around that once the best teacher in the world records their lecture, then teaching the subject is dead. Do you think that's correct?

fel writes:

-Writing. It's useful in many careers. It's extremely difficult to learn (well) on your own, though, because it requires practice and feedback. It's possible to practice on your own (though it requires a lot of self-discipline, and most people will limit themselves to things they're really interested in), but it's impossible to give yourself good feedback. This is because, by definition, you lack the expertise (why else learn?), and people in general are really bad at self-criticism.

-Acting. It forces you to become aware of how others perceive you, and public speaking is pretty helpful. And this is impossible to learn online; in order to practice public speaking, you need other people. Feedback is also pretty important here, too, as an expert will pick up on many subtleties that a novice will miss.

-Upper level math or logic in the philosophy department. These classes teach you to think more rigorously, always a good thing. This can be really hard, which makes it really tempting to skip the harder stuff as a self-learner. Also, feedback is crucial. I don't know how many times I handed in a problem set, thinking I had constructed such a clever proof, only to get it back and find I had missed something really subtle.

Clay writes:

Someone can learn a little plumbing, car repair, or programming and see that translate into better job options. STEM education doesn't work like that. Just a little STEM education without credentials is generally worthless in terms of practical usability and job opportunity. But a lot of STEM education is extremely valuable.

Alan, and the self-study learning masses, are typically focused on the short term, easy, tangible, non-abstract skills, that are either fun or have a very clear route to financial pay off. Alan names basic programming, auto repair, plumbing, and crochet which all illustrate this point.

Silas Barta writes:

@Scott_Wentland: That's really pushing it. A PhD program involves extensive interaction, clarification, feedback, enculturation, etc that is far more efficient at weeding out misinterpretations or miscommunications in a text. (I grant a bachelor's may not do this very well.) So libraries of the past could not accomplish that.

(On the other hand, since the internet's large-scale, many-to-many interaction model does allow close facsimiles to that and so has much more potential to give you a PhD's worth of expertise.)

In the language of compsci, the amount of books you would have to read -- if books were your only means of obtaining knowledge -- increases super-exponentially in the amount of expertise you want to achieve, due to the inability to ask a polynomial amount of questions of the PhD-program oracle.

Alan Shields writes:

@Clay: you make an interesting point! Let's take it a bit further, especially in light of other comments.

I taught a couple of programming classes for non-credit at the research institute where I was employed. I've also spent time helping junior programmers. I'd separate teaching activity into lecturing, answering questions, and mentoring (working through someone's completed solution and advising).

The lecturing portion and some limited question answering were well served by the library back in the day. Newer self-study resources like YouTube and various websites seem to do an even better job, but still lack mentoring.

So thinking about the problem again: if I had two years of uncredited class time, I would either opt for concrete skills I want to attain or I would opt for intensive mentoring in skills I wanted to advance. The former is still illustrated by the rise in materials for self-study.

Having said that, while intensive mentoring is most certainly available for college students during office hours, I think most professors can attest how few utilize that most valuable time.

Looking in other places, talking programmers in their early-to-mid twenties, the availability of mentoring at their job is a big feature. I know a few people who have taken less pay in return for a job with access to better mentoring - a revealed preference in cash.

To go a bit more nerdy, you can buy teaching time from professional StarCraft players. You pay them by the hour and they'll cover anything you want. Advice from fans is commonly: it will help you immensely IF you have your basics down and have played enough. In other words: once you've passed the big gains from self-study.

If people are buying mentoring for jobs and videogames, I think this reveals where signaling is least powerful.

Charlie writes:

You ask an interesting question - 2 years free, but not certification coming from it... I'd probably go and take all the classes that I couldn't fit in during my undergrad. There's a fair amount of hands on EE that would be interesting. The advantage of a university comes from access to the labs that have all of the necessary equipment for completing the projects (soldering stations, oscilloscopes, bins of components, etc) or even software (high end cad/verilog packages that would cost thousands to buy).

I think the equipment access would also make it's way into things like machine shop, auto-repair, welding and other more hands on classes. Usually, coming out of any of these classes you have a thing and you can point to it and say "i made this" and someone looking for that set of skills can look at it and see if you mastered the necessary techniques for future work.

Other interesting things might be classes like networking - possibly focused on some industry recognized certification. Classes provide lab environments where you can put the knowledge to work and see how various changes to the environment actually impact it. While the class doesn't provide certification, it may prepare you for a certification exam offered by a third party.

Alan Shields writes:

I'd also like to say that availability of independent materials does not indicate where college is necessarily worse, unneeded, or obsolete. It's to show where it is at its least signal-y.

College may still have a signaling premium for these areas, but where people would otherwise still want the service is a good sign.

Comments for this entry have been closed
Return to top