Bryan Caplan  

Signaling Versus Educational Innovation

Ron Bailey on Confirmation Bia... New Thinking in Macroeconomics...
Tyler wants to use my little signaling model to predict the future of online education.  At risk of looking a gift horse in the mouth, I'm afraid a much richer model is required to address Tyler's question. 

In the interest of parsimony, my model assumes that education is purely a signal of IQ; Tyler also considers a variant where education signals conscientiousness instead.  So far, so good.  But as I've said several times, in the real world education is also a signal of conformity.  One of the main things a stack of degrees says about you is, "I uncomplainingly submit to social expectations." 

This makes educational innovation inherently difficult.  Why?  Because the first people to sign up for innovative alternatives to traditional education are usually people who have a beef with the powers that be.  As I've told Arnold before:
[E]ducation doesn't just signal intelligence and conscientiousness; it's also signals another character trait employers pragmatically cherish: conformity.  This leaves us in a catch-22, because experimenting with new ways to signal conformity is a strong signal of... non-conformity!
You could of course reply, "All that's going to change.  The future is coming."  I'll bet against it.  In fact, I already have.  It's easy to imagine a society where traditional educational credentials could collapse at a moment's notice.  But that society is not ours. 

Take out your sociological goggles and look around.  In our society, smart, hard-working, conformist kids go to old-fashioned brick-and-mortar colleges.  Their elders expect them to do so.  Their peers expect them to do so.  They feel like losers in their own eyes if they don't go.

The normativity of conventional education isn't a passing phase.  College attendance is a central tenet of our society's secular religion.  A student who scoffs at all these expectations probably has a serious problem with authority.  Would-be employers treat him accordingly.

There may well be a niche for online education.  Maybe it will attract the best students who currently don't go to college and the worst students who currently do: the top of the bottom plus the bottom of the top.  But until we sharply reduce subsidies for traditional education, traditional education will continue to dominate, warts and all.  Middle class jobs will no longer require college only after middle class kids can no longer afford college.  Hail austerity!

COMMENTS (22 to date)
Tyler Cowen writes:

You don't need to overturn all convention. The top schools could shift at the margin, as they have many times in the past, and suddenly the conformist thing to do is to have ?? percent of your classes be on-line, and so on. In virtually any other context you would see the flexibility of the market here! No major credentials need to collapse, if it turns out that cannot happen easily.

Vipul Naik writes:

There are other pathways.

First, online educational tools may emerge that are mostly aimed at covering the (useless?) college curriculum. Students will still go to college, but will rely more and more on the online educational tools for their actual learning (of the useless stuff).

Second, once people are deriving their actual "learning" from online sources, there may be a push for these online learning sources or tools to cover topics and skills that are more relevant to employers or job-specific skills.

Third, once the superiority and greater relevance of online learning sources is established, employers may start feeling more comfortable giving credit for such learning, over and above college degrees.

At some stage, the college degree will start losing importance.

This seems like something that could happen with reasonable probability over the next 150-200 years.

Glen Smith writes:


I'm going to assume that those "useless" classes are pretty much the ones I'd consider "useless". For me, those "useless" classes were the only ones I needed classes for since those are the ones where I needed work ethic and discipline. The classes in my major took almost no discipline or work ethic because they were fun and useful.

Doug writes:

Well maybe things will change when the new economy pushes employers to value conformity less. Or at the very least value it less in the very high cognitive elite.

If (at least some) of the most prestigious employers valued highly intelligent non-conformists then Alt-Harvard might have a chance to compete against the real thing. This doesn't seem so far off when the largest company in the world's motto is "Think Different."

PrometheeFeu writes:

I think you need to more precisely define what you mean by conformity Bryan. I can definitely see how employers want employees to conform to basic social norms: don't resolve disputes via fistacuffs, don't tell everyone about your sex life in the middle of the quarterly planning meeting, etc... But outside of that, employers desire "culture match" AFAICT. So some employers might wish a high degree of deference to authority (don't want the assembly line worker making a fuss about the color of the widgets you're making) while some may wish a much lower degree of deference to authority. (if your Stanford-trained engineer thinks his boss' idea will make the bridge collapse, you want him to speak up and vigorously defend his point of view)

Anton Sherwood writes:

If the market becomes freer so that the demand for bosses declines (i.e. self-employment rises), demand for signals to potential bosses should also decline.

Zippy writes:

So . . . a question that I don't think you've ever addressed. How much conformity is actually necessary to function within organizations? Yes, it helps if you bathe at least semi-regularly and have some degree of ability to get along with others.

But do you actually need to be an "Organization Kid" of the sort who adroitly jumps through the hoops needed to get into an Ivy League school in order to be a productive worker?

Could a company find a competitive advantage targeting its hiring among high IQ non-conformists?

Currently, use of IQ tests by employers is pretty much illegal, because of all the race stuff and "disparate impact." If the theory of disparate impact went away tomorrow, and employers were at liberty to give IQ tests directly, might it (at least somewhat) break down the useless credentialism?

DThinker writes:

I suspect that Professor Caplan is massively overemphasizing the value of conformity. In particular, in an economy that is moving towards the information age, I think independent thinking is becoming more economically valuable.

Being able to get along with others and work in teams and conformity are two different things.

I also think that Professor Caplan is massively underestimating the value of actual concrete skills in our modern economy. Do you know how to program in Java? Python? Ruby on Rails? Do you know how to handle Big Data? Have you ever used Hadoop? Do you have experience with Source Control Management systems? Can you write SQL queries? Optimize them? Apply normalization? Do you understand the principles of creating database tables?

These are just examples of the sorts of concrete knowledge that is highly valued in today's information age economy and which schools actually help impart. Online education is going to make this sort of knowledge all the more accessible and accelerate the transformation of our economy.

The idea that a university education primarily signals intelligence, conscientiousness, or conformity may be true for some majors, such as in economics, where not as many real world skills are actually learned. But when it comes to fields like software engineering, electrical engineering, civil engineering, or any other sort of engineering, this simply is not true.

A lot of economics education is based on theory. Of course employers are going to want to hire really smart economics majors for the same reason they hire really smart philosophy majors; for the critical thinking skills. But this is not the case for some other fields, and in particular, for the fields that are increasingly dominant in the information age economy.

ChrisA writes:

I am sympathetic to the signalling model but I have some doubts. My wife likes to go to Zumba dance classes. I guess there is no need for her to do this; she could view the dances off the internet and perform them in her own home. I don't think a signalling model explains why she takes the time to travel across town to be in a class with a bunch of other people with a real live instructor which she has to pay for.

It's similar to the fact that I am still travelling to meet people for business meetings, often at great expense, rather than just calling them on Skype. I also attend conferences which cover my topic of interest, rather than just viewing the videos over the internet. There is no signalling model that can explain that.

I think therefore there must be some utility and/or efficiency that come with a "real" instead of a virtual interaction. I don't really know how to describe the difference, but it is there. That may be what recruiters are buying?

John Roccia writes:

Professsor Caplan,

I work for a major online university, and I see every day evidence that massively supports the signaling model.

One of the most frequent questions we get from both prospective students as well as information-seekers on their behalf (guidance counselors, parents, etc.) is: "Can an employer tell that the degree came from an online school?" Since the university I work for has ground campuses as well, all degrees are conferred the same way, and there is nothing on the diploma to indicate the student took their courses online as opposed to at a ground campus. This is a major plus for our prospective students, who have their various reasons for wanting to take classes online, but (even though they don't know to phrase it like this), don't want to signal a lesser education.

Also, I can tell you strictly with numbers from our career services division that even BEING in school helps job prospects. Students that haven't even graduated, but that put on their resume that they're ATTENDING school get a higher percentage of callbacks, interviews, etc. than students with similar resumes who don't. If the signaling model wasn't correct, why would students who haven't even finished school be more valuable to employers? But the fact is, even attempting school signals to employers that you have more ambition and drive than students who do not attempt.

I don't want to say that college is USELESS, mind you. I'm sure for many professions, like engineers, there is a valid amount of learning that college gives you. But I also offer the following amusing anecdote:

This past Easter holiday, I happened to be talking with my cousin "Jake" (names changed to protect the guilty). Jake has a successful career in finance; he's worked for the same company for 15 years and has risen quite high in the ranks. Jake and I were talking, and the conversation happened to drift to the signaling theory. He chuckled and said, "Well, I can definitely tell you that's true. You don't really think I have a degree in finance, do you? I never attended a day of school past 12th grade. I just said I did, got the job, and did fine. I never encountered anything I didn't know how to do, so what does that tell you?"

It tells me Bryan Caplan might be on to something, Jake. I wonder how many stories like this are out there?

Dan Carroll writes:

I tend to agree that Prof Caplan is overrating conformity. I think it is important, but mostly in professions that shouldn't require a degree. I think the bias against "online degrees" comes from the fact that the online pioneers catered mostly to the AA and technical school crowd. Conformity is probably still be a factor, but partly due to the fact that most individuals are not going to experiment with their career on an untested model, and employers are going to be reluctant to do the same kind of experimentation. However, often employers barely read the resume. So a lot of the signaling is better described resume padding - putting the right keywords on your resume.

I think the strongest case for the stability of the current education model is the government subsidies and mandates, which are built on the basis of soundbites. I think there is a strong case for the signaling model, but a weak case for the stability of the current system absent government support.

Chris Stucchio writes:
Could a company find a competitive advantage targeting its hiring among high IQ non-conformists?

Zippy, many tech startups do exactly this. Mine certainly does.

For example, on interviews, I don't ask about degrees or experience. I ask people to code. My last hire never told me if he went to school until months after he came onboard, when it came up in casual conversation.

Umpa Lumpa writes:

[Comment removed for supplying false email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Nathan F writes:

As an instructor and soon to be (hopefully anyway) professor, I see the online education as a challenge and threat.

I believe the end result will be this - we will have to economize on what we teach. By that I mean, if something is so easy that you can learn about it from the internet, then we will no longer teach that in person. A lot of courses I took in undergrad were easy enough that no instruction was really necessary, just have someone tell you, "hey learn this, I will test you on it."

However, the truly difficult subjects, the comparative advantage of teaching, is in the difficult stuff. The things that you could post online and would still require explaining. Nearly all my graduate work was like this, in addition to some key undergraduate courses.

Maybe I biased when I say I believe economics is like this. Its a subject that you can't simply say, "read this and you will get it." It often requires a good professor to really prod the mind to show the student the beauty of the field.

If online classes begin replacing actual instruction, the actual instructors and professors' optimal response will be to up the level of difficulty to maintain our comparative advantage. I see this as a positive outcome, the students get a high quality education.

BZ writes:

Although I'm firmly in the "Of course Education is signaling; you mean there was another theory?" camp, I also think the conformity thing is vastly overrated. One can get several degrees while being the jerk in the class with his hand always up questioning the professors outlook on things. It happens.

Steven Kopits writes:

You know, that's an unbelievably cynical post.

I really enjoyed it.

John Roccia writes:

@BZ: Sure, but nowhere on your degree does it say "Cum Laude - But He Was A Non-Conformist Jerk."

The unspoken assumption is that you can't have been TOO much of a jerk, or you would have either gotten frustrated and quit or been kicked out. Either way, a degree signals that you're willing to play by the rules in order to get the Brass Ring. For employers, that means that in order to get the Brass Ring of your paycheck or that next promotion, you're willing to do what it takes.

Lonely Libertarian writes:

So as I read Professor Caplan's post I immediately thought...

1. Steve Jobs...
2. Bill Gates

Arguably two of the most significant contributors to our way of living - they were both the very definition of non-confomists...

So I am thinking the Professor is right - traditional education paths signal conformity - which in today's world may be of little or no real value...

Perhaps we are seeing one of the possible benefits that on-line education can leverage - delivering highly curious, highly motivated, creative, non-conformists...

David S writes:

I think that college provides both education and signalling - the amount of each depends on the major / college. The different benefits provided by colleges will result in differently optimized online colleges.

One could foresee a very picky online college, for example, giving a very high signalling value. It may teach nothing - essentially just providing an IQ test for admittance, and requiring you to click a link each morning. From a marketing standpoint, this is a hard sell - you need to convince both the students and HR bosses that you have high standards and won't sell out.

On the other hand, giving an actual education is easier - teaching engineering, for example. It is also easier to prove that your students have learned engineering to prospective students and HR bosses.

So in the short term, I'd expect to see a rapid rise in education focused online colleges - specifically in engineering and related fields. Longer term, you will see the signalling online colleges as well - unless Harvard wises up and beats them to the punch.

Dan Jelski writes:

I don't think conformity is a very important criterion, but something related is - an alum's social network. A Harvard grad, for example, arrives with a Rolodex (is that still a word?) full of other Harvard grads. Of course a certain amount of conformity is built in - all Facebook Friends are alike - but the Admissions Office assures that upfront.

infopractical writes:

If employers change their thinking, that opens the door to change. As one of those employers, I see others around me -- we talk about these things -- who agree that we'd rather hire somebody who can "get the job done" rather than somebody who devoted so much time to polishing their signaling mechanism.

We may not be the majority -- large corporations hire more, and are going to be more conformity oriented. But it's hard to see a reason why the ranks of successful hackers of life won't grow to take up a larger share of the markets.

Vera writes:

Among the other margins mentioned by commenters by which the conformist's path might shift gradually: /re/-education. Someone who already has a degree who wants to change paths can supplement their standard college education with online classes without sacrificing their signal of conformity.

Comments for this entry have been closed
Return to top