Bryan Caplan  

Education and Signaling: Reply to Bill Dickens

GDP and Well-Being... Monetary Theory Question...

As you'd expect, Bill's case against the signaling model doesn't convince me.  Here are my thoughts, point-by-point:

I take it that you think that nearly all of the value of schooling is signaling?
My best guess is that 80% of the private return of education reflects signaling, and the remaining 20% reflects genuine skill acquisition.  I'd add, though, that even modest concessions to the signaling model drastically slash estimates of the social return to education.  Suppose just 20% of the private return reflects signaling with zero social rate of return.  Then even ignoring tuition (i.e., foregone earnings are the only cost of education), an 8% private return implies a 6.4% social return. 
I used to take that view too, but the accumulation of evidence that I've seen leads me to believe that isn't the case.
I'm curious to hear more about what convinced you in the first place before you changed your mind.
For one thing I find it very hard to believe that we would waste so many resources on a nearly unproductive enterprise. There are plenty of entrepreneurs out there trying to make money by selling cheaper, in time and money, versions of education and they aren't very successful. Mainstream schools have experimented with programmed learning, lectures on video, self-paced learning, etc. and none of the methods have caught on. Why wouldn't they if they worked?
Your objection makes great sense if school signals nothing more than intelligence.  But what if school also signals conscientiousness - or, more plausibly, a blend of intelligence and conscientiousness? 

Suppose you open a school that supposedly accomplishes in two years what regular schools do in four.  What kind of students enroll?  On average, lazy shirkers who are trying to get their degrees the easy way.  And the smarter your students happen to be, the more negatively employers will rationally judge their conscientiousness.  After all, a kid with a 150 IQ can easily get a four-year degree; why's he slumming at Quickie Tech?

But what if your alternative school makes a strong effort to weed out lazy students in order to avoid this negative equilibrium?  You'd still face another serious problem: Choosing an alternative school signals weirdness almost by definition.  And if weird/non-conformist students make bad workers, you can easily get stuck in a "normal" but highly inefficient equilibrium. 

On top of all this, conventional education has an extra edge: large government and private subsidies.  You might say that donors wouldn't support the status quo unless they valued it, but I disagree.  Maybe donors support the status quo because the market for charity has very weak feedback on effectiveness, and they falsely believe that education is a great way to build human capital.
Of course its hard to believe that reading novels and poems contributes much to ones productivity on the job.
I'm delighted that you admit this point.  I feel like most economists who work on education just want to download data - and throw all their first-hand knowledge about actual classrooms down the memory hole.
So how do I square curriculum content with my view that education is productive? Here goes:

1. Education isn't mainly about learning specific subject matter. Rather education is mainly about practicing the sort of self-discipline that is necessary to be productive in a modern work environment. High school allows you to practice showing up on time and doing what you are told. College allows you to practice and work out techniques that work for you that allow you to take on and complete on time complicated multi-part tasks in an environment where you have considerable freedom about how you spend your time.
Relative to sitting around smoking pot, I agree that school inculcates some productive character traits.  I'd count character development in the 20% of schooling that actually builds human capital.  However, I deny that education inculcates productive character traits relative to having a job

Work inculcates the worker ethos; school inculcates the student ethos.  The two are only moderately correlated.  The most obvious differences: Work offers much more tangible rewards for good performance, and much harsher punishments for bad performance, than almost any school.  School teaches students the wrong life lessons: Excellence doesn't lead to money or status, and disruptiveness won't get you fired.

Even worse, school often indirectly inculcates counter-productive character traits.  Students spend a lot of their energy trying to show their fellow students that they're defiant, cool, etc.
2. Education is a consumption good. This should be self explanatory. At the margin school may be work, but infra-marginally at least some (if not most) people actually enjoy the reading, the lectures, the homework, etc.
It's a consumption good for some people; it's a consumption bad for more.  Think about all the people who continue to gripe about the boredom they felt in econ classes decades after they escaped our clutches!  And merely having consumption value>0 isn't enough to make the consumption of education socially beneficial.  When you estimate the consumption benefit, you need to subtract the opportunity cost of students' time from their gross willingness to pay.  I'm almost sure that's negative in the aggregate, especially when teachers take attendance.
3. Education is not just investment in work capital, its also an investment in consumption capital and social capital. I feel much more at home in the world due to the fact I understand certain cultural references. For example I know what someone means when they refer to someone else as a Prufrock. I also understand what someone is saying about a character in a music video if their costume invokes the evil robot Maria from Metropolis. I learned those things in college. The shared culture produced by the education experience expands our common language with a lot of meaning, and that produces huge network externalities. Knowing history does help me do my job, but it is much more important that it allows me to make analogies that will be understood by acquaintances. What good does it do to talk about Vietnam Syndrome with those who didn't live through that era if they don't know anything about the Vietnam war and its effects on our politics? What good does it do to denigrate McCarthyism if people don't know what that is? Obviously this list could go on and on. "The original position," "fair game," "zero-sum game," "categorical imperative" etc. etc. All very commonly used expressions whose meanings would probably be completely lost on someone who had three years of trade school after grade school and then went straight to work. But anyone who attended college has probably been exposed to those ideas through conversation if not through attending classes.
On consumption capital, see my reply to #2.  On social capital, I'd say that the world of work also produces social capital with network externalities.  And these network externalities are a lot more useful for society than the ones that you learn in school.  Frankly, most of your examples just seem like consumption concepts that siphon nerds like us away from the productive sector of the economy.  What's "the original position" ever going to do for the world?
4. Some classes are very very valuable at work. Reading, writing and numeracy are all obviously important. Most people may learn those things in grade school, but a lot of people are still advancing in HS. If you do powerpoint presentations for English you are learning skills you may very well use later on the job.
I largely agree.  But if you look at a typical curriculum, reading, writing, and numeracy get a lot less time than you'd think.  And much of that time is poorly spent on e.g. Shakespeare - learning how not to write in the modern world.  Still, reading, writing, and numeracy belong in the productive 20% of education.
Math courses have big returns even controlling for IQ (I believe) and that would seem to indicate that they have value preparing people for a wide range of work. Its not just engineers who use math on the job these days. Many blue collar workers are programming numerically controlled machine tools and have to understand statistical quality control.
The return to math could still easily be signaling, of course.  And I think that most of it is.  Most college graduates do not use higher mathematics on the job, and never will.  When my wife was in law school, the students practically rioted when their tax professor tried using algebra!

Bottom line, Bill: Your first instinct was correct.  What students learn in the classroom appears to have little real-world application - and appearances are revealing. :-)

P.S. If anyone naturally overestimates the on-the-job usefulness of education, it's academics like us.  After all, by definition a big part of our job is to teach students the material we learned when we were in their shoes.

Comments and Sharing

COMMENTS (22 to date)
Dan Hill writes:

Dickens seems to totally ignore the concept of opportunity costs. You can't just argue whether there is substantive value in spending four years in college, but you must also ask does it deliver superior value to the alternative uses of that time, money and effort?

As for his argument about common references, does he really think the typically college graduate could define "categorical imperative"? Seriously?

Curt writes:

To what extent is Bryan specifically referring to post-high school education? Some of the points indicate that there is some difference between high school and college (i.e. "Rather education is mainly about practicing the sort of self-discipline that is necessary to be productive in a modern work environment. High school allows you to practice showing up on time and doing what you are told.")

I tend to think that high school does provide some skills that are of specific value in many workplaces, while college can indeed tend more toward signaling.

Chris T writes:

I find the idea that you should pay thousands of dollars to learn about a subject for fun when you can do it for free using the largest information repository ever created to be very puzzling. The only rational reason to pay large sums for a certificate is in the expectation that it will lead to greater economic rewards.

You don't need a university to fulfill a love of learning.

Telnar writes:

The extent to which university attendance is required by the job market as a signal varies widely by field. For example, skill at developing software is relatively measurable and the software industry has an above average tolerance for weirdness, so it's not that difficult for skilled high school graduates to get a job which will eventually lead to a successful career.

Colin k writes:


It does work well if you want to find single, socioeconomically complementary people of a compatible sexual orientation with whom to discuss Fitzgerald, Hemingway, Foucault, or, god help you, Annie proulx. Especially if you hope to get to second base, or beyond.

I suspect higher education would look very different if it didn't market primarily to early-mid twentysomethings.

Norman Maynard writes:

"It's a consumption good for some people; it's a consumption bad for more."

Relative to sitting around smoking pot, I agree. But as you emphasize in your previous point, the correct comparison is to having a job. And while people regularly lament specific classes, I think it's very rare to find someone who doesn't occasionally pine after the leisure time and ease of their college days after they've entered the workforce.

GU writes:
You don't need a university to fulfill a love of learning.

This is a bit of an overstatement. If you'll allow for some inductive reasoning, I will note that in areas where I engaged in self-study, and in which I later took a university course, I usually found the professor added valuable things to my education that would be difficult to replicate by using the Internet (and my own intellect) alone. I suppose if I didn't know what I was missing, I wouldn't necessary feel unfulfilled, but I don't think that pure hedonism is the goal for lovers of learning (presumably accurate and deep knowledge count for something).

But it is true that the Internet (or a good library) can contribute a decent portion of a liberal education.

blink writes:

I have two questions about your very thoughtful post:
1) Do you see different mixes of signaling across countries/educational systems and, if so, what accounts for the different equilibria?
2) If you're right and could eliminate signaling from education, then what? Wouldn't this merely push the game back one level, possibly with bad consequences?

Troy Camplin writes:

Allow me to take issue with the reading novels and poetry slam. If taught properly, literature classes actually teach you how to understand more clearly what you read, to empathize with others by allowing you entry into another's world, and to think more complexly.

More, reading literature contributes to the conditions needed for the creation of the economic spontaneous order, as I point out here:

People who read and understand literature are more complex thinkers. Why wouldn't an employer, especially in the modern economy, not want more complex thinkers?

Tracy W writes:

Suppose you open a school that supposedly accomplishes in two years what regular schools do in four.

In New Zealand in most subjects it's possible to skip the first year of your university degree, based on good enough exam results at the end of high school, so you can do an honours degree in 3 years, not 4. That's what I did and I know quite a few other people who did. It's normally regarded as a plus.

eccdogg writes:

I think a lot of the signalling to capital formation depends on the university and course of study.

I went to a land grant university , the original mission of those schools was to create engineers and farmers so the focus was much more based on forming productive skills. This philosophy also bled into non science/technical fields that were later added.

That experience really colors my view on this subject as I think at least half of what I did there was ultimately skill based and that seemed to be the case with many of my classmates who were engineers, farmers, food scientist, accountants, teachers etc.

Hyena writes:

But do academics like you have enough "on the job" experience to really qualify your claims that work also builds these traits?

For all your digging on Shakespeare, fine arts are probably a better course of study than people believe. Probably better than undergraduate sciences. People who work in the visual or literary arts learn to solve design problems using research and experimentation. You then have those solutions validated or destroyed by peers.

I suspect that if any person completed more than a third of their required hours from a literally creative/constructive area of study, that they are a more valuable employee and it shows in their pay after 5 years at work.

floccina writes:

My view is the testing/signaling component of education squeezes out education in schools. So were schooling is mostly signaling education is valuable. The basic principles of chemistry, biology, physics, accounting are simple but in order to make the signal valuable we make those subjects hard on purpose loosing most of the population.

floccina writes:

BTW another point most education is now avaible free, but the signal costs a lot!

Two Things writes:

"On social capital, I'd say that the world of work also produces social capital with network externalities. And these network externalities are a lot more useful for society than the ones that you learn in school."

Though I mostly agree with you I think you're wrong about this point. In the world of work most people (apart from salesmen) meet only fellow employees of one firm. (Academics may meet more people from outside their departments or institutions.)

I don't know whether camaraderie among co-workers is especially useful for society, but I think single-firm social capital is is less useful for individuals than multi-firm social capital.

Labor mobility, especially upward mobility, depends on contacts which many people only make at school.

Seth writes:

Does the network and affiliation have any value?

Mark Little writes:

This is a very good post and I think it contains a great deal of truth. However, two additional points deserve emphasis.

While signaling may explain much of the value of the typical undergraduate education, many professions do require genuine higher education. If one wants to be an engineer, an MD--or an economist--attending a university provides essential value beyond signaling. There is great heterogeneity in students and in the services provided by educational institutions, and the signaling model would be stronger with better differentiation of the cases to which it is to be applied.

As to what is being signaled, a missing factor in the analysis is social class. A college diploma provides social status, and traditionally is the means of entry into the professional classes. The costs of college thus can serve as a barrier to entry similar to occupational licensing, but with additional social benefits beyond increasing the wage level for the exclusive group.

A diploma is no longer as much of a barrier to entry today because higher education has been democrotized, but the perception that education equals social and economic advancement remains powerful. I suspect that it is this promise of advancement in social status which represents much of the perceived value that the education sector is selling. Rather than enabling students to signal personal qualities, status signals may be more important.

I would like to see Bryan address this angle in the education signaling model.

Russell writes:

A good example of education being about signaling is looking at what courses people take who don't care about signaling.

For example, at Chicago Community colleges Hispanic immigrants will often enroll just to use the computer labs and use the Rosetta Stone software to help them learn English.

If signaling is so important, reform can only happen at the bottom of the educational hierarchy.

Roger Sweeny writes:

2. Education may not be a consumption good for many but "going to college" is. Being on your own in a relatively safe and free environment, with lots of people your own age to do things with--from watching niche movies together to having sex. Lots of people really like that. Many like it so much that in order to get a lot of it, they will read and study less than their professors think is optimal.

sconzey writes:

But let's not forget the most important role of public schooling: indoctrinating students in the awesomeness of democracy and public schooling! </tinfoil_hat>

No, seriously, as a recent CS graduate (now employed :D) I have experienced first hand the odd mix of the useful and useless on University courses. Great: learning new languages, software management techniques, BIG projects with no sub-deadlines forcing you to do time-management :P.

However, we also had to take courses on "professional issues" which instructed us on important things like: "You shouldn't lie to your employer." Likewise, although not so obviously, the study of obsolete or obscure languages, techniques and non-problems in gratuitous mathematical detail. Clearly signalling.

At interviews what they wanted to know wasn't my module marks or GPA, but about my projects: what did I do, how did I go about achieving it.

I came to a Navrozov moment, when I realised that learning to be a good software developer had only been an externality of my course. The signalling thesis is as good as any. I'd argue that students with loans express an expectation about their future earnings which companies trust.

I maintain CS is a Craft not a Science, like woodworking or metalworking. :P Blue-collar not white. :D

Roger Sweeny writes:

3. This sounds like E.D. Hirsch's "cultural literacy." But most people in higher education consider Hirsch kind of a crank and don't take him seriously. Over the last half century, there was been a strong movement away from courses that all students must take and ideas that all students must be exposed to.

Seriously, can you imagine a faculty senate anywhere passing a resolution saying, "Beginning with the class of 2015 no one shall be eligible for a diploma until they demonstrate a clear understanding of J. Alfred Prufrock, the graphics of Fritz Lang's Metropolis, Vietnam Syndrome, McCarthyism, the original position, fair game, zero-sum game, and categorical imperative"?

Ivin Rhyne writes:

Dr. Caplan,

After reading this, along with others in the series of posts on the signaling value of education, I would like to propose a thought experiment that could be fleshed out into a formal model rather easily, but doesn’t require it.

Let’s imagine a country called “CapLand” where there are four universities that each teach the same four subjects. These subjects are important to the businesses of CapLand who each year compete for the best and brightest among the graduates. In the country of CapLand, each business is competitive and willing to pay a wage based on the contribution of each employee (otherwise known as the marginal product of labor). This means those employees who are most productive pull down the biggest salaries, and those who are best in each of the subject are most productive.

Education in CapLand has one severe limitation. It fails to add any real knowledge or competency beyond a threshold level that is determined by innate ability. In other words, it doesn’t matter WHICH school you go to from a productivity point of view since your innate ability will determine your productivity. And not everyone in CapLand has the same level of ability. In fact, let’s assume that the abilities in the “Big 4” subjects are distributed in a bell curve distribution that puts most people near the mean and relatively few at the very high end.

So here is the thought experiment – under these conditions, will all of the universities have equal tuition and equal stature in the eyes of the labor market (both students and potential employers) since none of them can really add value over the other? The answer should be an obvious NO.

If we were to launch all four of CapLand’s universities simultaneously, they would undergo a fiercely competitive period to get the teachers and scholars of highest repute. Eventually one school would emerge as the “Ivy League” of CapLand, charging higher tuitions than the others and generally producing more CEO’s than the other three. Why? Because despite the fact that no university can add productivity beyond innate ability, each school is providing a valuable service to the marketplace: they are sorting the labor pool along their abilities and in doing so adding information to the marketplace that employers find valuable.

Think of it from the employer’s point of view. Without the stratification of schools, they would have to carefully comb over each of the four schools to find the best and brightest. Instead, they can focus their efforts on the “Ivy League” school with a high degree of confidence that they are getting candidates of higher productivity. As a result, employers are willing to pay these Ivy League graduates higher starting salaries, which in turn allows the school to charge higher tuition.

It’s from here that the “Ivy League” school begins to spread the fallacy that they are somehow responsible for the productivity of their students. This is because the narrative is powerful and simple – good schools produce good graduates. The implied cause and effect relationship is alluring and most people fall for it. Instead, at least in CapLand, good graduates are what make good schools.

A fairly mathematical version of this argument can be made, although I doubt its value since such a model tends to hide the reasoning behind the mechanics of the formulas. Instead, I think from here you may have a natural launching point for your signaling model of education being integratable with the concept that IQ (or innate ability) matters as well.


Ivin Rhyne

Comments for this entry have been closed
Return to top