Bryan Caplan  

The Magic of Education

PRINT
Kevin Drum vs. Priceless Econo... Bryan Caplan is Not Like Warre...
I've been in school for the last 35 years - 21 years as a student, the rest as a professor.  As a result, the Real World is almost completely foreign to me.  I don't know how to do much of anything. While I had a few menial jobs in my teens, my first-hand knowledge of the world of work beyond the ivory tower is roughly zero.

I'm not alone.  Most professors' experience is almost as narrow as mine.  If you want to succeed in academia, the Real World is a distraction.  I have a dream job for life because I excelled in my coursework year after year, won admission to prestigious schools, and published a couple dozen articles for other professors to read.  That's what it takes - and that's all it takes.

Considering how studiously I've ignored the Real World, you might think that the Real World would return the favor by ignoring me.  But it doesn't!  I've influenced the Real World careers of thousands of students.  How?  With grades.  At the end of every semester, I test my students to see how well they understand my courses, and grade them from A to F.  Other professors do the same.  And remarkably, employers care about our ivory tower judgments.  Students with lots of A's finish and get pleasant, high-paid jobs.  Students with a lots of F's don't finish and get unpleasant, low-paid jobs.  If that.

Why do employers care about grades and diplomas?  The "obvious" story, to most people, is that professors teach their students skills they'll eventually use on the job.  Low grades, no diploma, few skills. 

This story isn't entirely wrong; literacy and numeracy are a big deal.  But the "obvious" story is far from complete.  Think about all the time students spend studying history, art, music, foreign languages, poetry, and mathematical proofs.  What you learn in most classes is, in all honesty, useless in the vast majority of occupations.  This is hardly surprising when you remember how little professors like me know about the Real World.  How can I possibly improve my students' ability to do a vast array of jobs that I don't know how to do myself?  It would be nothing short of magic.  I'd have to be Merlin, Gandalf, or Dumbledore to complete the ritual:

Step 1: I open my mouth and talk about academic topics like externalities of population, or the effect of education on policy preferences

Step 2: The students learn the material. 

Step 3: Magic. 

Step 4: My students become slightly better bankers, salesmen, managers, etc. 

Yes, I can train graduate students to become professors.  No magic there; I'm teaching them the one job I know.  But what about my thousands of students who won't become economics professors?  I can't teach what I don't know, and I don't know how to do the jobs they're going to have.  Few professors do.

Many educators sooth their consciences by insisting that "I teach my students how to think, not what to think."  But this platitude goes against a hundred years of educational psychology.  Education is very narrow; students learn the material you specifically teach them... if you're lucky. 

Other educators claim they're teaching good work habits.  But especially at the college level, this doesn't pass the laugh test.  How many jobs tolerate a 50% attendance rate - or let you skate by with twelve hours of work a week?  School probably builds character relative to playing videogames.  But it's hard to see how school could build character relative to a full-time job in the Real World.

At this point, you may be thinking: If professors don't teach a lot of job skills, don't teach their students how to think, and don't instill constructive work habits, why do employers so heavily reward educational success?  The best answer comes straight out of the ivory tower itself.  It's called the signaling model of education - the subject of my book in progress, The Case Against Education

According to the signaling model, employers reward educational success because of what it shows ("signals") about the student.  Good students tend to be smart, hard-working, and conformist - three crucial traits for almost any job.  When a student excels in school, then, employers correctly infer that he's likely to be a good worker.  What precisely did he study?  What did he learn how to do?  Mere details.  As long as you were a good student, employers surmise that you'll quickly learn what you need to know on the job.

In the signaling story, what matters is how much education you have compared to competing workers.  When education levels rise, employers respond with higher standards; when education levels fall, employers respond with lower standards.  We're on a treadmill.  If voters took this idea seriously, my close friends and I could easily lose our jobs.  As a professor, it is in my interest for the public to continue to believe in the magic of education: To imagine that the ivory tower transforms student lead into worker gold. 

My conscience, however, urges me to blow the whistle on the system anyway.  Education is not magic.  Professors can't make students better at whatever job awaits them with learned lectures on arcane topics.  I'm glad I have a dream job for life.  I worked hard for it.  But society would be better off if taxpayers saved their money, students spent fewer years in school, and sheltered academics like me finally entered the Real World and found a real job.
 

Comments and Sharing





COMMENTS (79 to date)
foosion writes:

School is part signaling function, part training, part learning our shared culture and history and part socialization (and some other functions I'm no doubt forgetting).

Ignoring any of these functions leads to faulty conclusions, such as school being solely signaling or solely training.

iamreddave writes:

Great post

"Good students tend to be smart, hard-working, and conformist - three crucial traits for almost any job."
Does the signalling theory explain why a student who went to an ivy league university would be ranked higher than one who went to a less prestigious one? The conformity should be the same in both unless Harvard is obviously more regimented. Smart could be tested with an IQ test much more quickly and cheaply then years of school.

What role does prestige play in the signalling theory of education?

londenio writes:

Good article! Very good point about signaling "conformism" apart from ability and hard work. There is no way to make it in the corporate world, consulting or banking if you are not a conformist and do not play by the rules.

Like in most signaling stories of education, these insights are more or less applicable in different "majors". For instance, in Chemical Engineering or Accounting, what you learn is often useful for what you will work on. I guess that is also true for many professional schools (Law, Business, ...) but then those schools are built to mimic the Real World. Like a toy version of the job you'll have.

Joshua writes:

I'd guess the signalling model explanation of school prestige is that the prestigious institutions are harder to get into, so a student's attendance there signals meeting that school's more rigorous admissions standards.

RPLong writes:

The modern view of education pretty much comes from the G.I. Bill after WWII. Before that, nobody "needed" an education to get work in The Real World. College existed pretty much to serve academia. Academia performs a certain niche market function, and that's how it was.

The government's massive education subsidy produced a glut of college-educated members of the workforce, who won out on jobs over their less-educated - but equally skilled - competitors. This gave the less-educated additional incentive to go to college, which they did.

After a few iterations of this, we entered a parallel universe in which college education is "necessary for finding a good job." It doesn't have to be this way, and we got along just fine without it. The only reason the system self-perpetuates this nonsense is because it's good for the rent-seekers (university faculty/staff).

I have taught myself a variety of computer programming languages, software applications, and technical skills on-the-job. There was never any reason for me to have to attend a university to develop those skills. The whole thing is sham.

IMHO.

rkw writes:

Writing books is a real job.

Joe writes:

Bryan: "It would be better for society if I were fired, but until then I'm gonna keep my dream job."
Commenters: "Hear, Hear"

Warren Buffett: "It would be better for society if I were taxed more, but until then I'm gonna keep my money."
Commenters: "You should just pay more yourself."

perfectlyGoodInk writes:

Well said. And this argues strongly against anybody going straight into grad school from undergrad, and in favor of the system to weight non-academic letters of recommendation equally with academic ones.

Jeff B writes:

@RPLong:

I have taught myself a variety of computer programming languages, software applications, and technical skills on-the-job. There was never any reason for me to have to attend a university to develop those skills. The whole thing is sham.

While I too have self-taught quite a bit in regard to software engineering, I would say that while I don't consider a 4 year university absolutely necessary, the work and best practices you pick up by taking a broad range of courses versus just trying to get through it yourself, is something I have quite a hard time calling a "sham"

Perhaps quite a few (if not most) degrees are, but calling highly technical ones a sham seem pretty short sighted, especially given the extremely large breadth of areas one can go into in regard our example - computing. Would you also say backgrounds in chemical, biological, electrical and medical engineering fields are a sham? I have a hard time believing I'd prefer a self-taught chemist over one that went to a university.

Back to computing, I imagine most people "self teach" with PHP these days, but I also imagine that they don't necessarily learn good engineering practices by doing that. Alternatively, they do it much more slowly than those who are trained in some way.

perfectlyGoodInk writes:

londenio: Like in most signaling stories of education, these insights are more or less applicable in different "majors". For instance, in Chemical Engineering or Accounting, what you learn is often useful for what you will work on.

As someone whose undergrad was in computer science and spent fourteen years as a software engineer (before diving into economics), what I learned in school was probably a little more useful than most.

The weed-out concept was recursion (a program calling itself), which we never used in industry because it uses up a lot of stack space compared to iterative approaches. Almost everything I needed to know I learned in the classes that taught C and assembly language. Everything else was pretty useless, and the thing that taught me the most about being a good software engineer was a book I read after graduating, The Practice of Programming by Kernighan & Pike.

Software engineers also typically make horrible interviewers, and a resume with a good school on it gets you far. It's totally signalling.

joeftansey writes:

"Good students tend to be smart, hard-working, and conformist - three crucial traits for almost any job."

This is dodging the issue. Can professors make their students smarter, harder working, and more conformist? Or are they like that to begin with?

I believe in the latter, but this seriously needs proof on all counts. Its not enough to show that you can't teach people how to think, you also have to show that you can't make them more conformist.

But another benefit of this topic is that it informs students about whats going on. It allows them to rationalize a liberal arts degree that might be causing them anxiety, or lets them choose extra curricular activities in an attempt to send even stronger signals of social conformity to employers.

Though I do think there's a case to be made that liberal arts majors signal laziness, lack of ambition, and self absorption... Case in point - not every job accepts every bachelor degree. In fact, most of the good ones don't.

Becky Hargrove writes:

Bryan, it is in part because I believe so much in the magic of education, that I wrote this yesterday in a fit of frustration: "the best place for action still remains in the real world of people who interact with one another at local levels." I'd like to contrast what I said with your comment, "If you want to succeed in academia, the Real World is a distraction." For me the contrast is a problem in that it splits up the way people relate to one another in sometimes irretrievable ways - not to mention the split that can occur in one's psyche. But how do we find balance between the life of the mind and that of one's more physical energies? That's a process that has yet to begin.

I care deeply about the kind of work that you do. Again, my frustration relates to what I believe is one of humankind's most important aspirations once basic needs are met: everyone needs to find a place in their lives where they can finally teach or heal others. I dream of knowledge integration - person to person - because I believe everyone needs that just as much as myself. And this is especially important now because of the solutions that need to be found in the present. Such solutions cannot be imposed by the few, they have to be participated in by the many. The knowledge of all is vital to rebuild the bridges between one another that continue to be burned. Lets build those bridges while there is still time.

[typo in name fixed--Econlib Ed.]

Gepap writes:

Please...

There are no "real" jobs. Liberterian economists need to be sent on mandatory anthropological sabbaticals to study Amazonian or Papua New Guinean tribes - those are the folks living the "real" human existance as good ol' mother nature would intend. Any society more complex is a human artifice, and any attempts to claim that any profession in those artificial communities is more "real" is nonsense.

Besides, I though you believed in rational market actors. Clearly signalling is critical otherwise why would so many rationally behaving economic actors pay so dearly for it? What makes you the great central planner know-it-all?

Tom West writes:

The only reason the system self-perpetuates this nonsense is because it's good for the rent-seekers

That's complete nonsense. Companies use the signal because it's reasonably useful. (Sure you lose a few good potential employees, but the filter only has to be better than nothing to make it worth employing in most cases.)

Why should the company care if the signal costs the potential employee immense amounts of time and resources?

This is a simple market failure. The only question is whether it's a costly enough failure that government should do something about it. I'm going to bet the answer is a big fat no.

So relax Bryan, I fully expect just as one needs a BA to get a job in the mail-room in most larger companies now-a-days, you'll need your Master's degree in another 10 or so years.

Yancey Ward writes:

You can measure intelligence and social conformity in an afternoon, and do so rigorously and cheaply. However, we have chosen to farm this process out to colleges who spend four or more years doing the same thing, and probably not nearly as well. We really would do well to change this, but I am not hopeful.

Thomas DeMeo writes:

Why do you stay if you feel this way?

Jack writes:

Good argument, but I think it overlooks two key aspects:

(1) Not all college degrees are pure or nearly-pure signalling. Ask my engineering brother if he could do the same work without the courses he took. But could teaching be more efficient, more relevant? Sure.

(2) Prof. Caplan ignores professors' research, which is (depending where you teach) 20-80% of your time. Some believe all university research could be instead done in the private market, but that is naive. (To be sure, much research can be and is conducted by the private market.) Moreover, complaining as some politicians do about foolish-sounding sociology research topics is a straw man: the humanities and qualitative social sciences receive little funding.

We could also talk about the demand-side of education: why do students willingly pay a fortune to be educated when they can get a good, inexpensive college education? Why do they take 5 years or longer to finish? These questions are as important, if not more so, than the supply-side of higher ed. Thus, overall, good argument by Prof. Caplan, but an oversimplification of the situation.

Randy writes:

Re; "Students with lots of A's finish and get pleasant, high-paid jobs. Students with a lots of F's don't finish and get unpleasant, low-paid jobs."

I agree with the pay, but not the idea that high paid jobs are pleasant or that low paid jobs are unpleasant. In fact, in my 40 some years of real world work experience it is usually the other way around. High pay is nice, but it often comes with a high cost. The high paid people I work with are all constantly stressed out.

joeftansey writes:

Isn't one of the biggest libertarian complaints about public schools is that they turn kids into conformists?

Neal writes:

Mind, sheltered academics like you have enough intelligence, political skill, and show-up-to-it-iveness to greatly succeed in the "Real World." You'd probably also be better paid. I'm not seeing how this is a loss for you :)

Joe T. writes:
"... insisting that "I teach my students how to think, not what to think." But this platitude goes against a hundred years of educational psychology."

While I didn't read the referenced book, I'd daresay that theoretical math teaches one to think. It's all proofs. By my junior year of undergrad, I never made another reasoning mistake. Incompletes, yes, where I couldn't finish the proof. But nevermore any false reasoning.

Matt writes:

So what exactly are you calling for? The death of the university? Fewer government subsidies? I doubt these are just casual observations. What else did you have in mind?

RPLong writes:

JeffB,

Perhaps quite a few (if not most) degrees are, but calling highly technical ones a sham seem pretty short sighted, especially given the extremely large breadth of areas one can go into in regard our example - computing. Would you also say backgrounds in chemical, biological, electrical and medical engineering fields are a sham? I have a hard time believing I'd prefer a self-taught chemist over one that went to a university.

Oops - Good point! I mentioned something similar in a comment to Bryan's last post on this topic.

I carve it into two concepts: One is, "Is education a signalling or human capital investment?" which I answer with, "Both, but it depends on the course of study."

The other concept is, "Is university education necessary?" which I answer with, "In practice, yes; in reality, we could probably satisfy the same need more efficiently by promoting greater use of an apprenticeship model."

My last comment only addressed this second point. All that is to say, I agree with you! ;)

adam writes:

How do you explain Oreopoulos AER 2006 with your signalling model? That is my litmus test--when everyone gets more education nothing should happen, but instead the UK experience is that an extra year of compulsory education (which affected a majority of UK students) made them more productive.

http://homes.chass.utoronto.ca/~oreo/research/compositions/Estimating_Average_and_Local_Average_Treatment_Effects_of_Education_When_Compulsory_Schooling_Laws_Really_Matter.pdf


More generally, how do you reconcile this idea with the literature on teacher value added? How is it possible for some [kindergarten/elementary school] teachers to regularly produce students who, as adults, earn more than students of other teachers? This evidence clearly indicates that teachers provide some sort of skills that are valuable in the labor market.

stephen j smith writes:

I couldn't agree more, except with technical degrees. And with technical degress, the firms could just give IQ test, then apprenticeships, which many parents might be better off b/c they would forgo educational costs.

David Worley writes:

Most of what I would be critical of in this post has already been commented on. I particularly agree with "Jack @ 10.56A" regarding key things Kaplan missed.

That said, it seems to me that the whole economic calculation about the monetary value of the educational industry misses the point. Certainly their are more economically efficient ways to produce workers; this isn't really the question. I think a better question would be "in what way does education serve society?" In posing this alternative inquiry I think we might conclude that the University (even the economically dispensable 'humanities') is a useful human endeavor. Furthermore, this discussion seems to reveal that the point-of-view of many economists and Libertarians is too myopic and frankly a little depressing.

Luke G. writes:

One of the questions I hope you address in your book, Bryan, is how educators like you and me should respond if we assume the signaling model education is correct.

I’m an English professor at a small, private Midwest liberal arts college. I teach mostly general ed requirements in composition and literature surveys. I used to buy in to the idea that I was teaching my students “how to think” and rounding them out as human beings. The signaling model, however, erodes that considerably. So, how should I teach? If I’m still teaching the same general ed courses, is there a way to help my students become more intelligent, more conformed, more hard-working?

Miguel Madeira writes:

I think that people with degrees in antrophology has high unemployment than people with degrees in computer engineering. How you concoloate that with the signaling model?

Richard Belloff, DBA writes:

Brillant! As a professor who spent over 35 years in the "real world" before entering academia, you are dead on. I encourage you to complete your book!

Troy Camplin writes:

IF it were but signalling, my Ph.D. would signal that I am an excellent worker, extremely smart, and a conformist's conformist -- and I would have a very high-paying job by now. Instead, I am unemployed. How does one account for that? Could it be because my degree is in the Humanities? If so, content does matter. A lot.

English Professor writes:

A little history might help. Until the 19th century, only liberal arts degrees were thought of as real education; professional degrees (esp. law and medicine) were thought of as learning a trade. The things that Bryan now praises--practical utility from schooling--were looked down on. Most liberal arts education was carried on in Latin, and one read works in Latin and Greek. An education of this sort prepared you for one of two things--to be a gentleman (i.e., probably a land owner, without regular employment) or a member of the clergy. This sort of education was a marker of one's social class. Gentlemen could thus talk to one another with the confidence that their interlocutors were "educated" men. (Women, of course, were largely excluded from this sort of education.)

In 19th-century England, men like Matthew Arnold thought that the ideal situation would be to educate the children of all working-class families in the same manner as gentlemen. This was intended to improve their lives by giving them access to the great ideas of preceding generations, and it would also be a major step in increasing social equality across Britain. Over time this required a dumbing down of the academic curriculum. The Arnoldian vision of expanding a young person's intellectual horizons provides the basic underpinning for all the modern claims that education raises human capital. For some people, it provides an enormous opening up of the intellectual world; for others (perhaps the majority?) it provides little more than a superficial introduction to serious thought.

There was also a counter-movement in the late nineteenth century when education reformers began to argue that education should aim at utility. At this point, practical majors like engineering and agronomy became common in college. Law schools and medical schools also gained greatly in prestige, as did all other sorts of postgraduate education.

So even from a historical perspective, Bryan is basically right: liberal education always included a large component of signaling. A man with a university degree was acknowledged to have achieved a certain social status and could be counted on to be a gentleman. But the 19th-century reformers were equally adamant that a liberal education was useless for those who would simply hold practical jobs. I suspect that the liberal arts have continued to survive because many working-class people (like my own parents) still see this sort of education as marking a step up in social class. And for parents who already have a college education, they want their children to do the same so that they won't slip in their class status.

Jonathan writes:

@Troy Camplin:

Your problem is that you have over-signalled. yes, you're extremely smart, a good worker and a conformist's conformists, but you have ruined it all by also indicating that you are actually interested in a life of the mind -- that undercuts the other skills except in academia. If only you'd stopped at a BA....

Daniel Dostal writes:

Bryan, read your comments. If one could correlate the misinterpretations of your post with college success, I'm confident you would find a strong inverse relationship. I am a software engineer and the tools and skills that I use day-to-day have generally been learned on the job. However, without college my ability to send thoughtful and articulate emails would be drastically reduced. My ability to communicate with other highly intelligent folks would be greatly hampered as university is the only place where I have been able to hone that skill. Interpersonal skills are not well learned on the job, as follies lead to write-ups and being fired, a signal that surely injures any worker. College is a relatively safe place to learn these and many other relevant skills. My peers and I learned so much more in college than could ever be gleaned in the workplace. I've worked with several self-taught engineers and they really do suffer because they lack all these undervalued skills. Sadly their thinking often blinds them to their ineptitude as well. Certainly education could be better tailored, but you're suggesting throwing out the baby with the bath water. Also, remember that the business guys are the flip-side of your "we can't teach what we do not know". My peers and I explored new concepts in college that the industry wouldn't pick up without us suggesting them. Also, businesses may complain that universities do not train the right skills, but that's because businesses do not want to invest in their employees what could be learned elsewhere. Universities will never teach enough for businesses.

Dan Weber writes:

I believe I actually learned things from my Computer Science (or Computer Engineering? Or Electrical Science? Oh, whatever) courses. I can certainly tell when I'm dealing an algorithm that someone wrote that takes O(n^3) time that they didn't get a good education in this. I'm sure the name-brand of my school is also very very valuable.

Software development is still a field very friendly to the self-taught, which I think is great for both them and society.


How is it possible for some [kindergarten/elementary school] teachers to regularly produce students who, as adults, earn more than students of other teachers?

Bryan looked back on this study earlier (search for October 14, 2011). It was vastly overstated.

English Professor writes:

To those who wonder why the signal is failing for many liberal arts degrees, I would suggest two things. First, there is now a superabundance of people with those degrees. The supply currently exceeds the demand. (This is certainly true of the supply of PhDs in almost all the humanistic fields.) Second, many liberal arts degrees are not rigorous. As this has become known, employers have concluded that they no longer signal the ability to think rigorously or to write clearly.

PrometheeFeu writes:

"At the end of every semester, I test my students to see how well they understand my courses, and grade them from A to F. Other professors do the same. And remarkably, employers care about our ivory tower judgments. Students with lots of A's finish and get pleasant, high-paid jobs. Students with a lots of F's don't finish and get unpleasant, low-paid jobs. If that."

Failing students out of college may change their outcomes in the real world, but I doubt your grades otherwise make much of a difference. Out of the people in my cohort with whom I have kept in touch, there seems to be little correlation between classroom achievement and current professional success outside of government jobs and graduate school. (That's about 4 years after graduation) I had terrible grades but am also one of the most successful having landed a stable intellectually stimulating job with substantial income. Nobody else seems to care.

@Miguel Madeira:

Perhaps because your interests influence both your major and your career. I didn't major in CS, but I'm a software engineer and get the corresponding salary. On the other hand, my CS friend who didn't want anything to do with tech professionally does just as badly as the history majors.

Bill Hocter writes:

Bryan-I can recall probably 10 or 12 things that my teachers said during 27 years of formal education that affected the trajectory of my life and made me a better physician, naval officer, husband and father. The accumulated time to say those things was brief, but the years of schooling prepared me to appreciate what they said.

A broad education is a valuable tool for those who can reflect upon it. Translating parts of Homer's Odyssey in high school and college caused me to reflect upon the value of family and perseverance, among other things. Taking Xenophon's Anabasis to Iraq was a most poignant experience. Learning something about economics helped me to be a better psychiatrist-the concept of opportunity cost has great applicability to psychotherapy.

I'm most grateful to the men and women who have taught me over the years, many of whom had never held a "real job" in their lives. I've told them so when I've had the opportunity. I've learned a lot reading Econlog as well. I suspect some of your students have benefitted more than you realize. Cheers.

Tom West writes:

Yancey Ward:

You can measure intelligence and social conformity in an afternoon, and do so rigorously and cheaply.

That would be approximately one afternoon minus ten seconds more effort than one needs to scan for a BA/BSc. And again, if saving $50 worth of testing costs the prospective employee $100,000 and 4 years, it's not your problem.

Anything that cuts 500 essentially identical resumes to 200 essentially identical resumes and doesn't actually *reduce* the good employee to resume ratio is worth-while for most businesses.

Mik writes:

(1) There are plenty of papers out there, published in fine journals, showing benefical effects of rolling out more public education overall (not only the AER paper); which would speak against that all that happens is where you end up relatively.

(2) Several papers have also shown that increasing resources for schooling have significant and broad non-cognitive returns. All those effects are harder to explain in Caplans model of schooling as pure signalling.

(3) There are several papers showing significant learning and wage returns due to increased resources for schooling such as teacher value added, smaller class size in early grades (yes, the latter clearly if you take a neutral read of the literature and base it on the most recent high-quality studies). I am ending by going all unscientific here, but that all these results would come from pure signalling and changes in the relative outcome does not pass the laughter test and requires a lot more of strong assumptions than the other hypothesis; that schooling actually does teach something valuable..

On the other hand, my view is that the marginal return of research funds is clearly lower and a bad deal for most developed countries. I think it would be a significant Hicks-Kaldor improvement if we re-allocated money from publicly funded research into early schooling.

akasavani writes:

One idea is to change the grading structure - for example get rid of weekly homework assignments etc where students need to spend time trying to solve problems etc that are too narrow. Also make attendance requirements even more lax. Have 2 exams total per semester, have questions in them that can be answered mostly by deductive reasoning, as opposed to having to count on pure knowledge. All open book perhaps...
This way students don't waste too much time trying to learn something they are not going to use, but also will use their heads in a rigorous/analytical way and learn skills that are more universal. So basically designing exams only loosely based on the subject matter...

Ricardo Vicente writes:

I commented on this article in my blog post at
http://ricemagic.blogspot.com/2011/11/education-as-much-more-than-just.html:

First, formal studies DO teach things related and about the real world. One learns the real world from studying at school, at least in my country (which is very far from being the best in OECD's PISA). Of course, spending one semester abroad teaches you more than reading about life abroad. Opening a car's hood and playing around with the engine helps you more if you want to be a real-world mechanic than reading a book about cylinders and pistons. Still, "indirect" learning does improve one's knowledge of the real world. Still, reading that book about engines improves, even if only a little, your knowledge of cars.
Second, Bryan Caplan's argument is based on a fundamental premise: people study to find jobs and only for that. Assuming that premise, I agree that people study way too much for the jobs they get, they study things way too unrelated to those jobs. And, keeping that premise, a huge part of studying and doing exams is not but costly signalling. Then, the conclusion of the argument is right: since what matters is the relative ability of potential workers, decreasing the "standard level" of education would save costs while still letting education work as well as a ranking mechanism aimed to allocate people into jobs.
...

See also my epilogue at
http://ricemagic.blogspot.com/2011/11/epilogue-to-previous-post.html.

[Comment elided and two separate comments combined. Please do not paste/re-post entire blog posts to EconLog. It's okay to quote an excerpt and to link to your material that is posted elsewhere. --Econlib Ed.]

Gene writes:

Whether your particular degree, from your particular school, provides the best bang for your buck is clearly in question.

But, I'm having trouble coming up anything that is more important than signalling.

Isn't accurately signalling your potential as an ally, a mate, a threat, or falsifying those same signals to your benefit, part of almost every interaction of one living creature with another.

PrometheeFeu writes:

"Other educators claim they're teaching good work habits. But especially at the college level, this doesn't pass the laugh test. How many jobs tolerate a 50% attendance rate - or let you skate by with twelve hours of work a week? School probably builds character relative to playing videogames. But it's hard to see how school could build character relative to a full-time job in the Real World."

First, don't knock video games. Spending 12 hours straight doing anything has got to teach you something about perseverance... ;-)

I think the flip side is that the real world is often much more flexible and rational than academic life. In school, you get an assignment and you must perform the assignment because you got it. That is leagues from the real world. In the real world, when I get an assignment, it's basically a collaborative exercise where my manager and I try to solve a common problem. If I think his solution is bad, I am expected to say it and push back. This is the opposite of a school environment.

PrometheeFeu writes:

@Tom West:

"That would be approximately one afternoon minus ten seconds more effort than one needs to scan for a BA/BSc. And again, if saving $50 worth of testing costs the prospective employee $100,000 and 4 years, it's not your problem.

Anything that cuts 500 essentially identical resumes to 200 essentially identical resumes and doesn't actually *reduce* the good employee to resume ratio is worth-while for most businesses."

But there are cheaper ways to do this than college. If indeed you can do an intelligence test in an afternoon for $50, there should be a company out there that administers and certifies such tests and is trusted by employers. You could put your "IQ Inc. score" on your resume. So you can toss all the resumes that don't have an IQ Inc score above a certain number.

JLA writes:

It's a bit ironic that Bryan argues so strongly for the signalling model of education - I would bet that many of his students at Mason learn pretty valuable insights in his classes that actually help in the real world.

Sam writes:

Well said. And this argues strongly against anybody going straight into grad school from undergrad, and in favor of the system to weight non-academic letters of recommendation equally with academic ones.

Actually, that's the one thing it doesn't do. The people who go straight to grad school are the trainee professors, and if you want to learn to do research and teach in a University, that's the right thing to do.

As far as letters of recommendation go, they're only even slightly useful to me if I have a way of evaluating the person writing the letter. If a well-known professor at a good university describes someone as the best student he's had in 10 years, that's worth something. The same recommendation from John Doe, owner of Doe's Hardware Store, is worth much less, unless I happen to know Mr. Doe.

Troy Camplin writes:

The responses to my query confirmed my point: the major matters. Thus, schooling per se is insufficient as signalling. There is still some expectation of content.

Of course, the "conformist's conformist" remark was ironic. If I were such, I would definitely have a job by now. I, however, do literary scholarship and philosophy from an evolutionary biological, evolutionary psychological, Austrian economics, self-organizing network perspective -- meaning I am really, within the humanities, a rebel's rebel. They hate that.

Yancey Ward writes:

Tom West,

What PrometheeFeu wrote.

College students need a less, far less expensive way to signal to prospective employers that they have a skill set, or set of characteristics that might make them a worthwhile hire. Also, even if the employer is paying for the testing, there might be value added if they can hire employees who aren't burdened with $50,000 in student loans when they leave school.

Glen S. McGhee writes:

The best thing about Spence's signaling theory is how it warns of the danger of overschooling -- over-investing in the signal, for its own sake.

When a third of all working college grads are working in jobs that do not require a degree, the level of overschooling is clearly evident. [Also worthy of consideration is how and why business organizations have come to rely so heavily upon credential screening, to the neglect of their own training and skills testing, resulting in major labor market distortions.]

But, as English Prof suggests, only history can tell us how we got into this mess. Economic sociology can illuminate some of these puzzles, including how vo-tech education in the 1910s and 1920s came to be stigmatized by the academics in the US, BUT NOT in European countries.

Even worse, the problem we have yet to address, is the gradual abandonment of the workplace as the original locus of learning -- resulting in structural misalignment and kids that have no clue about what work is, and what they could be working at. How will education be re-coupled with the workplace?

Curt Doolittle writes:

???

Employers don't look at grades. The only effect your grades have on students is when you fail them, or graduate them. It's binary. That they learn something that is useful ... even just one or two major general patterns, that they can apply in the 'real world', and have the confidence and discipline to hack away at real problems despite repeated failure, is what employers see. And they very likely have a hard time seeing that.

In general, a PhD in engineering is usually better than his less credentialed co workers. An MBA is generally better than his less credentialed co workers. However, we graduate so many MBA's that the filtering process has diminished the value of an mba.

Schools perform sorting and filtering. They condition people to work using similar ethics of cooperation. I'm not sure they perform all that much actual education. And I'm not sure much education is actually valuable in the real world.

And I think Sowell in particular would agree.

Mark Brophy writes:

Brian's article is enlightening, and it’s rare that anyone would endanger a job that he enjoys by revealing the scam. Why won’t David Copperfield show us how he does his magic tricks? Has President Obama ever had a job outside politics and academia in the Real World? If not, does that disqualify him from creating laws that affect the Real World?

I liked the article so much that I excerpted it in my blog post, The Confession of a College Professor. And, for further thoughts, see Software Publisher Proves College is Unnecessary.


ChrisA writes:

A comment to those proposing to replace colleges with an IQ/Aptitude test. There are two problems with this approach; 1) it doesn't test for ability to sustain effort over an extended period of time. and 2) For most technical and professional jobs, even a 22 or 23 year old is in the short term a ZMP (actually negative). Companies take on graduates as an investment which pays off over time. Hiring at 18 would increase the time needed for payback meaning the needed wage was below sustenance. So you need to find something for people to do between 18 and 22/23. An 18 year old is (except in rare cases) just not mature enough to be of value in today's business environment. It's sort of like cooking with a 6 year old.

Tom West writes:

PrometheeFeu & Yancey Ward

I understand that there are probably a lot of ways that things could be a lot better for the prospective employees (i.e. some sort of trusted certification, etc.) and for businesses in the long term (less taxes spent on superfluous education).

But the benefit of switching to a better system goes to the student, while the cost of having to learn a new system falls on the employer, so it isn't likely to happen.

I think this meets the strict definition of an externality (the prospective employee would gladly pay for the company's expense in using a new system, but there's simply no market to allow that to happen), but I honestly don't think anyone's got the appetite for a government intervention to knock the current system out of its sub-optimal equilibrium state.

DK writes:

That's what it takes - and that's all it takes.

Wrong. It takes all that PLUS a lot of luck. You won a lottery - that's what it takes. The other 99 who were just as meritorious as you weren't as lucky. Don't forget this simple fact when you are pushing the 'I simply excelled in my coursework year after year' fairly tale on unsuspecting readers.

Ari T writes:

"I have taught myself a variety of computer programming languages, software applications, and technical skills on-the-job. There was never any reason for me to have to attend a university to develop those skills. The whole thing is sham."

I have developed a lot of computer programming skills that "work in real-life" by myself (I started coding when I was ~11) but to be honest, there's a lot of advanced technical knowledge that is better learned in a dedicated environment. Learning differential equations, digital signal processing, complex analysis, statistical signal processing, statistical pattern recognition etc. "on the job" is going to be very difficult. Maybe not the people with IQ 150 but for the average worker that is.

Also when you learn by yourself, there's a good chance you learn some bad habits. Its like learning music by your own. It works for some people but getting a teacher is usually a very good idea. Likewise figuring out laws of physics on your own isn't a good idea either.

I think there are a lot of IT / engineering jobs where education is useless, but I also think there are a whole bunch of jobs where its not. I don't know if I'd like if my air plane was designed who picked up aerodynamics and strength calculations "on the job". Bryan Caplan made another good example: you wouldn't want someone doing your brain surgery who didn't bother with book learning. Its not the matter of having some understanding of the job, but being sure that the person knows all the necessary skills to do the job without skipping over some "boring stuff". Just being able to do a job, does not mean you do it well (a lot of programmer jokes and frustrations have been developed around this).

For jobs that quality matters, it is probably efficient for a 3rd party to check your understanding generally of your field to avoid any conflicts of interest. Its neither the employer's and employee's comparative advantage to specialize in such QA. Not for all jobs, bigger companies can afford own QA, but smaller ones are better at outsourcing the QA to institutions that specialize in that.

Caplan's signalling model of education sounds highly plausible though.

Will Wilkinson writes:

I truly would like to hear Bryan's explanation of why he thinks it is not immoral to continue providing a useless service in return for a handsome salary financed by taxpayers.

VAD writes:

Bryan, looking back at your own education, I would be interested to hear what proportion was signalling and what proportion was human capital acquisition, and how did the proportions change?

N. Joseph Potts writes:

There's nothing wrong with education that getting the government out of it wouldn't fix. Nor anything wrong with employment and hiring practices that getting the government out of THAT wouldn't fix, either.

This stuff is SO EASY . . .

PrometheeFeu writes:

@Tom West:

"But the benefit of switching to a better system goes to the student, while the cost of having to learn a new system falls on the employer, so it isn't likely to happen.

I think this meets the strict definition of an externality (the prospective employee would gladly pay for the company's expense in using a new system, but there's simply no market to allow that to happen), but I honestly don't think anyone's got the appetite for a government intervention to knock the current system out of its sub-optimal equilibrium state."

I think it would be a market failure, but not an externality. However, I think you are underestimating the power of the market to jump out of such an equilibrium. The most straightforward way I can think of would be for the testing provider to pay companies to send their HR people to be trained in regards to the test. But I can imagine many other approaches such as already-reputed educational institutions offering the testing service. Sure, they would loose their business as a college but they would probably pick up the testing business for the whole population. But more generally, think of the huge immediate savings an individual company would see if they hired out of the pre-college market where there is a lot less competition... I think the reason why we are not seeing a switch over to cheap IQ testing is that college doesn't just signal your intelligence. There is something else going on.

ChrisGreen writes:

Its not so much a signal of the ability to conform to rules, its a signal of the ability to figure out what the rules are in the first place ("What is my professor really looking for in this paper?"). That ability can sometimes translate well into, "What does my boss want this report to contain?"

Duncan Frissell writes:

One difference between school and work is immediately apparent...

The first page of the book was written in 2007. It is almost 2012. Still no book. Academic productivity may be a bit low. The Authorized (King James) Version of the bible took 7 years. Better finish quick. I know they had more bodies...

I did Preschool through Law School without a break. I learned a lot of useful stuff and have used some of it. Mostly reading and writing. Since I was hooked early, I continued doing it. Math helped too though I didn't like it as much. My exposure to lots of different bits of information has helped me to produce gestalts when faced with new concepts. I don't know if college helped all that much. Certainly Primary and Secondary school are useful.

Too bad the US has decided to forgo Primary and Secondary education for most children.

Scott writes:
I truly would like to hear Bryan's explanation of why he thinks it is not immoral to continue providing a useless service in return for a handsome salary financed by taxpayers.

Ditto for me

EJD writes:

DISCIPLINED MINDS by Jeff Schmidt is highly recommended, along the lines of this post. (Schmidt was fired by his employer upon publication.)

Daniel Menes writes:

Bryan,

I think there is much truth in what you say, but my own experience suggests a valuable counterpoint.

I took a bachelor's degree in intellectual history and, having found this to be unremunerative, I completed most of the work for a second bachelor's in mathematics. Since then, I have achieved considerable success in the "real world," where I find that I use every day the concepts, facts and intellectual habits that I learned studying both math and history. Some of my colleagues tell me, however, that they do not use what they learned in college, and I have no reason to disbelieve them. Clearly their experience agrees more than mine with the thrust of your post.

So why do I benefit from what you call "magic" where others don't? The difference may be that I have pursued a different path to real-world success. Specifically, I offer the ability to look at problems from different, unexpected, and often surprisingly practical directions. This ability comes at the expense of some of the "conformity" measured by college grades.

Clearly there is a tension between a credential that measures conformity and an education that is most valuable to the non-conformist.

On the labor market, I have found my species of non-conformity to be both a hindrance and an advantage. The advantages have come to dominate as I have grown older and have competed for higher, more responsible positions.

Based on this experience, I have concluded not that higher education is useless, but that our institutions get the timing wrong. Most people could more usefully pursue such education on an ongoing basis, as an adjunct to their career, rather than as a four-year hiatus before beginning their career. I hold this conclusion particularly strongly with regard to "impractical" subjects such as history and the more abstruse topics in mathematics.

With this in mind, it is clear that our current system of public grants and subsidized loans creates precisely the wrong incentives, being designed around the assumption of a four-year full-time college program.

Asher writes:

@ PrometheeFeu

There's this little thing called the Civil Rights Act. When you use a measure that produces racially-disparate outcome you are required to demonstrate how the measure directly pertains to the job requirements. Case law is pretty clear that courts will not accept IQ as a measure of hire-ability. Sure, case law is wrong on the facts, but, then, law has an irritating tendency to trump reality.

WG writes:

What a breath of fresh air. This is a kind of commonsensical honesty that one rarely finds at any level of the education industry. It is time that we took a hard look at the huge investments of time and money that people devote to earning (and unsuccessfully trying to earn) academic credentials. Is it really worth it?

You might be interested in the "100 reasons NOT to go to grad school" blog for similar thoughts: http://100rsns.blogspot.com/

Mark V Anderson writes:

I agree that a college education is mostly signaling. I also disagree that college teaches anybody to think. It teaches students how to think in the manner proscribed by the particular subject one is in -- it doesn't teach how to think in a broad sense. Thus one learns how a sociologist thinks or an engineer thinks or an economist thinks. True creative thought has to come from the individual.

Many have said that education in a technical field really is about the learning. I suspect that even this is mostly untrue. I have a 4 year Accounting degree. I could have gotten the accounting knowledge I needed in a year or less. The rest is signaling. Based on my discussions with engineers, I suspect the same with them. Do engineers and scientists really pack four years of needed schooling into their degree, or do they really just have a dozen or so classes they actually use when they get Real Jobs?

On the other hand, my Masters in Tax was mostly content based. I took my classes at night and was able to use almost every class I took at my real job. So I do agree it is possible to get a true content-filled degree; it's just that it doesn't take nearly as long to do this as traditional degrees require.

Porkopolis writes:

Some colleges are tackling the "how to think challenge" head-on. Take for example the cross-disciplinary program at Lehigh University called Some colleges are tackling the "how to think challenge" head-on. Take for example the cross-disciplinary program at Lehigh University called South Mountain College:

South Mountain College is an academic residential program dedicated to the exploration of ideas across disciplines. Situated within the College of Arts and Sciences, SMC consists of a community of people interested in student-driven, immersive learning.

South Mountain students are challenged to assume responsibility for their educations and make connections across disciplinary boundaries. Assisted by core faculty—along with faculty and staff ‘friends’ of the program—South Mountain students draw from the curricular resources and intellectual expertise of the University as a whole...


Floccina writes:

IMO education and grading/signaling are at cross purposes and that grading often squeezes out education.

A separation of education and grading might help.

Also IMO education is very valuable to all humanity, schooling is far less valuable.

Diplomas are very valuable to the individual in the job market but not very valuable to all humanity.

Again a separation of education and grading might help.

More education and less effort spent on signaling would be good.

freemarketer writes:

Bryan, India hath need of thee! Your argument is very pertinent to the situation in my country where Tax payer's money is criminally wasted on subsidizing college education which imparts no useful skills .

The very sensible move in a couple of states to shut down departments offering dead-end degrees in subjects like philosophy, literature and history was successfully stalled by the fierce opposition from the teachers who would have lost their well-paid paid but socially useless jobs. The teachers who defend the dead-end degrees want their own children to enroll only in engineering or medical schools. Yet they insist that what they teach is as important as engineering or medicine!

I look forward to your book , although I am aware that policy makers in my country will continue to waste tax payer's money to support university departments which exist so that the students who enroll for their courses can claim they are doing something rather than nothing.

Eugene Dillenburg writes:

I must disagree with the following:

Think about all the time students spend studying history, art, music, foreign languages, poetry, and mathematical proofs. What you learn in most classes is, in all honesty, useless in the vast majority of occupations.

29 years ago, I earned my BA in Advertising. I never worked a day in the field. Instead, after bouncing around a few low-level jobs, I ended up at a science museum, planning exhibits, researching them, writing the text. Though I am not an author, the poetry and writing classes I took made me a better writer, which is something I use on my job every day. Though I am not an artist or curator, my art history classes exposed me to new ways of looking and thinking, and gave me a rich body of metaphors to help understand everyday issues. The smattering of introductory science courses -- chemistry, biology, astronomy, etc. -- ended up being directly relevant, even though I had no way of knowing I'd end up with my current job (and, in fact, spent many years desperately trying to land a different one). But the bulk of my classwork was in advertising, TV, radio, journalism, etc. -- in other words, communication. And, again, I have never worked a day in any of those fields. But I have spent my career communicating through exhibits. I simply transferred my training from one medium to another.

At 22, nobody knows what life is going to hold. Your best bet is to get a broad base of knowledge so you can be ready for anything.

Tom West writes:

Will Wilkinson writes:

I truly would like to hear Bryan's explanation of why he thinks it is not immoral to continue providing a useless service in return for a handsome salary financed by taxpayers.

That's easy. For many (most?) professors, their value is their research. Having to teach students is pretty much an accidental side product of history. If the teaching part turns out to be useless, it was a negligible part of the professorial service anyway.

At least that's the way I would defend it. Of course many taxpayers see colleges and universities as primarily teaching institutions, but it's certainly never been the way the majority of professors see it.

Jean writes:

I do agree that technical programs are valuable, and would be harder to teach on the job. But they can lean too much on the theoretical as well.

I remember having an argument with the Dean of Engineering at our local (good) Canadian university. I was in the position to hire co-op students from their faculty every term. (Co-op is a program where you alternate work and study terms. Very popular!) In our area, there was also a good community college which had a 2-year program in Electrical or Mechanical engineering.
Students from the college who had good grades at graduation could "bridge" into the 4-yr program at the uni.

Our company, and many if not most engineering firms in the area, vastly preferred co-op students who had come through the bridge over students who had only gone to the university. On paper, the 4-yr program kids were smarter - they got into university on the first go. But in reality, they were all but useless. The uni was frustrated that the "best" students, ie. those with the best grades all along, were in fact the hardest to place.

My point to the Dean was that you want a co-op student to do the grunt work, using a CAD program or circuit layout, testing circuit boards etc. They needed hands-on experience, which the college program gave them in spades. The uni kids didn't even touch a CAD program until 4th year!

The college kids weren't lacking in theory either, since they completed a six-month bridge program to make up the gap before starting 3rd year. But they had the practicality to actually build things. Best of both worlds.

Noah Yetter writes:

Education is less narrow than the (ivory tower!) literature leads you to believe.

My economics degree makes me a better software engineer.

Mik writes:

@ Tom West:

And, in comparison to teaching, do you mean that the social return of social science research is positive? I would say the probablity of that is lower than the probablity that their teaching has some positive social return.

riskasaurus writes:

@iamreddave

"Good students tend to be smart, hard-working, and conformist - three crucial traits for almost any job." Does the signalling theory explain why a student who went to an ivy league university would be ranked higher than one who went to a less prestigious one? The conformity should be the same in both unless Harvard is obviously more regimented. Smart could be tested with an IQ test much more quickly and cheaply then years of school.

The model would be useless if it couldn't. The significance of a signal is that it is costly information for the candidate to relay to a selector under imperfect information. The selector knows very little about both candidates, but DOES know that it is much harder to even get into Harvard--much less get a 4.0-- than any other school.

It is a much more costly effort for a job candidate to have gotten into Harvard and graduate with a 4.0 than another kid to do the same thing at a state school. The odds of an untalented person graduating 4.0 from Big StateU are much higher than an untalented person doing the same at Harvard. Literally anyone could tell a hiring investment bank that they'd be good at investment banking. Since investment banks get to pick from the top of the talent pool, they pick the kids that are sending the costliest signals of having high talent.

Harvard grads get to send the strongest signal of being talented, and that's why Goldman Sachs recruits there. 17yo's must know this because they kill each other each fall to get in to Harvard.

Student Media writes:

"Education is less narrow than the (ivory tower!) literature leads you to believe.

My economics degree makes me a better software engineer."

I am fully agree with Noah Yetter, I am an IT person but my 2 additional degrees in Mass Communication and Politics made me more powerful to write articles for my own sites

derskitaplari writes:

And, in comparison to teaching, do you mean that the social return of social science research is positive? I would say the probablity of that is lower than the probablity that their teaching has some positive social return.

rjp writes:

rickasaurus wrote: The model would be useless if it couldn't. The significance of a signal is that it is costly information for the candidate to relay to a selector under imperfect information.

Agreed.

The selector knows very little about both candidates, but DOES know that it is much harder to even get into Harvard--much less get a 4.0-- than any other school.

Agree with the part about harder to get into. Disagree about difficulty in graduating 4.0 being more difficult than state school.

It is a much more costly effort for a job candidate to have gotten into Harvard and graduate with a 4.0 than another kid to do the same thing at a state school. The odds of an untalented person graduating 4.0 from Big StateU are much higher than an untalented person doing the same at Harvard.

Disagree. A 4.0 is difficult anywhere, and at the state school a few students would have actually flunked out -- something nearly impossible to do at Harvard -- which would signal more significant grade inflation at Harvard, which would in turn mean more A and B graduates.

Literally anyone could tell a hiring investment bank that they'd be good at investment banking. Since investment banks get to pick from the top of the talent pool, they pick the kids that are sending the costliest signals of having high talent.

Harvard grads get to send the strongest signal of being talented, and that's why Goldman Sachs recruits there. 17yo's must know this because they kill each other each fall to get in to Harvard.

Actually that is slightly incorrect. The Harvard graduate will have developed relationships the state school student could not have developed ... access to capital via relationships.

In the list of top Quant schools by QuantNet, Harvard is no where to be found. (http://www.quantnet.com/mfe-programs-rankings/)
At #19 NC State and at #22 is the lowly Kent State. Quant school is where smart goes. MBAs from the Harvards are for salesmen.

Comments for this entry have been closed
Return to top