Arnold Kling  

College and Summer Camp

Mixed Signals: Why Becker, Cow... Posner on Health Care...

Bryan writes,

I've been defending the signaling model to other economists for 15 years, and always met fierce resistance. Frankly, I think that this resistance mostly stems from a failure of introspection. If you really compare what you learned in school to what you actually use in your job, the disconnect is too large to ignore.

The challenge is to reconcile two observations.

1. College graduates earn more than people who do not graduate from college.

2. Introspection suggests that what we learn in college is not very important on the job.

Then perhaps the relationship between earnings and a college education is one of correlation without causation. Perhaps people would prefer to spend age 18-21 on campus, at least if they have sufficient aptitude so that without too much effort they can avoid the unpleasantness of bad grades.

So college, like summer camp, is a consumption good. And you might observe just as high a correlation between earnings and summer camp attendance as between earnings and college attendance. But in neither case is there much causation.

If your parents are wealthy, and they can afford to send you to summer camp, then you are likely to have high earnings. Or, if you know, based on how well you do in high school, that you will have high earnings, you can afford to go to summer camp. So the people who go to summer camp (college) will tend to have high earnings, without there being a causal relationship.

My problem with the signalling model of education is that it suggests that there is a huge unexploited profit opportunity for employers and employees who can come up with alternative signals. And yet nobody tries to set up a system for identifying and hiring smart high school graduates. I suspect that is because very few talented high school students are willing to miss out on the equivalent of a summer camp experience.

I think that the "summer camp" model explains why colleges have done more in recent years to improve their amenities than to improve education. It may explain grade inflation, since you want to keep the campers happy. It may explain why rural small colleges have fallen out of favor, while universities with top-ranked basketball teams have become more popular.

Comments and Sharing

TRACKBACKS (6 to date)
TrackBack URL:
The author at Division of Labour in a related article titled Summer Camp U writes:
    Arnold Kling compares going to college to going to summer camp. He's onto something. Bascially I think a lot of people go to college because it's a fun thing to do. And unlike summer camp, you can drink and have... [Tracked on February 6, 2006 9:40 AM]
The author at Newmark's Door in a related article titled Signaling, summer camp, and Alchian's model writes:
    Bryan Caplan thinks the signaling model of education is underrated. Arnold Kling and Robert Lawson prefer the summer camp model. Two years ago I briefly noted Armen Alchian's model of higher education. I found it intriguing then, but if it's [Tracked on February 14, 2006 5:56 AM]
COMMENTS (20 to date)
John P. writes:

College as consumption -- FWIW, I loved college and law school, and I would gladly go back for still more of that experience if I could afford it.

Vorn writes:

Another thing about the introspective point that what you learn in college does not help you on the job is that it definitely depends on major.

As a computer science major (before law school), I can say that my experience has been exactly the opposite. When I was a software engineer, I was using skills I learned in college everyday. I am sure it is the same story with the hard sciences. A physicist or chemist or engineer would typically find the things they learn in college to be very helpful.

Dave writes:

Vern makes a good point; as a graduate electrical engineer (in the early fifties) there was a direct correlation between my college courses, especially the math and physics, and my career as an electronics-engineer-cum-businessman.

I also believe that the college experience in general helps form attitudes and manner of thinking, which should be worth dollars on down the line.

spencer writes:

Yes, the signaling function is part of the story.

But the very fact that starting salaries by major have almost a 100% range -- a recent chemical engineer graduate gets a starting salary almost double a recent liberal arts graduate -- implies that it is only part of the story. If all it was was signaling you shouldn't see the very wide range in market payments for new graduates.

David Thomson writes:

“When I was a software engineer, I was using skills I learned in college everyday. I am sure it is the same story with the hard sciences.”

I never mock the hard science Ph.D.. One almost certainly has to earn such an advanced degree. My general contempt is reserved for those of the soft science variety. I’m sure, for instance, that Harvard University’s hard science majors are fantastic. The intellectual whoring goes on in its liberal arts courses.

Foolish Jordan writes:

One thing I remember learning in college was that in the more-racist past, there were serious debates over whether Shakespeare really intended "Othello, the Moor" to be black. I certainly never used this fact in my career so far.

What I did do in my Shakespeare class was read a lot of books with slightly weird English and write a lot of papers. I don't really remember that as "something I learned in college", but certainly I do a lot of reading of things in slightly weird English and writing in my job!

Perhaps everyone just has a perception bias -- nobody thinks they use what they learned in college because they internalized it and no longer remember "learning" it.

caveatBettor writes:

3 of the 5 in my immediate family are hard science PhDs, and they are the biggest life buffoons, when measured say by the quality of their marriages or their network of friends. (I am the least educated in my family, and make more than the other 4 put together, although income doesnt measure much either).

Zac writes:

Vorn: I once talked to Bryan about the very issue you bring up. We concluded it would be a very good topic for a paper, but neither of us were convinced that engineering programs significantly challenged the signalling model.

I am a recent graduate and many of my friends majored in computer science, physics, chemical engineering, and the like. All of them have reported that while their classes gave them some necessary background, they are by and large being retrained with actual job skills. My friend the chemical engineer now works almost exclusively with a specific polymer to develop products with a medical purpose - this has as much to do with many of his classes as my job has to do with my courses in macroeconomics (nothing). My friends who are software engineers have all had to learn different skills - .NET programming for instance - that are largely different from the Java classes that made the core of their curriculum, and a far cry from the Assembly language classes that they were required to take.

And what about that student who decided to take another computer science class and not his general education course in Aristotle or Tolstoy? What do you think his job prospects are like? That illustrates that even if a small number of college majors involve significant vocational training, the primary purpose behind a college education is signalling your ability to be a good automaton, err, employee.

spencer: Chemical engineering is a much harder major than philosophy; it requires more memorization, longer hours, more labs, more intelligence. So it signals a better potential employee by all the metrics discussed in the singalling model. But there are smart philosophy majors, you say? Many would make good chemical engineers? Yes, but they chose to major in philosophy instead. That signals they don't care too much about the money they could make, they clearly aren't too concerned with becoming a chemical engineer. Nevertheless, philosophy is a popular major for pre-med students wishing to inflate their GPA. Do you care if your doctor majored in philosophy instead of getting that double in bio and chem? Do you think it really matters?

Arnold, your two observations are quite easily reconciled by the singalling model. As far as your concern about unexploited profit, this is because you assume what is being signalled is intelligence. As discussed, that is only part of the story. The other part is signalling your ability to work hard and conform. Clearly college is an effective signal for the qualities employers are looking for.

The principal problem is that education subsidies make the problem worse.

Vorn writes:


I can't speak about the other majors, only having been a computer science major / software engineer.

What you learn in computer science is not some trivial thing like how to program in .NET or in Java or whatever programming language is the latest fad. What you learn are the PRINCIPLES of programming. I can learn ANY programming language quite quickly. (BTW, if your friend learned Java in college, he practically learned .NET. .NET is object oriented and uses basically the same sort of concepts, just with different syntax and obviously different functions to call in the .NET library. )

Of course, in the field of programming, there are hacks and there are professionals. Hacks are basically people who do not transfer very well from one programming language to another. Such people demand lower salaries in the long run because they have difficulty adapting to an evolving marketplace.

Basically, in my experience, people who emphasize learning particular languages, like .NET or ASP or Java and complain they weren't taught a particular programming language in class are usually, but not always, hacks. It just isn't that big of a deal to learn particular languages, once you know the principles of computer science. Computers are machines that run according to certain principles and all programming languages need to conform to those principles. Then there are principles of design. If you understand the principles, particular manifestations of those principles in the form of particular programming languages just are not that big of a deal. We all have our favorites, and different languages are optimal for different tasks.

Anyway, that is my take. I didn't major in computer science to learn Java. I went to learn the principles of computer science. As a result, I am an expert in Java, C++, MATLAB and many other programming languages and tools, as a matter of course. If I so desired, I could pick up any programming language fairly quickly. Just because I have to learn C# (a major rip-off on Java and a .NET langauge) for a particular project doesn't mean that I am not using what I learned in college.

Computer science is a hard major; there are some people who barely scrape by without really learning the core principles. It is probably this type who emphasize knowledge of particular languages and particular technologies; it is simply easier to grasp a particular language or technology than a principle since one is concrete and particular and the other more abstract and general.

Alcibiades writes:

In my own case, the fact that my parents paid the $40,000+ per year for me to attend a private college had a big impact on my decision to go there. I doubt I'm alone. Additionally, the myth of that says college-is-essential-and-special compared to all other experiences is exceedingly strong. My parents, and every other adult I mentioned it to, thought me nuts for wanting to graduate in an accelerated 3 years.

/typing from a neuropsychology class. It remains to be seen whether the knowledge I am currently not absorbing will be significant in my future.

Bob Hawkins writes:

I had a bunch of proposals to review this weekend. So I did it while the Super Bowl pregame and game were on. That was a skill I perfected in college, and I'm still using it.

I learned how to deal with a bureaucracy at Ohio State and the University of Denver. Different kinds of bureaucracy: OSU was big and relatively efficient, DU smaller and very inefficient.

Really, if you spend several years at college, and don't learn anything you can use later, I want a sample of your skull.

Michael writes:

I think you are for the most part right, but, at least from my perspective, the difference in earnings between college grads and non-college grads is (except for specialized fields, like law or medicine) based on reputation. While I have no clue whether statistics back this up, I wouldn't be surprised to see graduates from schools that are stereotyped as "party schools" making about as much as graduates from other colleges.

Glen Raphael writes:

I learned all sorts of things in college. For instance, I learned to play guitar and found a better way to fold socks. It's impossible not to learn a few useful things over the course of several years. But who's to say you couldn't have learned a similar amount of stuff that "you could use later" given the same time investment in something other than college? Say, summer camp and a series of job internships?

Dan Landau writes:

There are still a few problems with the signaling model, even if a lot of people feel what they learned in college was worthless. I by the way don’t, I got a terrific start in economics as an undergrad that helped a lot in graduate school. Maybe there is something wrong with me, but I teach many things I learned at both levels.

The signaling model ignores the question of learning how to learn. I don’t think very many people come out of high school with that capability. That is the most important thing learned in college.

The signaling model also implies economics is worthless. You the employer pay the college grad a lot more because he spent so much proving he is a better employee. Why don’t you pay 3 or 4 people to develop a battery of tests, paper and activity tests, which will be able to distinguish who has the qualities signaled by a college degree and who doesn’t. Then with these tests you find high school grads who can do the work of college grads. Since they are only high school grads, you will get them at a much lower salary. I haven’t heard of this being done. Either economics is stupid or signaling is wrong.

Dr. T writes:

I think the signaling model is appropriate for the majority of colleges and majors. I spent 3 years on a medical school admissions committee, and we looked at college in that way. Quite a few required pre-med courses had little relevance to the study of medicine (calculus is the best example), but the ability to get through a year of calculus with an A or B average signaled that the student had a combination of intelligence, learning skills, and perseverence that predict success in medical school.

I agree that hard science, engineering, and computer majors learn much that will be relevant to their jobs. But, most successful students in those fields could master the needed fundamental and specialized information with a year or two of independent learning followed by on-the-job training. What they would lack is a college degree, which signifies that the person has qualities associated with successful employees. Is that really necessary? I don't think so, but (so far) businesses do.

rakehell writes:

Landau: If you never "learned how to learn" how would you have completed your degree?

A battery of tests might assess intelligence, but not motivation or productivity. There is also the issue of training. Firms don't want to train workers because of raiding.

Bill writes:

The signaling model ignores the question of learning how to learn. I don’t think very many people come out of high school with that capability. That is the most important thing learned in college.

If true, this indicates to me that K-12 education is mostly time wasted for most people. I learned how to learn before I reached puberty. Teachers were mostly wasted on me. On the other hand, good textbooks were quite useful.

Brad Hutchings writes:

Signalling or not, perhaps the main value of a college education (or summer camp) is what you weren't doing instead. The college experience teaches frugality in a controlled environment. It may be selection bias at countless levels, but my college educated friends pretty much stay away from severe stupidity, while my uneducated friends have risked the rent money on all sorts of folly.

It would be interesting to look at earnings mobility and job mobility of people over, say the past 20 years, and control for whether people were in school at time of employment. I suspect you'd find that waiters who were not going to school remained waiters. Same with retail sales. But if they were going to school, then I'd suspect the get out and move up. Maybe even look at types of jobs people applied for. I bet the uneducated stuck in their niche, while those who were in school saw their positions as temporary and applied for other things.

Chris writes:

It almost seems like College is like an HR department, where what you learn is not as important as the "screening" they do for you. Could explain why they really don't choose the talent from High Schools and why some employers have preference with particular schools. It's a thought.

Dr. Drey writes:

Here's a potential additional, albeit likely small and incomplete, explanatory variable:

College is a consequential labor market distortion induced by cartelization. This may be most obvious in the requirements of cartels like the American Bar Association (ABA). You have to go to Law School, whether you want to or not -- whether its useful or not.

So why don't firms circumvent this and have lawyers learn on the job, as they used to decades ago? Too risky for the employees. Imagine joining the only non-ABA firm: if it goes bankrupt, you get fired or you hate it and wish to join another firm, you can't. I doubt any salary could compensate for that risk.

Could this be generalized into a broader explanation? Perhaps. More generally, the labor market forms an "effective cartel" from a kind of network effect/sunk cost phenomenon. Those with degrees have incentives to protect their value, so they won't hire those without. Once again, you can't circumvent the cartel (network) because if you need to change jobs, you can't.

That said, I'm an economic amateur. What do you guys think?

Comments for this entry have been closed
Return to top