Bryan Caplan  

Credentialism versus Signaling

Gresham's Law of Comments... Credentialism Trap Rejoinder...

Arnold's not too happy about credentialism:

I remember hearing a company founder in the Dotcom era saying that he only wanted to hire MBA's from top schools. I thought to myself that this was silly. When you are a start-up, you need to find people who are better than their credentials. The last thing you can afford to do is pay a premium for credentials. But this guy, who had an MBA from a top school, was evidently uncomfortable around people with more talent than credentials.

On the surface, Arnold's complaint and the signaling model of education have a lot in common. Both positions doubt that fancy schools teach much that you need to know in the real world.

However, there is a critical difference between the two views. If Arnold's story is right, there wouldn't really be a trap. Some firms are bound to be run by people who value profits more than pride in their alma mater. These firms will do what Arnold recommends: Hire talented people regardless of what diplomas they hold, and make money hand over fist. Faced with such competition, firms run by Harvard men for Harvard men would financially bleed to death.

Think this "just about to happen"? Don't hold your breath.

In contrast, if the signaling model is right, the Harvard men are, on average, really worth the extra money. Maybe not in every case - Arnold makes a good point about start-ups. But according to the signaling story, the workers who jumped through the toughest academic hoops tend to have the Right Stuff. An idealist who hired based on "talent, not credentials" is destined for disappointment. Simple question:

"If he's so talented, why did he take the easy way out and go to Podunk State?"

"But he's a genius!" you reply?

"Well, then it should have been child's play for him to excel at Harvard. Why didn't he try?"

If, like me, you're tempted to answer, "Because he's a free spirit who didn't care for Harvard's attitude," you've hit the pay dirt of the signaling model. Free spirits are an inspiration, but they rarely make good employees.

COMMENTS (24 to date)
dearieme writes:

There's an old saw concerning the degree classifications used in Britain. "First class men hire first class men and second class men hire third class men." This suggests that whatever correlation you seek, don't expect it to be linear.

Chris writes:

There are many valid reasons why a very talented, bright person would choose not to go to a top notch school. Money and location merely being the easiest examples.

In my experience, the only firms that hire based on credentials are the ones more concerned with their image than their performance.

Vorn writes:


If one were so bright, woudln't they realize that the difference in tuition is one that will easily be made up by pay differentials and increased opportunity resulting from going to a more prestigious university over a less presitigious one. With respect to location, wouldn't a very ambitious person often make a sacrifice for increased opportunity?

While I wouldn't dispute your suggestion that in particular instances, very bright people may have good reasons not to go to a better school, I think typically, they will jump at the opportunity. Thus the high correlation between brightness and the level of prestige of a university and the ability of such universities to play the signaling function mentioned by Byran.

Bill Stepp writes:

True story:

In 1990 I saw an ad in the Wall Street Journal for a trading position (the name of the firm wasn't mentioned). Curious which firm it was, I called the phone number in the ad. The person I spoke to asked what school I graduated from.
I told him, and he said they wanted only someone from an Ivy League school, and wouldn't talk to anyone else.
I don't know what happened to the firm, but I couldn't find a record for it after looking a couple years ago.

marco writes:


I think that Arnold is on target. The logic he offers is both conceptually and practically correct. All things being equal one would always hire talent over credentials.

Let's see how this may work within the world you offer. Let's say that we have two candidates who graduated from some prestigious university with the same degree, work experiences and grades (I do this to make it easy for you to see the point that Arnold is making given that you wish to over analyze a simple point). Which one should the employer hire? I won't wait for you to reply and go ahead and say the right answer the one with more talent (across all dimensions).

In the real world consulting firms and investment banks practice the fine art of credential-based hiring. Why? Three reasons:
(1)they have been successful with the approach;
(2)signaling value to clients that what they offer is of great value;
(3)(3) hire people like you (it's comfortable).

The fact that credential-based hiring has produced success does not answer an important question that a scientifically inclined individual would ask. How do those that are not hired perform relative to both those were hired and succeeded and those who failed. In order to understand this question we need to enumerate the appropriate cells – hire/successful, hire/unsuccessful, not hire/successful, not hire/not successful. Having worked in a very large prestigious consulting firm that practices this discipline, I can attest we were surprised what we found. It was not what we expected. The old boy network still rules for signaling purposes.

Dan Landau writes:

Theorizing about real learning versus signaling is not too hard. Empirical study is lot more complex. First of all, what are the students getting? Is it only things you can get from a book? Second, l how much does it differ between MBAs and professors? It is fun to throw examples at the theory, but it isn’t going to prove much. For example where did the Nobel Prize winners go to grad school? Will the answer prove anything about your average top 50 university professor?

aaron writes:

In addition to the credential, they are also in "the club". They are often attached to a priveliged and talented network.

Silas Barta writes:

Another problem with Kling's theory is that getting a degree doesn't signal support for credentialism at all. It could also mean you're going to "do what it takes to get by, even if you don't disagree with the system", similar to how, as libertarians note, paying taxes doesn't mean you support taxation-funded government. Of course, that you are willing to "do what it takes to get by, even if you don't disagree with the system" is also valued by employers, so Kling's right that there could be a kind of self-reinforcing trap here, but it's not because of credentialism.

rakehell writes:


"Ronald G. Ehrenberg, director of the Cornell Higher Education Research Institute in Ithaca, New York, insists, “on average, there is significant gain in going to a top private school,” both in access to better graduate schools and in higher lifelong earnings." (I highly recommend Ehrenberg’s labor econ textbook).

NERA posted a study that found that income correlated strongly with high school grades and SATs, regardless as to where one went to college. I take those results with a grain of salt. Perhaps income correlates, but there may be many other factors are not captured: choice of career, choice of subfield, working conditions, location, etc. With a top notch credential behind your name, you may be able to pursue a career in a field that has average lower salaries, knowing that that you have a signal that will give you more job and career security.

As for out-of-class learning, I think that connections are important. At the better schools I have attended, I simply had more in common with my colleagues and formed more lasting friendships. My first two jobs out of college I got were from alums and the experience I got there led to another position. I also think “socialization” (in the social scientific sense of the word) is important. My colleagues at the elite schools were much more disciplined and had a more confident “can do” attitude. They were better role models. They also had a better knowledge of labor markets, where the good opportunities were, and planned their careers with more care.

PJens writes:

In practice, I think employeers use both cedentials and signaling to make hiring decisions. Especially in entry level positions. It is not an either or situation most of the time.

I also think that by the second, or third job, previous work performance is more valuable v. a degree from school X.

Dr. T writes:

There seems to be an incredibly strong belief among the authors and posters on this site that independent thinkers or free spirits or nonconformists make bad employees, and that it is important to use 'signalling' or 'credentialling' to decrease the likelihood of hiring such folks. My questions are: 1. What kinds of employment are you people referring to? 2. Have any of you actually been bosses?

I don't know about the rest of you, but I find that having flock of sheep employees results in a crappy work environment with low productivity; lots of bureaucratic nonsense; and markedly decreased creativity, problem-solving, innovation, etc. I want at least a few employees who think outside of the box, who do not automatically check which way the wind is blowing before moving, and who don't always fit in with the majority. And, I don't want to work for anyone who thinks all employees should be conformists.

Bill writes:

If one were so bright, woudln't they realize that the difference in tuition is one that will easily be made up by pay differentials and increased opportunity resulting from going to a more prestigious university over a less presitigious one.

Brightness has nothing to do with it. Knowledge of the signaling nature of education is much more important. A person with a genius IQ is no more than a wild animal without the accumulated knowledge of civilization.

In my case, I chose San Francisco State University over UC Berkeley for financial reasons. I came from a poor family, and I received almost no help from them for my education. No one in my family had a university degree. At the time, I was unaware of the added value of a degree from Berkeley. Now, I'm self-employed and do well, but I have considered going back to school. I would certainly choose a big name school if I choose to return. Why? Now I'm aware of the value of a school's name. I'm no brighter, just more knowledgeable.

Personally, I hate credentialism and signaling. I'd burn my degree tomorrow if these were replaced with a system of testing for IQ, appropriate knowledge, and the ability to "work well with others".

Chris writes:

Vorn, Going to a prestigeous school isn't a free check for higher salaries out of school - so no an intelligent person would not automatically go for the credentials. Additionally, there is a signifigant price difference between Ivy league schools and state colleges, especially if you have to take travel into consideration. That price difference isn't something that everyone can finance.

Don writes:

Y'all might bear in mind that Krueger/Dale QJE calls into question the entire premise of this discussion.

Don writes:

On second thought, the paper itself doesn't question the premise of this discussion. But if folks are aware of this evidence (and if the result is supported with additional empirical evidence), then poof.

Silas Barta writes:

Don: could you elaborate on this article? Where can it be found? What relevant information is in it?

Taeyoung writes:


If, like me, you're tempted to answer, "Because he's a free spirit who didn't care for Harvard's attitude," you've hit the pay dirt of the signaling model. Free spirits are an inspiration, but they rarely make good employees.

and Dr. T's comment, I think "free spirit" is here kind of a euphemism for "lazy." Even if you're a genius, getting into Harvard requires jumping through a lot of hoops, nowadays, mostly involving racking up extracurriculars or experiences to pad your resume sufficiently that your 4.0 and 1600 SAT stand out from the pack of 4.0's and 1600s. And plenty of extremely smart people don't have the grades to get into a top school anyhow, because high school grades are often based primarily on busy-work (homework), and lazy people / "free spirits" find the tedium of busy-work oppressive.

"Free spirits" may, depending on what industry you work in, be excellent investments, as employees, but I think it's much more common for brilliant people who do not attend a top school (not necessarily Harvard, mind, or even an Ivy league or Ivy league-wannabe) to be lazy, with poor work habits, than for them to be "free spirits who didn't care for Harvard's attitude" or the attitude of any other prestigious or academically challenging school. Maybe that's what they tell themselves, but I don't think that's what the rest of us, who probably know someone like them, actually think.

Don writes:

Silas, the article is QJE a couple of years ago. It looks at income as a function of school quality, but uses interesting and novel control variables. Everybody knows that Harvard grads earn more, but is that because of learning while at Harvard and/or networking contacts made there, or is it because Harvard gets kids who are smarter in the first place? It turns out that once you control for characteristics unobservable to the econometrician but observable to admissions directors, the return to school quality goes away. That is, kids who got into Harvard but went elsewhere earn just as much as kids who went to Harvard.

It's not conclusive, but it is important and interesting.

David Thomson writes:

“Free spirits are an inspiration, but they rarely make good employees.”

I completely agree. An employer should think twice before hiring a “free spirit.” This individual may be able to offer an insightful intellect and an iconoclastically beneficial approach to the never-ending challenges of the business world---but at very high price. Alas, geniuses are sometimes hell to work with on a daily basis. Is the risk worth it? This is a question that best not be ignored. One abstractly wants to find someone who is not inclined to place their wet finger in the air to see which way the wind blows. But sometimes, in the real world, you are better off hiring an employee who keeps their mouth shut and merely goes along to get along.

Credentials are admittedly very important. There’s no way to get around that harsh fact. A degree from a leading university signifies a certain intelligence level, maturity, and work ethic. The employer, however, is foolish to limit their selection to these applicants. General rules should never become absolutes.

Dr. T writes:

Taeyoung is under the mistaken impression that getting into Harvard is academically, socially, and logistically difficult. One does not need 4.0s and 1600 SATs to get into Harvard. One does not need lots of extracurricular activities to get into Harvard. What one needs are halfway decent grades, connections (political, alumni, or business), and either money or minority status.

In my comments above I was not responding to the "free spirit" comment. I was responding to earlier postings about the use of signaling to find employees who were conscientious conformists.

Collin writes:

marco writes:

I think that Arnold is on target. The logic he offers is both conceptually and practically correct. All things being equal one would always hire talent over credentials.

All things are most likely not equal -- the cost of discerning one's talent is probably much greater than discerning one's credentials. By assuming away the cost of discovery it seems that you have avoided the issue.

I believe that the signaling model holds true whenever the costs of finding out enough about one's "true talent" is prohibitively costly relative to a close-enough substitute (here credentials).

If it's significantly easier to just look at a transcript than it is to conduct a series of interviews, personality tests, intelligence tests, etc., and if it ends up with the same cracker jack chance of a "good" employee, why not?

Rick Stewart writes:

Assuming Harvard can administer an IQ test less expensively than I can as the employer seems somewhat farfetched to me, and there is a greater correlation between IQ and successful hiring than between any of the other 'normal' hiring considerations (interview, previous experience, recommendations, degree, etc.).

So why can't we get 'certified' IQ scores sent to prospective employees who merely start at the top and make job offers until someone accepts?

Of course there would be hiring failures, but we already have hiring failures. And virtually eliminating the cost of hiring would be an enormous boon to not only individual firms but the entire economy.

Edward O'Connor writes:

Free spirits are an inspiration, but they rarely make good employees.

This reminds me of Eric Sink's reply to Paul Graham (re: great hackers): Great Hacker != Great Hire

Ronny Max writes:

Those who went to an Ivy League school proved they acquired much knowledge. Those who succeed in startups proved they utilized other people’s knowledge. It’s the difference between an expert and a leader.

Comments for this entry have been closed
Return to top